Journal of Scientific Exploration- Volume 14: Number 1-4

March 27, 2018 | Author: SonyRed | Category: Time Series, Stochastic Process, Markov Chain, Statistics, Randomness


Comments



Description

JOURNAL OF SCIENTIFIC EXPLORATIONA Publication of the Society for Scientific Exploration Volume 14, Number 1 Page 1 19 35 53 73 91 Research Articles Investigating Deviations from Dynamical Randomness with Scaling Indices Valentich Disappearance: New Evidence and a New Conclusion Protection of Mice from Tularemia Infection with Ultra-Low, Serial Agitated Dilutions Prepared from Francisella tularensis-Infected Tissue The Correlation of the Gradient of Shannon Entropy and Anomalous Cognition: Toward an AC Sensory System Contributions to Variance in REG Experiments: ANOVA Models and Specialized Subsidiary Analyses Publication Bias: The "File-Drawer" Problem in Scientific Inference Essay Remote Viewing in a Group Setting Harald Atmanspacher and Herbert Scheingraber Richard F. Haines and Paul Norman Wayne B. Jonas and Debra K. Dillner Edwin C. May, S. James, I? Spottiswoode, and Laura V Faith R. D. Nelson, R. G. Jahn, Z H. Dobyns, and B. J. Dunne Jefsrey D. Scargle 107 Russell Targ and Jane E. Katra 115 Guest Column The Sovereignty of Science Book Reviews Best UFO Cases-Europe by Illobrand von Ludwiger The Last Laugh by Raymond Moody The Discovery of the Cold Fusion Phenomenon by Hideo Kozima The Truth in the Light by Peter Fenwick and Light and Death by Michael Sabom Psychedelic Drugs Reconsidered by Lester Grinspoon and James B. Bakalar The Meaning of Consciousness by Andrew Lohrey Erratum 121 124 129 132 135 137 139 Richard E Haines Robert Almeder John O'M. Bockris Ian Stevenson Charles Eisenstein John Belof 141 145 149 SSE News Items Nineteenth Annual SSE Meeting Announcement Fifth Biennial SSE European Meeting Corrected Index for Volume 13 [sji JOURNAL OF SCIENTIFIC EXPLORATION A Publication of the Society for Scientific Exploration Volume 14, Number 2 Page 163 195 217 233 Research Articles Overview of Several Theoretical Models on PEAR Data The Ordering of Random Events by Emotional Expression Energy, Fitness, and Information-Augmented Electromagnetic Fields in Drosophila melanogaster A Dog That Seems to Know When His Owner Is Coming Home: Videotaped Experiments and Observations York H. Dobyns Richard A. Blasband Michael J. Kohane and William A. Tiller Rupert Sheldrake and Pamela Smart Essay 257 275 What Can Elementary Particles Tell Us About the World in Which We Live? Modern Physics and Subtle Realms: Not Mutually Exclusive Ronald A. Bryan Robert D. Klauber 281 282 287 289 293 295 296 Book Reviews The User Illusion: Cutting Consciousness Down to Size by Tor Ngrretranders Larnarck's Signature by Edward J. Steele, Robyn A. Lindley, and Robert V. Blanden; Epigenetic Inheritance and Evolution by Eva Jablonka and Marion J. Lamb Authentic Knowing: The Convergence of Science and Spiritual Aspiration by Imants Baruss Alien Abductions: Creating a Modern Phenomenon by Terry Matheson At the Threshold by Charles F. Ernmons Minds in Many Pieces: Revealing the Spiritual Side of Multiple Personality Disorder by Ralph B. Allison Earth Under Fire: Humanity's Survival of the Apocalypse by Paul LaViolette SSE News Items Arnold L. Lettieri, JK Michael Levin James A. Scott Damien Broderick Charles Eisenstein Robert E Creegan John L. Petersen JOURNAL OF SCIENTIFIC EXPLORATION A Publication of the Society for Scientific Exploration Volume 14, Number 3 Page 303 304 Tribute Editorial Peter A. Sturrock Henry Bauer I 2000 ~ 307 353 365 383 Research Articles Plate Tectonics: A Paradigm Under Threat The Effect of the "Laying On of Hands" on Transplanted Breast Cancer in Mice The Stability of Assessments of Paranormal Connections in Reincarnation-Type Cases ArtREG: A Random Event Experiment Utilizing PicturePreference Feedback Can Population Growth Rule Out Reincarnation? A Model of Circular Migration 411 David Pratt William E Bengston and David Krinsley Ian Stevenson and Jiirgen Keil R.G. Jahn, B. J. Dunne, K H. Dobyns, R.D. Nelson, and G. J. Bradish David Bishai 421 431 Commentary The Mars Effect is Genuine: On Kurtz, Nienhuys, and Sandhu's Missing the Evidence Bulky Mars Effect Hard to Hide: Comment on Domrnanget's Account of the Belgian Skeptics' Research Essay What Has Science Come to? Letters to the Editor Is the Activity of the Cerebral Cortex a "Brake" on the Functioning of the Brainstem? Understanding the Nature of Racial Prejudice: Comment on Hoyle and Wickramasinghe's article with response Book Reviews With the Tongues of Men and Angels: A Study of Channeling by Arthur Hastings Reaching for Reality-Seven Incredible True Stories of Alien Abduction by Constance Clear Project Mindshift: The Re-Education of the American Public Concerning Extraterrestrial Life, 1947 to the Present by Michael Mannion Cryptozoology A to Z: The Encyclopedia of Loch Monsters, Sasquatch, Chupacabras, and Other Authentic Mysteries of Nature by Loren Coleman and Jerome Clark (continued on next page) Suitbert Ertel and Kenneth Irving Suitbert Ertel 447 Halton Arp 1 I 455 458 Jesse Hong Xiong Richard S. Hahn 461 463 465 Stephan A. Schwartz Robert Nordberg Alexander Zmich I 468 Larissa Vilenskaya 1 JOURNAL OF SCIENTIFIC EXPLORATION A Publication of the Society for Scientific Exploration Volume 14, Number 4 Page 497 Editorial 2000 499 557 571 583 Research Articles Mind/Machine Interaction Consortium: PortREG Replication Experiments Unusual Play in Young Children Who Claim to Remember Previous Lives A Scale to Measure the Strength of Children's Claims of Previous Lives: Methodology and Initial Findings Reanalysis of the 1965 Heflin UFO Photos R. Jahn, J. Mischo, D. Vaitl, et al. Ian Stevenson Jim B. Tucker Ann Druffel, Robert M.Wood, and Eric Kelson Joel M. K a u B a n Joop M. Houtkoopel; Dieter Vaitl, and Ulrich Timm Stefan Schmidt and Harald Walach Philip J. Klass 623 643 Should You Take Aspirin to Prevent Heart Attack? Letters to the Editor A Further Note on Walach and Schmidt's "Empirical Evidence for a Non-Classical Experimenter Effect" Reply to " Further Note on Walach and Schmidt's A 'Empirical Evidence for a Non-Classical Experimenter Effect"' Response to "Valentich Disappearance: New Evidence and a New Conclusion," by Richard F. Haines and Paul Norman Authors' reply to Philip Klass 644 646 647 Richard F: Haines and Paul Norman Michael Levin Henry H. Bauer J. O'M. Bockris Alexander Imich Robert L. Raymond John Belof Robert F: Creegan 649 654 657 659 662 663 664 Book Reviews Mysterious Flame: Conscious Minds in a Material World by Colin McGinn Dancing Naked in the Mind Field by Kary Mullis Consciousness by D. Rakovic and D. Koruga Searching for Eternity: A Scientist's Spiritual Journey to Overcome Death Anxiety by Don Morse Gardner's Whys & Wherefores by Martin Gardner Paranormal Experience and Survival of Death (Suny Series in Western Esoteric Traditions) by Carl B. Becker Scientific Development and Misconceptions through the Ages, A Reference Guide by Robert E. Krebs SSE News Item 665 669 Twentieth Annual SSE Meeting Announcement and Registration Form Index: Volume 14 r i JOURNAL OF SCIENTIFIC EXPLORATION s A Publication of the Society for Scientific Exploration (ISSN 0892-3310) Editor-in-Chief, Bernard Haisch*, Henry Bauer** Executive Editor, Marsha Sims*, Managing Editor, Sarah Cunningham*" * For material published through Vol. 14, No. 2 and for several articles in later issues. * * For all new material Associate Editors, Stephen Braude, Dean Brown, Michael Epstein, Dean Radin, and Mark Rodeghier Editorial Office and Manuscript Submission: Journal of ScientiJic Exploration, Attn: Sarah Cunningham Allen Press, 8 10 East 10th St. Lawrence, KS 66044 ([email protected]); 1-800-627-0629 ext. 284; Fax: 785-843-1274 World Wide Web - http://www.jse.com Society for Scientific Exploration Website - http://www.scientificexploration.org Executive Book Review Editor, Michael Epstein Analytical Chemistry Division, MS 8391, National Institute of Standards and Technology 100 Bureau Drive, Gaithersburg, MD 20899- 0001 ([email protected]) Associate Book Review Editor, David Moncrief Editorial Board Dr. Mike1 Aickin, Center for Health Res., Kaiser Permanente, Portland, OR Prof. RCmy Chauvin, Sorbonne, France Prof. Olivier Costa de Beauregard, University of Paris, France Dr. Steven J. Dick, U. S. Naval Observatory, Washington, DC Dr. Peter Fenwick, Institute of Psychiatry, London, UK Dr. Alan Gauld, Dept. of Psychology, Univ. of Nottingham, UK Prof. Richard C. Henry (Chairman),Dept. Physics &Astronomy, Johns Hopkins Univ. Prof. Robert Jahn, School of Engineering, Princeton University Prof. W. H. Jefferys, Dept. of Astronomy, University of Texas Dr. Wayne B. Jonas, National Institutes of Health, Bethesda, MD Dr. Michael Levin, Cell Biology Dept., Harvard Medical School Dr. David C. Pieri, Jet Propulsion Laboratory, Calif. Inst. Technology, Pasadena, CA Prof. Juan Roederer, University of Alaska-Fairbanks Prof. Kunitomo Sakurai, Institute of Physics, Kanagawa University, Japan Prof. Ian Stevenson, Health Sciences Center, University of Virginia Prof. Peter Sturrock, Ctr. for Space Science &Astrophysics, Stanford University Prof. Yervant Terzian, Dept. of Astronomy, Cornell University Prof. N. C. Wickramasinghe, School of Mathematics, Univ. College Cardiff, UK SUBSCRIPTIONS AND BACK ISSUES: Please use the order forms in the back. COPYRIGHT: Authors retain the copyright to their writings. However, when an article has been submitted to the Journal of Scientific Exploration for consideration, the Journul holds first serial (periodical) publication rights. Additionally, the Society has the right to post the article on thc Internet and make it available via electronic as well as print subscription. The material must not appear anywhere else (including on an Internet Website) until it has been published by the Journal (or rejected for publication). After publication in the Journal, authors may use the material as they wish but should make appropriate reference to the prior publication in the Journal, for example: "This paper (material) first appeared in the Journal of Scientific Exploration, vol ... no... pp... under the title... ." Journal of Scientific Exploration (ISSN 0892-3310) is published quarterly in March, June, September and December by the Society of Scientific Exploration, Allen Press, 810 East 10th St., Lawrence, KS 66044. Private subscription rate: US $50.00 per year (plus $5.00 postage outside USA). Institutional and Library subscription rate: US $100.00 per year. Periodical postage paid at Lawrence, KS, and additional mailing offices. POSTMASTER: Send address changes to: Journal of ScientiJic Exploration, Allen Press, 8 10 East 10th St., Lawrence, KS 66044. Journal of Scientijic Exploration, Vol. 14, No. 1, pp. 1-18,2000 0892-33 10100 O 2000 Society for Scientific Exploration Investigating Deviations from Dynamical Randomness with Scaling Indices Institutfir Grenzgebiete der Psychologie Wilhelmstrasse 3a, 0-79098 Freiburg, Germany and Max-Planck-Institut fur extraterrestrische Physik Giessenbachstrasse, 0-85740 Garching, Germany Max-Planck-Institut fiir extraterrestrische Physik Giessenbachstrasse, 0-85740 Garching, Germany Abstract- The information contained in any given experimental time series can be utilized more exhaustively when transition probabilities between states rather than state probabilities alone are studied. Using advanced techniques of time series analysis, it is shown that deviations from dynamical randomness indicate evidence for unexpected temporal correlation features in selected data sets taken from a mind-matter experiment conducted at Freiburg (Germany). The techniques of analysis and a proper error estimation are briefly described, and some preliminary first results are presented. They encourage further inquiry into processual aspects of deviations from randomness in addition to more straightforward analyses of state probabilities. Keywords: transition probabilities - scaling indices - deviations from randomness - temporal correlations in mind-matter data. 1. Introduction Among other problem areas in the general framework of mind-matter research there is the basic question as to whether there are presently unknown relationships between physical systems and mental activities of human agents. Avoiding any speculations with respect to specific concepts of correlations, interactions, and causal or other influences between physical systems and mental systems, the mere formulation of this question presupposes a minimal methodological (not ontological) dualism of mind and matter without which it could not even be phrased. It has to be specified in order to lead to researchable problems. The special type of question addressed here is whether the randomness of a physically random system changes when human agents are asked to carry out certain "intentional tasks" that are defined so as to correspond to deviations from randomness in the behavior of the system. Randomness is itself a concept that needs to be considered in more detail. First, there is the much discussed question of whether randomness or chance is 2 H. Atmanspacher & H. Scheingraber an "ontic" property of the material world, without any reference to the knowledge of observers, or whether it is an "epistemic" expression of our limited knowledge about an ontic world. Another most obvious problem with which any working scientist is familiar arises from the fact that for empirical purposes the mathematical, measure theoretic definition of probability a la Kolmogorov has to be backed up by an interpretation in terms of a limit of infinitely many independent realizations of an event. In practice, this is impossible to achieve. We always deal with finitely many events, they are never identical in every respect, and it has to be checked carefully whether and in what sense they are really independent of each other. The finiteness of any empirical collective of events has the consequence that many mathematical theorems about probabilities (based on the limit N -+ o o ) must be applied with caution. For instance, in the area of complex systems research, many examples are known for which limit theorems (e.g., laws of large numbers) are not naively applicable and ergodicity must not be implicitly presupposed. Novel approaches and techniques of modern statistics such as second order statistics or large deviations statistics (Atmanspacher, Rath, and Wiedenmann, 1997) offer new insights beyond a standard characterization of a distribution in terms of its first two moments. For appropriate methods of analyzing finite collectives in such situations, it has been proven necessary to characterize the concept of randomness with reference to the range of N with which one is empirically dealing. Another important topic is the independence of events as members of a random collective. It is well known that distributions of states of a system can be perfectly random in the sense that they are perfectly represented, e.g., by a Gaussian distribution, but nevertheless there can be non-random features, e.g., correlations, as far as the transition probabilities between states (i.e., the dynamics of the system) are concerned. In such a case, individual events are not independent. They depend on their prehistory and show significant temporal correlations. Correlations are significant if they deviate (at a level to be specified) from the amount of correlations expected due to the finiteness of the random collective. A number of measures that characterize and distinguish these different kinds of randomness have been analyzed and compared with each other by Wackerbauer et al. (1994). For an extensive study addressing dynamical randomness due to different kinds of stochastic processes see Gaspard and Wang (1993). In Section 2 we give some basic arguments as to why dynamical aspects of randomness are important. Section 3 introduces a method of time series analysis capable of detecting deviations from (dynamical or non-dynamical) randomness with high sensitivity. This method is applied to data from an experiment described in Section 4.1. In Section 4.2 we sketch some details about proper error estimates for the analysis, and Section 4.3 discusses some first results from a small set of data involving human agent intention. Section 4 is a condensed version of a more comprehensive presentation in Atmanspacher et al. (1999). Section 5 concludes the paper with some perspectives. Deviations from Dynamical Randomness 3 2. Why Transition Probabilities? Let us consider a state space A with a (homogeneous or non-homogeneous) ~ partition P = {Ai}Z1 S U C that UAi = A and Ai n Aj = 0 b'i # j. If we assume that states w of a system can be represented as points in A then each state w can be assigned to a cell Ai of the partition. Given an overall probability measure p on A , the probability to find the state w in cell Ai , briefly denoted state probability, is Since the state is assumed to be somewhere in A , the state probabilities are normalized, Supposing that the states w are not stationary but evolve dynamically in A , each w has a predecessor and a successor. Let pij be the joint probability that the sequence w E Ai + w E Aj occurs in any two successive steps in the dynamics, for instance described by a stochastic process (for more details see Doob, 1953). Then is the conditional probability that, given w E A i , the successor of w is w E A,. The matrix in Equation (3) is called a transition matrix, its entries ' being the transition probabilities. Since each state has to be somewhere in the subsequent step, the transition probabilities have to be normalized in each row of the matrix, In another terminology, pi+j is a forward transition probability, designed for purposes of predicting a future state given the present. One can similarly define backward transition probabilities, designed for purposes of retrodicting a past state given the present, by pi+j = pij/pj. This would be the conditional probability that, given w E Aj, the predecessor of w is w E Ai. In general, ' ~ i + j pi-j . # After these definitions, it is easy to answer the question formulated in the title of the present section. Stochastic processes with diferent transition matrices pi+j can produce the same state distribution pi. This is to say that 4 H. Atmanspacher & H. Scheingraber information about the transition matrix can help to distinguish between different processes generating the same state distribution. If there are reasons to believe that (almost) identical state distributions are generated by different processes, then, in order to distinguish them, one has to look for differences in the transition probabilities. To discuss some straightforward examples, we consider two classes of stochastic processes, namely uniformly and doubly stochastic processes. Uniformly stochastic processes are characterized by an equidistribution of states, Doubly stochastic processes are characterized by the fact that transition probabilities in both rows and columns are normalized, Each doubly stochastic process is uniformly stochastic, but not vice versa (Feller, 1950, p. 399). For instance, a transition matrix of the form implies that pi = 1/N K , hence uniform stochasticity. Well-known examples are the fair coin ( N = 2 ) or the fair dice ( N = 6). Other examples for which the partition is usually much more refined are "fully developed chaos" in the logistic map (at r = 4 ), or any white noise process, i.e., processes of Markov order zero with no temporal correlations. Not quite so simple are doubly stochastic processes with transition matrices satisfying Equation (6) with different entries. For example, a process with also implies uniform stochasticity with pi = 112 for both states. More involved examples for this kind of doubly stochastic behavior are given by colored noise processes, i.e., higher order Markov processes with temporal corre- Deviations from Dynamical Randomness 5 lations. Doubly stochastic processes include the special case of strictly deterministic (e.g., periodic) processes. If the corresponding stochastic process is not doubly stochastic. This is the general case, and it is easy to imagine that in this case identical state distributions can be generated by processes with totally different transition matrices. Some well-studied examples are the golden mean process, the even process, or the Misiurewicz process. For more details, see Young and Crutchfield (1994). 3. Scaling Indices There are a number of standard techniques of time series analysis which have originally been developed for the analysis of linear systems. Tools such as autocorrelation functions and power spectra have been widely used in many fields of application. Particularly due to more recent interest in the non-linear dynamics of complex systems, there is a considerable amount of literature on non-linear time series (see, e.g., Priestley, 1988; Tong, 1990; Kantz and Schreiber, 1997); for applications see additionally Deco, Schittenkopf and Schiirmann (1 997), Kanz, Kurths and Meyer-Kress (1998) and Schreiber (1999). A particularly illuminating example for the power of advanced time series techniques, enabling the detection of phase correlations in non-linear systems, has been reported by Rosenblum, Pikovsky and Kurths (1996). Among general topics such as optimal prediction, noise reduction, tests for stationarity and linearity, it is one of the central problems for many applications to distinguish between random and non-random contributions in a given time series. An extremely sensitive technique in this regard is the scaling index analysis, a procedure strongly inspired by the concept of multifractals (Mandelbrot, 1974; Halsey et al., 1986). Scaling indices essentially represent a fairly sophisticated and compact measure of correlations between data points (Paladin and Vulpiani, 1987). For a given point set they describe how the local density of the set around each point changes with increasing distance from that point. A convenient way to visualize this is in terms of a histogram N(a!) showing the number N of points in the set as a function of the scaling index a . The scaling index a! (sometimes also denoted "crowding index"; Grassberger, Badii and Politi, 1988) is defined i i a . = log n ( € 2 ) - log n (€1) ' log €2 - log €1 where ni(~) the number of points within a box of size e , to be considered is 6 H. Atmanspacher & H. Scheingraber around point i of the entire point set (i = 1,2, ..., NtOt),and €1 < e2. The scaling index is thus defined for a specific range of e , which in turn defines a locality criterion with respect to which correlations are characterized by ai . Counting those boxes (to be constructed around points) that give rise to a i , a histogram N ( a ) is obtained. For more details and additional references, see Atmanspacher, Scheingraber and Wiedenmann (1989) and Atmanspacher et al., (1999). For perfectly random distributions in a d -dimensional space, the ideal N ( a ) histogram for infinitely many points is a 6 -function at a = d. Due to the restrictions imposed by a finite number of points, non-zero E , and finite binning of a , N ( a ) is broader and its mean is shifted toward values of a smaller than d . The same happens for regular patterns with topological dimension n < d , where the ideal N ( a ) for infinitely many points is a 6-function at a = n . It is intuitively clear that a scaling index analysis can be useful to discriminate between non-random features with n < d and a random background. For examples in various scientific areas see Atmanspacher, Scheingraber and Wiedenmann (1989), Morfill, Demmel and Schmidt (1994), Atmanspacher, Wiedenmann and Amann (1995), Rath and Morfill (1997), Wiedenmann, Scheingraber and Voges (1997), and Atmanspacher et al. (1999). For faint non-random contributions dominated by a random distribution such a task is difficult since -depending on their nature - they typically appear as small deviations in the left wing of an N ( a ) histogram characterizing randomness. For such situations, it is crucial to take care in selecting a good locality criterion (range of E over which ai is calculated) and in estimating errors in an appropriate manner. If a correlation analysis based on a scaling index analysis is to be carried out for a time series (temporal sequence of data) rather than a spatial pattern, then it is necessary to represent the time series in a space with dimension greater than one. A standard way to achieve such a representation has been proposed by Packard et al. (1980) in order to reconstruct attractors of dissipative dynamical systems and extract their invariants. Our goal in the present study is less ambitious since we do not look for invariants of a real physical process but simply for correlations in the temporal sequence of data points in a time series. Inspired by the delay coordinate technique of Packard et al., we use the original time series $(t) to construct a number ( d - 1) of additional time series, each delayed by a temporal interval At with respect to its predecessor. In this way, new coordinates xi are obtained according to Xl(t)= 4 ( t ) , x ( t )= 4 (t At), 2 . - . _ X d it) = 4'(t ( d - 1) At) + + Deviations from Dynamical Randomness with In this manner, the original (one-dimensional) time series can be represented in a d -dimensional space, and the corresponding histogram N ( a ) can be calculated. This allows us to discriminate regular (deterministic) contributions from random noise in the behavior of dynamical systems. In particular, nonrandom features in the transition probabilities (i.e., correlations) between individual members of the time series can be detected in addition to non-random state probabilities. Such a task is fairly straightforward in the case of low-dimensional attractors (fixed points, limit cycles); compare, e.g., Siffling (1996). For investigations of less prominent non-random temporal features in a random background process, the analysis becomes more sophisticated. In this case, one would ideally proceed to embedding dimensions d as high as possible since the a -range of N ( a ) characterizing random contributions increases with d whereas the a -range of the non-random part of N ( a ) drops back with increasing d . This means that faint non-random contributions appearing as deviations in the left wing of the dominating random part of N ( a ) become more pronounced and thus can be better discriminated as d is increased. However, severe restrictions on the size of d are imposed by the length of the original time series. For as few as 1000 data points per time series, the point distribution resulting from the delay technique in d = 4 is not dense enough to admit a statistically reasonable analysis. For 10,000 data points, d = 4 is a reasonable upper limit for d . In addition, it is mandatory to carry out a proper error estimate since the N ( a ) histogram for a random distribution of not more than 10,000 points deviates significantly from its limit for infinitely many points. Non-vanishing correlations are to be expected for random time series of finite length. The question is whether the scaling histogram of an empirically observed time series of finite length differs significantly from the expected histogram for a random time series with the same finite length. It makes absolutely no sense to compare empirically obtained N ( a ) histograms for finite time series with N ( a ) histograms as they are theoretically expected in the limit of infinitely long time series. Significant deviations from a random distribution of points as detected by a scaling index analysis of a time series embedded in a d -dimensional space can have different origins. In particular, such deviations are not necessarily due to deviations from dynamical randomness, i.e., temporal correlations in the time series. Moreover, the delay technique generating the embedding space can lead to delicate superpositions of different k i d s of correlations that are difficult to separate from each other. It is therefore inevitable to apply additional procedures if one wants to disentangle genuinely temporal correlations from others; this issue will be taken up later on. H. Atmanspacher & H. Scheingraber 4. Data Analysis 4.1 Experimental The prototype of the experimental setup from which data are taken for analysis goes back to Schmidt (1970). It has been refined and applied to a large range of empirical questions in the work of the Princeton Engineering Anomalies Research (PEAR) project within the last two decades. The general question behind that work is whether some "intentional activity" of human agents changes the output of physical systems which are expected to produce random events. The work of PEAR suggests that there are significant deviations from randomness when human agent intention is involved (Jahn et al., 1997; see also Utts, 1991). A replication of the PEAR studies has been started at the Institut fiir Grenzgebiete der Psychologie und Psychohygiene (IGPP) at Freiburg in 1996. The present investigations refer to a small subset of data obtained in the IGPP replication study. The material core of the experiment is a random event generator (REG) whose output (after some processing) consists of binary sequences. In the IGPP replication study, a portable random event generator, different from the REG originally developed and used by PEAR, was utilized. More details concerning these two sources of randomness can be found in Nelson, Bradish and Dobyns (1989) and Bradish (1993). Both sources are semiconducting devices, producing a mixture of quantum and thermal noise. After 200 bits (0s and Is) were generated by the REG, the number of 1s is counted and the result is taken as the outcome of a single "trial" for the experiment. Sequences of 100 and 1000 trials, respectively, constitute experimental "runs" consisting of successive integer "raw data" xi scattered around a mean value of 100 with a standard error of The trials of a run can be graphically represented (depending on feedback options) on a monitor. This representation tiis obtained from the raw data xi by m. or, recursively, by i The series t is thus cumulative in the sense that a constant value above (below) 100 in the raw data would produce a monotonically increasing (decreasing) graph for tion the monitor. If the REG produces random numbers, Deviations from Dynamical Randomness 9 the overall expectation (for N + oo) is that the curve tidoes not significantly deviate from the baseline 0. According to the experimental protocol, human agents are asked to intentionally try to cause the curve to rise above or fall below the baseline, or to be intentionally neutral with respect to the appearance of the curve in different runs. (They are not explicitly asked to achieve deviations from randomness.) The corresponding "modes of intention" are denoted as "high," "low," and "baseline." It is left to the agents how to realize each of these modes of intention cognitively. The particular intention per run is preselected randomly or by the agents themselves. An experimental "session" consists of 1000 trials per each intention. For runs consisting of 100 (1000) trials, a session therefore amounts to 10 runs (1 run) per intention, i.e., 30 (3) runs in total. For the analysis described below, data from agents with 10 sessions each have been used, i. e., 10,000 trials per agent and intention. To avoid artificial correlation effects, the analysis has been based on the raw data Xi rather than the cumulative data ti. Regarding the values assumed by xi (roughly between 70 and 130) as states, the histogram of those states represents a state distribution. The most prominent result obtained from analyses of the data pool collected by PEAR is that the mean value of the distribution of state probabilities (integer values xi) shows small systematic shifts above and below the expected mean of 100.0 for experimental conditions with high and low intention, respectively. The precise values as given by Table 1 in Jahn et al. (1997) are 100.026 for high, 99.984 for low, and 100.013 for baseline. Although the corresponding high-low separation of 0.042 is tiny, Jahn et al. (1997) report an overall z -score of 3.81 (with p = 7 x for the entire set of individual sessions. (For more details on the importance of z-scores and effect sizes in this and related studies see Utts (1991) and Delanoy (1996)). This is not the only result that was found by PEAR. After perusing the available literature, it appears that the analyses that have been done are all based on state probabilities and do not explicitly take transition probabilities into account. As shown in Section 2, it is interesting to focus on those transition probabilities between states since they can reveal information that is not available in the distribution of state probabilities alone. In the following, we describe a scaling index analysis of time series obtained from a physical random event generator (REG) as described above under two different conditions: 1. The output of the REG is sampled without any additional constraints on the experimental setup; the corresponding data are used for calibration and allow us to derive a reasonable error estimate; 2. The output of the REG is analyzed for situations in which human agents are asked to carry out "intentional tasks" that are defined so as to correspond to non-random contributions to the activity of the REG. The main conceptual difference between the scaling index analysis and the analyses carried out by PEAR so far consists of the fact that we do explicitly 10 H. Atmanspacher & H. Scheingraber address transition probabilities between states together with state probabilities. (Of course, there are other possibilities to realize such a purpose; see, e.g., Kurths et al., 1995.) As mentioned above, the data used for this purpose are data from the IGPP replication study. This study is not yet finished, implying that any assignment of individual runs to agent intentions is kept hidden so far. Therefore, our analysis is strictly bound to an analysis of deviations from randomness and does not refer to specific agent intentions. 4.2 Errors In experiments such as those briefly sketched in Section 4.1, one should by no means expect easily detectable, major deviations from a perfectly random distribution. For this reason, extremely careful error estimates are mandatory for a sound scaling index analysis. These error estimates have to be based on the same parameters as used in the analysis of those data taken under the influence of human agents. For time series consisting of 10,000 data points At = 1 and d = 4 have been selected. Scaling indices a! are binned in steps of 0.01. For an optimal locality criterion, c, = 4.6 and €2 = 12.7 have been determined by minimizing the differences between N ( a ) histograms for calibration data (Atmanspacher et al., 1999). This means that we look for correlations on the mentioned distance scale only. Correlations on larger scales, e.g., approaching the diameter of the point distribution as a whole, are left out of consideration. Calibration data are taken from the experimental random number device as described in Section 2. Figure 1 shows an example of a resulting N ( a )histogram. Fig. 1. Histogram of scaling indices N ( a ) for a set of 10,000 calibration data obtained from the IGPP replication of the PEAR experiment. The embedding dimension is d=4 so that the theoretical expectation for N+oo (and Aa+O ) is a 6 -function at a=4 . Deviations from Dynamical Randomness I1 The relative differences (normalized with respect to the total number Ntot of points) between any two histograms N ( C Y and N ( a ) z (for specifications see Section )~ 4.3) will be shown and discussed in integral form according to 4 2 ( a ' )= N:,t lo ( N (4, N - ( 4 2 ) Such an integral representation' has the advantage that consistent trends in the differences 6 over extended ranges of a become clearly visible even if they are small and noisy. Values aeXt at which A is maximal (minimal) correspond to the onset of negative (positive) differences 6 after an extended range a < aeXt of positive (negative) differences 6 . Hence, an extremum in A indicates that the differences 6 giving rise to it are at a < aeXt . of two As an example, the dotted line in Figure 2 shows the differences histograms. They are calculated for each a bin separately and normalized with Fig. 2. Example of differential and integral representations of the differences between two N ( a ) histograms. Dotted line: relative normalized differences 6 1 , ~ each bin. Solid for line: relative normalized integral differences A1, 2 (same as dotted line in Figure 4). 'AS always, integrals are numerically evaluated by a sum of finitely many terms, here given by the values of N ( a ) within bins of Aa=O.Ol . 12 H. Atmanspacher & H. Scheingraber respect to the total number NtOt of points. The solid line in Figure 2 shows the corresponding integral plot of A1,2, again as relative deviations. For a < aext = 3.18, the differences 61,2 are consistently negative. This trend bechanges at a,,, = 3.18, where Al,2 is minimal and the differences come positive. The relative deviation at aeXt= 3.18 is negative and amounts to 0 1 , 2 3.56%. In order to obtain reasonable error estimates for the analysis, the fluctuations of such histograms for 100 different realizations of calibration data were considered. Subtracting each of the 100 individual N ( a ) histograms from a mean N ( a ) histogram, a mean fluctuation (Acal) was determined as a function of a (see Figure 3). The applied procedure, in detail described by Atmanspacher et al. (1999), is basically heuristic; further work to back it up formally is in progress. Insofar as the calibration data can be regarded as surrogate data (Theiler et al., 1992), this mean fluctuation represents a proper error estimate for the histograms. Values of (Acai) that differ from zero quantify how much the density of an individual set of calibration data fluctuates around the density of an average distribution of calibration data with N,,, = 10,000 on purely statistical grounds (given that the overall density is constant). Deviations in the left wing of N ( a ) refer to "overdense" regions in the embedding space. They indicate that local correlations in an individual set of calibration data deviate from those of an average distribution of calibration data, i.e., are more or less homogeneous than that average distribution (with a finite number of points). "Underdense" regions such as voids (characterized by "anti - correlations) correspond to the right wing of N ( a ) where errors typically are much larger and prevent any significant detection (Atmanspacher, Scheingraber and Wiedenmann, 1989). In another publication (Atmanspacher et al., 1999), other sets of surrogate V Fig. 3. Mean values (solid line) and standard deviations (shaded area) of A(cr,,,) ma of 1 as a function of a,, . for the maxi- Deviations from Dynamical Randomness 13 data such as fitted Gaussian distributions and various kinds of Monte Carlo permuted data were studied in addition. It turned out that fluctuations as obtained from the calibration data are both most conservative and most realistic for an error estimate with respect to our present purposes. The detailed procedures and arguments can be found in Atmanspacher et al. (1999). 4.3 Some First Results The empirical material analyzed was taken from a series of experiments carried out at the Institut fiir Grenzgebiete der Psychologie und Psychohygiene (IGPP) at Freiburg. The purpose of the IGPP study is a replication of the results obtained by PEAR as briefly summarized in Section 4.1. Only those agents were selected for the analysis who had finished 10 sessions within the IGPP replication study by the end of September 1997. Due to the experimental protocol, the corresponding amount of data is 10,000 per mode of intention and per agent. Therefore, individual agents were coded by numbers (1, ..., 16), and modes of intention were randomly coded by numbers (1, 2, 3) for each agent separately before the data were released for analysis. Hence the analysis itself amounts to nothing else than an analysis of deviations from randomness. In particular, it does not make use of any information concerning correlations with psychological observables. Time series provided by sequences of 10,000 data points per intentional mode for each agent have been analyzed, regardless of any other possible discriminations. This is to say that we disregarded any information other than the distinction between different but unknown (not yet decoded) modes of intention. In other words: we assume that each sequence of 10,000 data points represents a statistical ensemble. The reason is that pilot studies showed that possible deviations from randomness have to be expected so small that a sensible scaling index analysis requires data series of at least 10,000 points to be analyzed in at least four dimensions, d = 4 . (Other parameters are as given in Section 4.2.) More detailed studies (with shorter time series) may be possible if agents can be identified for which major deviations from randomness are observed. The first step of the analysis was the determination of the relative integral differences between the N ( a ) histogram for a suitable representative for a random sequence and the three N ( a ) histograms corresponding to the three modes of intention for each agent. Selecting the mean histogram over all calibration data (Section 4.2) as such a suitable representative, was calculated for each agent and mode of intention. The solid line in Figure 2 represents as a function of a for intentional mode #2 of agent #16. There is a prominent negative peak of 3.56% at a = 3.18, where the corresponding error is 1.1 (see Figure 3). This means that the deviation from 14 H. Atmanspacher & H. Scheingraber randomness in this example is significant with 3.20 (corresponding to = 0.0017 ). It represents the most pronounced effect resulting from an analy. sis of the extrema of Acal,int The significances of differences A might be underestimated if one considers only the extrema of A . This is due to the fact that the errors in the critical range 2.9 < a < 3.3 depend strongly on a . As a consequence, deviations smaller than their extrema may be more significant than those at the extrema, particularly if they are located at small enough values of a . For this reason, it is interesting to check the dependence of significance of deviations on a in addition to the deviations themselves. For instance, the significance of deviations for agent #16/intention #2 turns out to be highest at a = 2.98 where an overwhelming 9.520 is obtained. Significances of deviations from randomness and additional features due to other agents and modes of intention are summarized in Figure 6 and Table 2 of Atmanspacher et al. (1999). It is remarkable that those data points contributing to significant deviations from randomness refer almost exclusively to states or transitions between states, respectively, located within a small stripe (= f 5 ) around the mean (x) = 100. This implies that major deviations are not caused by (transitions between) states far away from (x) . Lucadou (1986) reported a similar observation with respect to his preferred observable "pragmatic information." The relationship between this observable and dynamical randomness remains to be clarified. Due to the fact that the deviations for agent #16 are so prominent, it has been checked whether they can also be observed with 3-dimensional rather than 4-dimensional embedding. Reducing the embedding dimension to three makes it possible to analyze the 10 sessions (comprising 10,000 data points altogether) individually. The characteristic features indicating non-random contributions remain pronounced for all 10 sessions with 1000 data points each; quantitative significances have not yet been determined. It is remarkable that the characteristic features are homogeneously distributed over the sessions. This underlines the option to further analyze the empirical material concerning agent #16. Since the analysis as performed in this study is sensitive to deviations from randomness in the transition probabilities as well as state probabilities, it is desirable to distinguish between these different types of non-randomness in the experimental data. If the deviations are due to temporal correlations, they should vanish (or at least decrease) for a Monte Carlo randomization of the sequence of states. This can be tested by generating random permutations of the time series taken under agent intention and subtracting the resulting N ( a ) histogram from the mean histogram of the calibration data. It turned out that all significant deviations drop well below20 if the corresponding sequences are randomized. The hypothesis that we are dealing with deviations from randomness in the transition probabilities rather than randomness with respect to states can be p Deviations from Dynamical Randomness 15 further backed up by a simulation of small temporal correlations within an otherwise random time series. As an example, one can replace a small proportion, say 60 randomly selected sets of three successive data points ("triplets") in a random series of calibration data with N,,, = 10,000 by 60 ordered triplets of the form {91,93,95). Analyzing the modified time series according to the procedure described above leads to deviations A from the mean calibration distribution which are plotted as a solid line in Figure 4. The dotted line in the same figure reproduces (from Figure 2) the deviations obtained for agent #16 intention #2. 5. Concluding Remarks The applied analysis was basically used as a non-parametric procedure to check the hypothesis of randomness in the data. There are strong indications that this hypothesis is rejected for some agents. Therefore, our results do strongly encourage further investigations of deviations from randomness in physically random devices and their relationship to the intention of human agents with appropriate techniques. In particular, it seems worthwhile to focus on the dynamical aspects of such deviations in addition to more straightforward analyses of state probabilities. There are some immediate options for future lines of research. 1. As soon as the identification of agents and modes of intention in the IGPP study is revealed, it will be interesting to see in which way Fig. 4. Solid line: relative normalized integral differences A between the N ( a ) histograms of a distribution with 60 ordered triplets and a mean random calibration distribution. Dotted line: relative normalized integral differences A between the N ( a ) histograms of the distribution corresponding to agent #16/intention #2 and a mean random calibration distribution (same as solid line in Figure 2). 16 H. Atmanspacher & H. Scheingraber intentional modes (andlor other psychological observables) correlate with deviations from dynamical randomness. Moreover, it has to be investigated whether those agents ( e . g . , #16) generating such deviations are able to reproduce them. Experimental results from more agents than studied so far are to be analyzed. In addition to material that can be provided by IGPP and a similar replication study at the University of Giessen, there is a tremendously rich data pool at PEAR. These data will be systematically surveyed in the near future. The scaling index analysis allows us to identify those data points contributing to the observed deviations from randomness. Since they are typically located in a small stripe around the mean(x) = 100, a search for patterns in those points is considerably facilitated. It is not yet clear whether such patterns are of the type giving rise to the solid line in Figure 4. The relationship between significantly non-random features in the scaling indices and the underlying stochastic process is both highly nontrivial and non-unique. Further work is necessary to explore this issue in more detail. Finally, it should be emphasized that the results presented so far are not to be understood in terms of a "proof7for the "reality" of yet unknown relationships between physical and mental systems. Statistical analyses can indicate evidence for something but never "prove" its "reality." Furthermore, we do not think that adding more results of the same kind to the existing material would change the situation with respect to its broader acceptance. Such results will not be accepted in the corpus of serious scientific knowledge unless plausible concepts concerning the context under which they occur can be presented. For these reasons there is no particular value in an analysis of the overall evidence for deviations from dynamical randomness for the full sample of 16 agents with three modes of intention each. Our research strategy focuses on a step-by-step procedure oriented toward understanding the origin of and the boundary conditions for those deviations for which the applied analysis indicates evidence. If at all, this can be done most effectively with those agents who seem to be capable of generating them. Acknowledgments We are grateful to Werner Ehm, Jiirgen Kurths, and Gerda Wiedenmann for helpful comments. References Atmanspacher, H., Bosch, H., Boller, E., Scheingraber, H., & Nelson, R. D. (1999). Deviations from physical randomness due to human agent intention? Chaos, Solitons, & Fractals, 10, 935-952. Atmanspacher, H., Rath, C . , & Wiedenmann, G. (1997). Statistics and meta-statistics in the concept of complexity. Physica A, 234, 819-829. Deviations from Dynamical Randomness 17 Atmanspacher, H., Scheingraber, H., & Wiedenmann, G. (1989). Determination o f f ( a ) for a limited random point set. Physical Review A, 40, 3954-3963. Atmanspacher, H., Wiedenmann, G., & Amann, A. (1995). Descartes revisited: the endolexodistinction and its relevance for the study of complex systems. Complexity, 1, 3 , 15-21. Bradish, G. J. (1993). PEAR portable REG prescription for operation. Technical Note of June 15, 1993. Deco, G., Schittenkopf, C., & Schiirmann, B. (1997). Dynamical analysis of time series by statistical tests. International Journal of Bifurcations and Chaos, 7, 2629-2652. Delanoy, D. L. (1996). Experimental evidence suggestive of anomalous consciousness interactions. Biomedical and Lije Physics, ed. by N. Ghista. Braunschweig: Vieweg, 397-410. Doob, J. L.(1953). Stochasticprocesses. New York: Wiley. Feller, W. (1950). An introduction to probability pheory and its applications, Vol. 1. New York: Wiley. Gaspard, P., & Wang, X. J. (1993). Noise, chaos, and ( E , T ) -entropy per unit time. Physical Review A, 235,291-343. Grassberger, P., Badii, R., & Politi, A. (1988). Scaling laws for invariant measures on hyperbolic and nonhyperbolic attractors. Journal of Statistical Physics, 51, 135-178. Halsey, T. C., Jensen, M. H., Kadanoff, L. P., Procaccia, I., & Shraiman, B. I. (1986). Fractal measures and their singularities: the characterization of strange sets. Physical Review A, 33, 11411151. Jahn, R. G., Dunne, B. J., Nelson, R. D., Dobyns, Y. H., & Bradish, G. J. (1997). Correlations of random binary sequences with pre-stated operator intention: a review of a 12-year program. Journal of Scientific Exploration, 11, 345-367. Kantz, H., Kurths, J., & Mayer-Kress, G., eds. (1998). Nonlinear analysis ofphysiological data. Berlin: Springer. Kantz, H., & Schreiber, T. (1997). Nonlinear time series analysis. Cambridge: Cambridge University Press. Kurths, J., Voss, A., Saparin, P., Witt, A., Kleiner, H. J., & Wessel, N. (1995). Quantitative analysis of heart rate variability. CHAOS, 5, 88-94. von Lucadou, W. (1986). Experimentelle Untersuchungen zur BeeinJluJbarkeit von stochastischen quantenphysikalischen Systemen durch den Beobachter. Frankfurt: Herchen, p. 225. Mandelbrot, B.B. (1974). Intermittent turbulence in self-similar cascades: divergence of high moments and dimension of the carrier. Journal of Fluid Mechanics, 62, 331-358. Morfill, G. E., Demmel, V., & Schmidt, G. (1994). Der plotzliche Herztod. Neue Erkenntnisse durch die Anwendung komplexer Diagnoseverfahren. Bioscope, 2/94, 1 1-19. Nelson, R. D., Bradish, G. J., & Dobyns, Y. H. (1989). Random event generator qualification, calibration, and analysis. Technical Note PEAR 89001, April 1989. Packard, N. H., Crutchfield, J. P., Farmer, J. D., & Shaw, R. S. (1980). Geometry from a time series. Physical Review Letters, 45, 712-716. Paladin, G., & Vulpiani, A. (1987). Anomalous scaling laws in multifractal objects. Physics Reports, 156, 147-225. Priestley, M. B. (1988). Non-linear and non-stationary time series analysis. London: Academic Press. Rath, C., & Morfill, G. (1997). Texture detection and texture discrimination with anisotropic scaling indices. Journal of the Optical Society ofAmerica A, 14, 3208-3215. Rosenblum, M. G., Pikovsky, A. S., & Kurths, J. (1996). Phase synchronization of chaotic oscillators. Physical Review Letters, 76, 1804-1807. Schmidt, H. (1970). A PK test with electronic equipment. Journal of Parapsychology, 34, 175181. Schreiber, T. (1999). Interdisciplinary application of nonlinear time series methods. Physics Reports, 308, 1-64. Siffling, R. (1996). Phasenraumbetrachtungen bei synchronen Zeitreihen. Diploma thesis, Universitat Miinchen. Theiler, J., Eubank, S., Longtin, A., Galdrikian, B., & Farmer, J. D. (1992). Testing for nonlinearity in time series: the method of surrogate data. Physica D, 58, 77-94. Tong, H. (1990). Non-linear time series analysis. Oxford: Oxford University Press. Utts, J. (1991). Replication and meta-analysis in parapsychology. Statistical Science, 6, 363-403. Wackerbauer, R., Witt, A., Atmanspacher, H., Kurths, J., & Scheingraber, H. (1994). A comparative classification of complexity measures. Chaos, Solitons, & Fractals, 4, 133-173. 18 H. Atmanspacher & H. Scheingraber Wiedenmann, G . , Scheingraber, H., & Voges, W. (1997). Source detection with the scaling index method. In Data Analysis in Astronomy 5, eds. di Gesu, V. et al. Singapore: World Scientific, 203-21 1. Young, K., & Crutchfield, J. P. (1994). Fluctuation spectroscopy. Chaos, Solitons, & Fractals, 4, 5-39. Journal of Scientific Exploration, Vol. 14, No. 1, pp. 19-33,2000 0892-3310100 O 2000 Society for Scientific Exploration Valentich Disappearance: New Evidence and a New Conclusion 325 Langton Avenue Los Altos. CA 94022 Victorian UFO Research Society I? 0. Box 43 Moorabbin, 3189 Australia Abstract-This paper presents new evidence regarding the now-famous disappearance of Frederick Valentich, who was flying a Cessna airplane on the evening of October 21, 1978, somewhere near Cape Otway SW of Melbourne. The testimony of three witnesses is given, each of whom claim they saw an airplane descending downward at a steep angle with a much larger object with green lights flying just above it. A plot of the most probable flight path is also included. Based on this new evidence, taken in conjunction with the pilot's own in-flight reporting of sighting events, we have to conclude that there appears to be sufficient evidence to suggest that Valentich's airplane probably crashed into the sea SE of Cape Marengo between 3 and 12 miles offshore. The nature of the large object with green lights that accompanied the airplane during its steep descent remains to be identified. Keywords: pilot disappearance - accident analysis - UFO - crash investigation Introduction The in-flight disappearance of Frederick Valentich over Bass Strait, Australia, on October 21, 1978, has become one of the most well-publicized mysteries of aviation since Amelia Earhart disappeared on July 3, 1937. Accounts of this tragic event may be found elsewhere (International UFO Reporter, 1978; Bass Strait mystery, 1979; Haines, 1987; Norman, 1979; Pinkney and Ryzman, 1980; Valentich, 1980). Despite the coordinated efforts of private pilots and the Australian government's search-and-rescue airplanes immediately following the event, no trace of Cessna DSJ (its registration letters: "Delta Sierra Juliet") of any kind was ever found. What has made this event such a perennial and popular mystery was the existence of an air-to-ground radio (voice) transmission between young Valentich and a flight service specialist, Steve Robey, who was working at Melbourne International's "Tullamarine" airport at the time of the disappearance. Other pilots overheard this transmission and, because of intense and immediate pressure on the civil aviation authorities, the Department of Transport 20 R. F. Haines & P. Norman (DOT)released a printed transcript of the conversation long before the official accident report was issued. The authors also have listened carefully to this tape recording; the eyewitness description and other sounds therein should be of interest to those who are truly interested in UFO phenomena (Haines, 1981). A detailed account of the entire event is found elsewhere (Haines, 1987). There is nothing in this 13-minute audio tape that contradicts the new evidence presented below. Other than a short article published in Australia (Norman, 1991), there has been no new evidence that relates directly to this reported aerial encounter and subsequent aircraft disappearance Aircraft pacing and other forms of reported interference with airplanes by unusual and nonaerodynamically shaped objects is not exceptional. One of the authors (R.H.) has compiled the following list of such events for the general period 1948-1989: 55 cases involving airplane pacing, 15 cases in which the aerial object completely circled the airplane one or more times, 12 cases in which the object suddenly disappeared from the pilot(s) sight, 22 cases involving a head-on approach to the airplane and near-miss by objects that did not appear to be airplanes, and scores of incidents in which on-board electromagnetic hardware was affected only when the UFO was nearby (Haines, 1992; Sturrock et al., 1998). Selected Background Information Pilot Frederick Valentich, 20, made arrangements with Southern Air Service, located at the Moorabbin airfield SSW of Melbourne city center, to rent a Cessna 182L model, single-engine, propeller-driven airplane for his night flight. He submitted his flight plan to the briefing officer at the airfield at 5:20 p.m. and finally took off alone at 6:19 p.m. for what was to be a "full-reporting" flight. This means that he was supposed to check in by radio with flight service personnel at certain defined checkpoints for safety reasons. His destination was King Island, about halfway between the Australian mainland and the tip of Tasmania (see lower left inset in Figure 1). Flying at 120 miles per hour (neglecting wind effects), the journey from Cape Otway to the nearest point of land on King Island would be about 48 miles (24 minutes of flight) flying at 4,500 feet altitude. The sun would set at 6:48 p.m.; but it was almost 7:00 p.m. when Valentich finally reached his designated (radio) reporting point near Cape Otway. This conclusion is based on a complete flight path reconstruction, including prevailing wind conditions. His radio call at 9:00:29 stated, "Melbourne, Delta Sierra Juliet. (Now at) Cape Otway, descending for King Island." He was right on time. (Note the 2-hour time difference between local and GMT used in the official transcript. We will use GMT for the remainder of this paper). According to his flight plan, Valentich planned to climb to at least 4,500 feet altitude for his water crossing (for safety and visibility reasons). We assume Valentich Disappearance 21 es observed his blue and white Cessna from the resort town of Apollo Bay as it flew SW over the water at an unspecified distance. Several local pilots have pointed out that it is normal procedure to "cut the corner" at the cape when flying to King Island (i.e., not to fly all the way to Point Franklin, Crayfish Bay, or the lighthouse itself before turning left for King Island; Figure 1). Valentich had flown this same route in the past, and presumably, he cut the comer on this flight as well. Doing so would shorten his trip by about 6 miles, saving both time and fuel. Indeed, Norman (1991) interviewed fishermen who had camped along the Parker River (south of Point Lewis; Figure 1) that night. They apparently saw the Cessna make this turn Path (Dashed Line) u Otway ' 0 1 2 16Oo Assumed Altitude 3,000 feet \ \ \ 9 : 0 3 ~ ~ m w times with re voice tnnsc.pi t+etai. Last Radio Encine begins to lose power ---1' \-9:09:00 \ \ Probable h a t i o n of Initial Visual Contact Fig. 1. Enlarged scale chart of region from Apollo Bay to Cape Otway. 22 R. F. Haines & P. Norman about 3 to 4 miles ENE of the Cape Otway lighthouse. The eastern sky was now dark, although the western sky still possessed some orange glow from sunset. The scattered ground lights visible off Valentich's right side likely helped him maintain his general flight path direction up to this point. After changing his heading to the left, he probably continued on out over Bass Straight toward the nondirectional beacon (NDB; inset, Figure 1) on King Island (a magnetic heading of 154.5"). Flying at 4,500 feet altitude and between 110 and 120 miles per hour (there was a tail wind of about ten knots out of the NW), he reported by radio to Steve Robey, who was handling this particular air sector that evening, that he saw "a large aircraft below 5,000" (feet altitude). The time was exactly 9:06:14 according to the official transcript of this interchange. Table 1 presents the DOT voice transcript as it may possibly relate to the new evidence presented here. The key for Table 1 defines the speech timing and inflection symbols that were added based on a detailed analysis of the original voice tape by R.H. In the years following this event, one of the authors (P.N.) succeeded in locating and interviewing a number of people traveling or living in the region along Great Ocean Road, which runs north and south through Apollo Bay. Reports were obtained from 20 eyewitnesses in this region, describing an erratically moving green light in the sky at that same time of evening as Valentich's flight. In addition, P.N. learned of three primary eyewitnesses who shed valuable new light on this event. Their testimony is recounted here. They saw both the lights of a small aircraft and a very large green light traveling directly above it. The primary witness, Mr. Ken Hansen (pseudonym), who was 47 years old at the time, told his wife of what he and his two nieces had just seen on their way home, but she laughed at his story. The following morning at work he told his fellow employees, who believed what he said about seeing the airplane, but not about the large green object flying above it, the details of which are given below. Of course, at this early date, he could not have known anything about Valentich's description of a green light flying near him. Hansen decided to drop the subject to avoid further ridicule. Years later, he happened to discuss his sighting with a local policeman, who later mentioned the story to Guido Valentich, father of the missing pilot. Guido Valentich then told author P.N., who interviewed Hansen and his two nieces. Both girls gave the same basic details as their uncle. Site Visit to Apollo Bay During a visit to the area between Cape Otway and the resort town of Apollo Bay on March 17, 1998, both authors had an opportunity to meet Mr. Ken Hansen (pseudonym), who was then age 67. Hansen lives in the resort town of Apollo Bay. As he had told author P.N. in 1991, he said that he had seen, with his two nieces, an odd aerial event the same night that Valentich had disappeared. We asked if he would take us to his original observation site so that we Valentich Disappearance TABLE 1 Official Voice Transcript Between Flight Service (FS) and the Cessna Aircraft (DSJ) 23 Time (GMT) From DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ DSJ To FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS FS Text Melbourne, this is Delta Sierra Juliet. Is there any known traffic below five thousand?' Delta Sierra Juliet--No known traffic. Delta Sierra Juliet. I am-seems (to) be a large aircraft below 5,000. D Delta Sierra Juliet-What type of aircraft is it? Delta Sierra Juliet-I cannot affirm. It is four bright . .. it seems to me like landing lights. Delta Sierra Juliet. [This statement affirms to the pilot that the person on the ground heard his transmission.] Melbourne, this (is) Delta Sierra Juliet. The aircraft has just passed over me at least a thousand feet above. Delta Sierra Juliet-Roger-and it, it is a large aircraft-confirm? Er, unknown due to the speed it's travelling.. . is there any airforce aircraft in the vicinity?Delta Sierra Juliet. No known aircraft in the vicinity. Melbourne.. . it's approaching now from due east- towards me.Delta Sierra Juliet. IlOpen microphone for two secondsll Delta Sierra Juliet. It seems to me that he's playing some sort of game.'-He's flying over me two-three times at a time at speeds I could not identify.' Delta Sierra Juliet-Roger. What is your actual level? My level is four and a half thousand, four five zero zero.Delta Sierra Juliet.. . And confirm-you cannot identify the aircraft. Affirmative.' Delta Sierra Juliet-Roger... standby. Melbourne-Delta Sierra Juliet. It's not an aircraft'. . . it is llopen microphone for two secondsll [This duration measured as three seconds. No information appears to have been removed from the tape.] Delta Sierra Juliet-Melbourne. Can you describe the.. .er-aircraft? FS DSJ FS DSJ FS DSJ DSJ FS DSJ FS DSJ FS FS DSJ R. F. Haines & P. Norman TABLE 1 Continued DSJ Delta Sierra Juliet... as it's flying past it's a long shape' //open microphone for three seconds// (cannot) identify more than that. It has such speed //open microphone for three seconds//.It is before me right now Melbourne.' DSJ FS Delta Sierra Juliet-Roger. And how large would the -er-object be? Delta Sierra Juliet-Melbourne. It seems like it's (stationary). [Author R.H. has determined that this word should be "chasing me" based on special filtering]. What I'm doing right now is orbiting, and the thing is just orbiting on top of me also' .. . It's got a green light,' and sort of metallic (like)-. It's all shiny (on) the outside.Delta Sierra Juliet. Delta Sierra Juliet // open microphone for 5 seconds // [measured as 3 seconds] It's just vanished.' Delta Sierra Juliet. Melbourne would you know what kind of aircraft I've got?' It is (a type) military aircraft?' Delta Sierra Juliet. Confirm the.. . er-aircraft just vanished. Say again. Delta Sierra Juliet. Is the aircraft still with you?' Delta Sierra Juliet... It's ah.. . Nor //open microphone for two seconds// (now) approaching from the southwest. Delta Sierra Juliet Delta Sierra Juliet - The engine is, is rough idling. -I've got it set at twenty three-twenty four... and the thing is-coughing. Delta Sierra Juliet-Roger. What are your intentions? My intentions are-ah... to go to King Island-Ah, Melbourne, that strange aircraft is hovering on top of me again Nopen microphone for two seconds// it is hovering and it's not an aircraft. Delta Sierra Juliet. Delta Sierra Juliet-Melbourne //open microphone for 17 seconds// [A very strange pulsed noise is also audible during this transmission.] Delta Sierra Juliet, Melbourne. FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS DSJ FS FS DSJ DSJ FS DSJ End of official DOT transcript Note: - = a normal pause in communications (based on the first author's flying experience); ... = a longer than normal pause (i.e.,several seconds); ' = an upward ending voice inflection (such as an interrogative question); = a downward voice inflection. Parentheses ( ) enclose words that are open to interpretation because they are not clearly audible. Brackets [ ] enclose the authors' comments. - Valentich Disappearance 25 might reconstruct each step of his sighting. He gladly agreed to do so, during which time he gave us the following information. Sighting details obtained from Mr. Hansen. Mr. Hansen and his two nieces had been shooting rabbits on the late afternoon of October 21, 1978, in the hills about 2 km west of Apollo Bay in the direction of Marriners Falls. He said that it was dusk, but he could not recall the exact time. They were in his fourwheel-drive vehicle driving east on Barham Valley Road toward his home on the southern outskirts of the town. Figure 2 shows an enlarged scale drawing of the road on which they were travelling when they sighted the lights in the sky. Hansen was driving (in the left front seat), and one niece, Tracy, was sitting in the right front seat. His other niece was in the back right seat. Tracy first sighted colored lights in the sky on their right side. The automobile was travelling about 30 miles per hour at the time in the left lane. Suddenly, she said, "What is that light in the sky?" Point A of Figure 2 shows their location at this time. As the automobile continued, Hansen craned his neck to look out the right side window in the direction that she was pointing. He caught sight of some lights and said to her, "Those are only the lights of an airplane." "No," she replied, "I mean that other large green light above it!" He drove on and then turned to look again some 10 to 15 seconds later. At that point, he also was able to make out two separate sets of lights in the clear but darkening sky. They were now near point B in Figure 2. They continued down the road, although Mr. Hansen was now slowing down because of the left turn ahead and because he wanted to better see the strange set of aerial lights. Mr. and Mrs. Hansen live near a small airstrip located just south of Apollo Bay, and he is knowledgeable about aircraft and the appearance of their lights at night. He noted clearly the familiar lights of a small airplane (white navigation light; red wingtip light) that were visible. He told us that these colored lights on the aircraft Whess stood here (center of 1 200 deg (Right-hand photo) 2 photo) 160 deg (Center photo) Fig. 2. Sighting area of Mr. Hansen and his nieces, SW of Apollo Bay, Australia (north is up). 26 R. F. Haines & P. Norman were separated by about the sanie angle that is subtended by a marble (0.65 inches) held at arm's length (approximately 22 inches from the eye) or about a 1.7" arc. Both aerial objects had passed through a 30" arc toward the east during this initial sighting interval, which lasted about 28 seconds. Not wanting to stop on the small bridge crossing Barham River, he drove on at about 20 miles per hour and finally decelerated to zero at point C of Figure 2. The car's measured transit time from point B to point C was no more than 45 seconds. Although it is not uncommon to see the lights of sinall airplanes in the vicinity, the presence of the large green light was so unusual that Hansen decided to pull over, stop, and get out of his automobile. He said that when he did so, he clearly saw a second, large, greenish, circular light "like it was riding on top of the airplane." Its angular diameter was equivalent to that of a tennis ball held at arm's length (approximately a 6.8" arc), for an angular ratio for the two objects of about one to four. Its color was similar to the navigation lights on an airplane. He also said that it kept a constant distance above and slightly behind the airplane's lights at all times. He stood watching t'or another 15 to 20 seconds until both lights disappeared from sight. Thus, the entire sighting from point A to point C-3 lasted about 93 seconds. Figure 3 is a photo-collage taken by author R.H. at the three locations along Barham Valley Road referred to in Figure 2. Mr. Hansen said that it was so dark, he was barely able to make out the tops of the trees and hill against the southeast sky. Both the airplane and accompanying light (which appeared to fly in parallel with the Cessna) seemed to descend at an apparent 30 to 40" angle (above the horizontal) along a straight line approximately as shown by the dashed line in Figure 3. Both lights eventually disappeared behind the hilltop at a magnetic bearing of about 126" froin Hansen's location (left section of Figure 3). No sounds were heard coming from the direction of the lights at any Fig. 3. Three contiguous photographs of aighting area (from 126 to 200" rnagnetic bearing). Valentich Disappearance 27 time during this sighting. The witnesses never saw the airplane strike the ground or the sea. Estimating Airplane Position Although there are too many unknowns to calculate a definitive flight path, we felt that some attempt should be made to estimate the position of the Cessna if it had continued downward on a relatively linear path, as described by the three witnesses. One difficulty in this regard arises from the possibility that the airplane and the accompanying light may not have been flying in a plane of travel normal to the line of sight but obliquely toward or away from the witness's location to some degree. If this was the case, then even a level flight path could appear to descend toward the distant horizon when viewed from the ground. This well-known optical illusion would make it appear as if the airplane was descending when it was not. Of course, there is no way to test this possibility in regard to Valentich's flight. If the plane of travel was not normal to the witness's line of sight and the airplane was descending, then the location of the "splash point" could extend over a wide range of angles and distances or may not have occurred at all. All that is known for certain is that the airplane and accompanying green light traveled somewhere within the arc defined by the two lines C-200 and C-126 in Figure 1. Other difficulties have to do with the accuracy of the perception of temporal duration itself and memory accuracy long after an event. As Hawkinds and Meyer (1965) found, most people tend to underestimate duration when they personally attend to a task and overestimate duration if they were not personally involved in a task. Although individual differences make it difficult to apply these findings to this case specifically, it is likely that Mr. Hansen underestimated the total duration of his sighting to some degree. In spite of the above difficulties, we attempted, in each of the following sections, to estimate what might have taken place using both the eyewitnesses' testimony and Valentich's in-flight DOT narrative. Our objective was to try to establish the most likely "splash down" point of the airplane in Bass Strait, assuming that the airplane continued to descend along a linear path after it disappeared from the view of the three witnesses. Probable flight path using the witness's observing time estimates. If the small airplane Mr. Hansen and the girls saw was traveling at 100 miles per hour and was seen for a total of 93 seconds (between points A and C-3 in Figure 2), it would have traveled a distance of 2.58 miles. Indeed the splash point would be only 1.2 miles off the shoreline. This linear distance is plotted on Figure 1 as line A-N near Cape Marengo. If the Cessna had been flying more slowly, say 80 miles per hour, it would have traveled a total distance of only 2.1 miles, placing it even nearer to the shoreline. Although there were many tourists in town at the time and the spring weather was relatively warm, clear, and calm, only the three witnesses reported seeing anything at that time of the I 28 R. F. Haines & P. Norman evening. For these reasons, the above position estimates appear to be too close to the shore. Estimating the distance between the eyewitnesses and the airplane based on the airplane's subtended visual angle(from memory). We then attempted to estimate how far Hansen actually was from the airplane using his memory estimate of the angular size of its lights. The linear distance between the red wingtip light and the white tail light on a Cessna 182L airplane, viewed from the side, is about 12 feet, and this was said to be equivalent to the angle subtended by a marble held at arm's length. Therefore, the calculated distance to the airplane would be only 404 feet. This distance is clearly in error for several reasons: (a) a large scale topographic chart of the area shows that the distance to the hills seen in the left part of Figure 3, behind which the airplane allegedly was seen to disappear, is about 3,000 feet away; and (b) engine sounds from a small airplane would have been heard at such close range, regardless of wind velocity and direction, yet no sound was heard. It is more likely that his use of a "marble" as a reference object is too large. Valentich's engine was running at this time and is heard on the voice tape. Assuming this angular estimate is 50% too large, we are left with a subtended angle of a 5 1 min arc between the two airplane lights and a calculated separation distance from point C (Figure 2) to the airplane of just over 800 feet, which is still too small a value for the same reasons discussed above. Indeed, if the airplane had disappeared just behind the indicated hill and had not leveled out, it would have impacted the ground and not the ocean. Its wreckage would have been found immediately. Thus, Mr. Hansen's recollection of the angular size of the airplane's lights is too large by perhaps several orders of magnitude. What might the maximum distance be between the witness and the airplane? Another estimate can be made by knowing the distance acuity for unaided vision by someone who does not wear prescriptive lenses. This is reasonable because Mr. Hansen did not need to wear corrective eyeglasses at the time. Estimated distance between the eyewitnesses and the airplane based on normal visual acuity. The human visual system can correctly discriminate two point lights at night as being separate at very small angles (less than 0.3 min arc or less; Haines, 1980). If we use this lower visual acuity threshold for the above calculations (i.e., the angular separation between the two colored lights on the airplane that can be correctly perceived as separate), we find a practical maximum separation distance between the witness at point C (Figure 1 and 2) and the airplane of 137,457 feet (26 miles). This maximum distance estimate is far too large, considering the relatively short amount of time the lights were in view and the impossibly high velocity the Cessna would have had to fly to cover this total visual angle. Estimating airplane altitude and range at disappearance. The Cessna disappeared behind low hills to the SE of the witnesses, which were about 180 feet above sea level. Site measurements indicated that these hills ranged from 5-10" arc above the local horizontal. If the Cessna were 26 miles away, it would have been at an altitude of either 12,027 feet or 24,233 feet altitude, re- Valentich Disappearance I I 29 spectively. There is little reason to accept either value as correct for several reasons. First, other witnesses saw the small plane pass overhead earlier toward the south at an altitude of no more than 5,000 feet. Second, Valentich himself indicated that his altitude at 9:09:06 p.m. was 4,500 feet (less than 4 minutes before his final disappearance); indeed, this airplane could not have climbed fast enough to reach such altitudes in the available amount of time. Finally, the eyewitnesses' total viewing duration of about 93 seconds was much too short to account for an airplane flying at this large a distance and altitude. In short, the theoretical, maximum distance to the airplane of 26 miles is, again, far too large. Estimating distance to airplane by its assumed altitude. If the Cessna was at an altitude of 4,500 feet when it disappeared below the line of hills south of Apollo Bay (along line C-3 in Figure 2) and these hills were about 180 feet high (determined from topography chart), then the horizontal distance to the airplane would have had to be about 14 miles. This point would lie along an extension of line C- 126" (Figure 1). Interestingly, Mr. Hansen s estimate of the distance to the airplane was from 10 to 12 miles. We can assume that a lower initial aircraft altitude reduces this distance. Because the eyewitnesses saw the airplane descending at a fairly rapid rate, let us assume it was at an altitude of only 1,000 feet when it disappeared behind these same hills; this yields a horizontal range of only about 2.5 miles from the witnesses (point N on Figure 1). Point N is only about 1 mile farther from the shoreline than the extension of point C3 (discussed above). Viewing duration, airplane velocity, and distance traveled. The following time and distance estimates are based on the eyewitness testimonies and lie between viewing lines B-1 and C-3 of Figure 2. Assuming that the airplane and the strange light were (a) flying in a plane normal to the line of sight, (b) flying at 100 miles per hour (8,800 ftlmin.), and (c) viewed for 65 seconds, they would have traveled a distance of 9,504 feet or 1.8 miles. Next, assuming that the airplane was at an altitude of only 2,000 feet when first seen by Mr. Hansen (from point B, Figure 2) and descended at a constant 30' angle, it would have descended the 4,000-feet glide path distance to the ground in only 27.3 seconds. The airplane was viewed for a significantly longer period than thismore than two times longer-before it disappeared. One or more of the following factors may explain this anomaly: The assumed descent angle is too steep; the velocity of the airplane was less than 100 miles per hour; its altitude, distance, or both is in error; the airplane was not flying in a plane that is normal to the line of sight; or the airplane leveled out after it disappeared behind the foreground hills. For example, if we repeat this calculation using a smaller descent angle of 20°, a speed of 80 mph, and an initial altitude of 2,000 feet, the airplane would travel the 5,847-foot-long flight path to the ground in about 50 seconds, which is more nearly equivalent to what was described by the eyewitnesses. In summary, the authors are more inclined to accept the nearer distance estimates (i.e., 3 miles to sea) than the farther distance estimates (26 miles) of 7 30 R. F. Haines & P. Norman the airplane's final disappearance point into the sea because they are more in line with the eyewitnesses' temporal estimates than with the angular estimates. The minimum-controlled-flight (stall) speed for this model Cessna is 48 knots with no flaps, zero bank angle, and center of gravity in its most forward position. Traveling at this velocity for the total viewing duration of 93 seconds at a descent angle of 30" along line B-X (Figure 1) yields a distance traveled of only 1.43 miles (to the point of aircraft disappearance behind the distant hills). This distance is clearly too small (Figure 1). The descent angle that yields a flight path length that is most compatible with all of the above facts (2,000 feet initial altitude) is between 5 and 10". A descent angle of 10" yields a distance of 11,517 feet to the surface of the sea. Traveling at 48 knots, the Cessna would require 142 seconds to travel this distance. A descent angle of 5" yields a distance of 22,946 feet and a flight time of 4 minutes, 43 seconds, to sea impact. Yet another estimate of airplane flight path can be derived from knowledge of its engine-off glide path ratio, which is between 7:1 and 8: 1, yielding a descent angle of 8" and 7.4" below the horizontal, respectively. Here, the airplane would glide (at 7:1) the 2.73 miles to the earth's surface from a starting altitude of 2,000 feet in 2.4 minutes, assuming a speed of 60 knots. The corresponding glide distance (at 8: 1) is 2.94 miles in 2.6 minutes; both duration values are reasonable. Nonetheless, engine sounds can be heard in the background of the audio tape throughout this period, which suggests a higher forward velocity than 60 knots. The Valentich audio tape transcript. At 9:09:52, Valentich stated that the unknown aerial object near his airplane appeared to be "a long shape." At 9: 10:20, he said, ". .. its got a green light and sort of metallic like, it's all shiny on the outside." Then, almost two minutes later, at 9:12:09, he said, "My intentions are-ah.. . to go to King Island-ah, Melbourne. That strange aircraft is hovering on top of me again llopen microphone for two seconds11 it is hovering-and it's not an aircraft." These were the last words ever heard from the young pilot according to the audio tape. Note that these descriptions by Valentich correspond in color and general size with the testimony of the primary eyewitness on the ground near Apollo Bay. Significantly, the signal strength and audio quality of Valentich's radio transmission did not change at any time during the entire tape, indicating that his altitude was above at least 3,000 feet. Line of sight transmission is blocked to Melbourne below this approximate altitude at this distance. Estimating UFO size. It is reasonable to assume, on the basis of psychophysical research data, that Mr. Hansen's angular estimates were basically accurate in comparing the size of the two aerial objects because they were seen side by side at the same time. Psychophysical research supports this view. Thus, the UFO's apparent angular diameter was about 4 times larger than the distance between the airplane's two external lights. Using this ratio and the known dimension separating the two airplane lights (12 feet), we find that the Valentich Disappearance 31 UFO would be about 48 feet across, assuming it was at the same distance as the airplane. A hypothetical aircraftflight path. The dashed line in Figure 1 presents one possible flight path for Cessna DSJ, which is consistent with the voice transcript. (Unfortunately, the transcript does not contain any references to particular spatial locations after 9:00:29, and even this location is not known exactly.) Tick marks are approximately 1 minute apart (assuming an airspeed of 1 10 miles per hour) and taking the wind differential into account. The main objective of this flight path reconstruction is (a) to bring the aircraft's position into correspondence with where Mr. Hansen and the girls said they saw it located and (b) to identify a general "splash down" area in Bass Strait from which search operations should commence. This flight path reconstruction is based on the possibility that at about 9:06:30, Valentich either became disoriented and frightened and banked back toward the mainland for reascns of safety, or the presence of the unidentified aerial object somehow affected his compass so that he thought he was continuing on to his original destination. A number of other magnetic compass interference cases in aircraft have been reported (Haines, 1992; Sturrock et al., 1998). We assume the following: 1. The Cessna's altitude began to descend at about 9: 10:30, shortly after Valentich began to fly in circles ("orbit"). His engine began to malfunction at 9: 11:52 (engine trouble is audible on the audio tape). 2. The airplane continued to descend at approximately 500 feet per minute so that the airplane was at 2,000 feet altitude upon reaching point B (Figure 1). 3. Radio transmission ceased after 9: 12:45 because of progressive line-ofsight signal loss caused by the earth's curvature. Of course, there is no way to determine the accuracy of this hypothetical flight path. A small dark oval (UFO) object with a dashed trail is also drawn in at various locations on Figure 1 in accordance with Valentich's description. It is obvious that his aircraft was the focus of attention of this strange object. Summary Based on what is already known about his flight plan and what can be learned from the new eyewitness evidence, we have come to the following conclusion regarding the fate of Frederick Valentich. We conclude, on the basis of the evidence presented above, that Frederick likely crashed into Bass Straight. The ground witness testimony places the airplane's approximate flight path somewhere within the arc defined by the lines C-200" and C- 126" in Figure 1, ESE of Cape Marengo. The most likely range of distances from the witnesses is from 3 to 12 miles as is discussed above. Consider the following line of evidence. First, no wreckage of Valentich's airplane has been found. If he had crashed 32 R. F. Haines & P. Norman on land, search-and-rescue personnel would have found crash debris in the 20 years following the disappearance. Locating a crash at sea is a far more difficult task. Second, Valentich's airplane was seen flying in a southerly direction east of Apollo Bay around 9:00 p.m. (based on certain reasonable assumptions). Third, Valentich was clearly disoriented by 9: 10 p.m. at the latest and probably earlier. Many pilots will not admit this to the authorities for fear of pending medical investigations that might be required, which would put their flying career in jeopardy. Fourth, the Cessna could have flown a distance of 27.5 miles at 110 miles per hour during these 15 minutes. Of course, the main question is, in which direction was Valentich flying? Fifth, Valentich was flying in circles by 9: 10:20 (and possibly earlier) and admitted to being confused about the relative magnetic bearings to the UFO by 9: 11:23. Clearly, he did not know where he was at that point. It is possible that he could have flown back toward the mainland at some time after 9:06, either deliberately or by mistake. Perhaps he was somehow captivated by the strange object he saw flying near his airplane. Sixth, at 9: 12:09, Valentich reported that the strange aircraft was ".. .hovering on top of (him) again ... and (was) not an aircraft." Is this what Mr. Hansen and his nieces witnessed from the ground minutes later? If so, then the Cessna was descending within the area shown in Figure 1 and may have impacted the water somewhere within the dashed area shown in Figure 1. Underwater search activities should begin in this region of Bass Strait. According to an Australian Marine Research report (Ocean Currents, 1997), Bass Strait is a shallow continental shelf with an average depth of from 50 to 70 m. Tide and wind action results in the mixing of the Bass Strait and the Tasman Sea, causing the saltier, colder (1-3°C) surface waters to sink (downwelling) and fall, much like a waterfall down the continental shelf slope, "beginning midway between Flinders Island and the Victorian coast and extending north almost to Jervis Bay" (Ocean Currents, p. 5). The Bass Straight Cascade pours toward the east. Tides in Bass Strait "originate from the tidal wave traveling southward down the east coast of Australian. As the wave passes the eastern entrance of Bass Strait, some of its water is deflected into it, slowing down to 80 km per hour in the shallower water. The rest of the wave continues at high speed around Tasmania in a clockwise direction to reach the western entrance to Bass Strait some 3 hours later. The wave front entering from the west meets the wave front entering from the east, causing large tides along a north-south line in the middle of Bass Strait." Because of the velocity and force of these currents, it is likely that underwater debris may be carried a long distance. The relatively low mass aluminum structure of Valentich's Cessna airplane would not sink quickly, nor would it dig into the bottom surface very far as would an anchor or the hull of a heavy ship. It might be possible to locate a particular area where such debris would accumulate over time. Computer simulations should be run to develop estimates of the debris field on the sea bottom, given tides and currents in the Valentich Disappearance 33 We may never know exactly what happened to Frederick Valentich. Nevertheless, an attempt should be made to locate the airplane. An underwater search should be mounted, despite the 20 years that have elapsed since the event took place. References The Bass Strait mystery. (1979, February). Australian UFO Bulletin, pp. 7-11. Foreign forum. lntemational UFO Reporter (1978, December). 3, 2-10. Haines, R. F. (1980). Observing UFO. Chicago: Nelson-Hall. Haines, R. F. (1981). Results of sound spectrum analysis of a tape recorded radio transmission between Cessna VH:DSJ and Melbourne Flight Service: I. The metallic noises. The Journal of UFO Studies, 3, 14-23. Haines, R. F. (1987). Melbourne episode: Case study of a missing pilot. Los Altos, CA: LDA Press. Haines, R. F. (1992). Fifty-six aircraft pilot sightings involving electromagnetic effects. Proceedings of the 1992 MUFON UFO Symposium, pp. 102-128. Hawkinds, N. E., & Meyer, M. E. (1965). Time perception of short intervals during finished, unfinished and empty task situations. Psychonomic Science, 3, 473-474. Norman, P. (1979). Mystery deepens in pilot disappearance case. The MUFON Joumal, 141,s. Norman, P. (1991, March). On the UFO trail. Australian UFO Bulletin, 5. Ocean currents around Tasmania (1997). CSIRO Marine Research [on-line]. Available: http://www.marine.csiro.au/LeafletsFolder/tascurrents.html. Pinkney, J., & Ryzman, L. (1980). Alien Honeycomb. Sydney: Pan Books, pp. 81-85. Sturrock, P. A., et al. (1998). Physical evidence related to UFO reports: The proceedings of a workshop held at the Pocantico Conference Center, Tarrytown, New York, September 29-0ctober 4, 1997. Journal of Scientific Exploration, 12, 179-229. Valentich, G . (1980, September). My son Frederick. The Australian UFO Bulletin, 14. Journal of Scientific Exploration, Vol. 14, No. 1, pp. 35-52,2000 O 2000 Society for Scientific Exploration 0892-3310100 Protection of Mice from Tularemia Infection with Ultra-Low, Serial Agitated Dilutions Prepared from Francisella tularensk-Infected issue' Department of Family Medicine Uniformed Services University of the Health Sciences 4301 Jones Bridge Road Bethesda, MD 20814 e-mail: wjonas @rnxa. usuhs.mil The Department of Chemistry United States Naval Academy Annapolis, MD Abstract-Reports of immunomodulation with serial agitated dilutions (SADs) of cytokines, hormones, minerals, and whole tissue led to this inquiry as to whether exposure to a complex SAD preparation produced from Francisella tularensis-infected mice could alter the immune response and the effects of subsequent challenge with this pathogen in vivo. Six SAD preparations of reticuloendothelial tissue from I? ttularensis-infected C3HlHeN mice were produced through a process of serial loglo and logloo dilutions in 70% ethanol interspersed with 30-second agitation. SAD reparations were analyzed for protein content and for contamination with H-NMR spectroscopy. Three preparations contained detectable protein by Lowry and NMR analysis, and three were diluted beyond detection of protein. These preparations were administered orally for 1 month to 147 animals randomly assigned to SAD or dilutent control groups. All animals were then challenged with a lethal dose (LDS0or LD75) of F. tularensis and evaluated for time to death and total mortality. In a series of 15 trials, the SAD preparations consistently produced increased mean times to death (MTD; MTD SAD = 18.6 days [range, 12.9-25.61; MTD controls = 13.7 days [range, 11.6-15.6]), and decreased mortality (SAD: 53%; control: 75%) when compared with matched control groups given the dilutent only. Protection was not related to the level of dilution, the number of times vortexed, or the presence or absence of original substance from the tissue. Active and inactive solutions could be distinguished from one another using 'H NMR-spectroscopy. Two preparations induced specific anti-tularemia IgG antibody production before challenge. This anomalous finding needs independent repetition and further investigation. P Keywords: homeopathy - nosodes - tularemia - animal model - immunology - vaccination - ultra-low dilutions - SADs 35 36 W. B. Jonas & D. K. Dillner Introduction Dose-dependent reverse effects (DDREs), or hormesis, refers to the observation that when living organisms are exposed to very low doses of infectious or toxic agents, their reaction is often the opposite of that observed after exposure to high doses (Calabrese, McCarthy, and Kenyon, 1987; Luckey, 1980). Biological and physical agents of all types (including potent toxins, essential nutrients, pesticides, and penicillin) demonstrate DDREs on an equally broad range of living systems from single-cell bacteria and fungi to multicellular organisms and across a variety of phyla and species, including humans (Calabrese and Baldwin, 1997; Fursi, 1987; Neafsey, '1990; Stebbing, 1987). These reverse effects are assumed to occur because low-level exposures to otherwise damaging agents induce protective and reparative responses in cells (such as induction of heat shock proteins) without causing cellular damage (Boxenbaum, Neafsey, and Fourier, 1988; Stebbing, 1982; van Wijk and Wiegant, 1997). Recent studies report that under certain conditions, however, these effects occur with solutions diluted beyond the point when sufficient molecules remain to provide any specific molecular stimulus (Bastide, 1997; Endler and Schulte, 1994; Schulte and Endler, 1998). The majority of these studies report that vortexing the solution in a serial fashion during the dilution process can enhance or preserve this activity. Diverse effects from these serial agitated dilutions (SADs) have been reported in clinical studies (usually in reference to homeopathy; Boissel et al., 1996; Kleijnen, Knipschild, and ter Riet, 1991 ;Linde e t al., 1997) in toxicology (Anderson et al., 1990; Bascands et al., 1987; Cazin et al., 1987; Fisher and Capel, 1982; Fisher et al., 1987; Larue et al., 1985; Linde et al., 1994; Wagner, Kreher, and Jurcic, 1988), and in immunology (Bastide et al., 1987; Bastide, Doucet-Jaboeuf, and Daurat, 1985; Benveniste et al., 1991; Carriere, Dorfman, and Bastide, 1988; Daurat et al., 1986; Daurat, Dorfman, and Bastide, 1988; Davenas et al., 1988; Harisch and Kretschmer, 1988; Maddox, Randi, and Stewart, 1988; Poitevin, Davenas, and Benveniste, 1988; Sainte-Laudy and Belon, 1996b). Immunomodulating effects include increased cytotoxic and antibody-secreting cell production with SAD preparations of cytokines (Bastide, Doucet-Jaboef, and Daurat, 1985; Carriere, Dorfman, and Bastide, 1988; Daurat et al., 1986; Daurat, Dorfman, and Bastide, 1988), NK cell stimulation with hormones (Bastide et al., 1987; Daurat et al., 1986), histamine release, basophil degranulation and CD subset shifting with histamine and anti-IgE antibodies (Benveniste et al., 1991 ;Davenas et al., 1988; Harisch, Kretschmer, and von-Kries, 1987; Poitevin, Davenas, and Benveniste, 1988; Sainte-Laudy and Belon, 1996a, 1996b; others This research was conducted according to the principles set forth in the National Institutes of Health (NIH) Publication No. 85-23, Guide for the Care and Use of Laboratory Animals, and the Animal Welfare Act of 1986, as amended. The views, opinions and assertions expressed in this article are those of the authors and do not reflect official policy of the Department of Defense, the Department of Health and Human Services or the U.S. Government. SAD Prophylaxis of Infection 37 have reported no such effect from anti-IgE SADs, however: Hirst et al., 1993; Ovelgonne et al., 1992), and macrophage and mast cell activation with SAD preparations from minerals (Davenas, Poitevin, and Venveniste, 1987; Harisch and Kretschmer, 1988; Harisch and Kretschmer, 1989), nonspecific adjuvants (Poitevin, Davenas, and Benveniste, 1988), and whole tissue homogenates (Sainte-Laudy, Haynes, and Gershwin, 1986). Jacques Benveniste, a prominent immunologist and one of the foremost proponents of this ultrahigh SAD phenomenon, also claims that biologically active signals from SAD preparations can be captured and transmitted electronically to biological systems (Benveniste et al., 1997; Hadji, Arnoux, and Benveniste, 1991). Interestingly, homeopathic physicians since the 18th century have claimed that administration of SAD preparations of infectious tissue (called "nosodes" in that literature) can prevent the spread of epidemic diseases (Boenninghausen, 1887; Campbell, 1909; Castro and Nogueira, 1975; Davis, 1904; English, 1987; Fox, 1987; Gibson, 1958; Krishnamurty, 1970; Linn, 1904; Morgan, 1899; Rastogi and Sharma, 1992; Taylor-Smith, 1950). In addition, studies done in the first half of this century reported that SADs of infected tissue could alter immune reactions as evidenced by alteration of the Schick test (a now outdated method of screening for diphtheria immunity; Chavanon, 1932; Paterson and Boyd, 1941; Shephard, 1967). With rare exceptions, these studies have major flaws that make it impossible to judge the efficacy of this approach despite persistent claims. There are numerous speculative hypotheses as to how such information might be captured and stored in SAD preparations, if this indeed occurs. These hypotheses include differences in minority oxygen singlets (Berezin, 1990), water cluster formation in solvents (Anagnostatos, Pissis, and Viras, 1995), electromagnetic signal transfer (Endler et al., 1997), informational content of solutions (Bastide and Lagache, 1997), the interaction of "systemic memory" with complex feedback dynamics of living systems, and others (see Jonas and Jacobs, 1996, pp. 85-91 for a short summary). One current testable hypothesis is that it occurs through the arrangement of minority oxygen singlets in water or alcohol (Berezin, 1990). Consistent with this hypothesis are reports that the action of SAD-enhanced preparations are reduced or eliminated by extremely high or low temperatures, after exposure to electromagnetic radiation and when prepared in the absence of oxygen (Benveniste et al., 1991; Cazin et al., 1991). In addition, broadening in the hydroxyl region and other changes are said to occur in the nuclear magnetic resonance (NMR) spectra of these preparations (Demangeat, Gries, and Poitevin, 1997; Weingartner, 1989). In the present study, we analyzed whether a SAD preparation of infected tissue when given to mice can induce protection against an infectious challenge by the same organism. We chose murine tularemia (Francisella tularensis, live vaccine strain [LVS]) as the model because infection with this organism has been characterized in our lab and is rapidly lethal for C3HlHeN mice, providing a clear endpoint-death and days to death, as outcome measures (Fortier et 38 W. B. Jonas & D. K. Dillner preparations of tissue from F. tularensis-infected mice would respond no differently than would control mice given only dilutent to a subsequent challenge with a lethal dose of F. tularensis. Methods SAD Preparations Two specific pathogen free, 6-week old, male C3HlHeN mice (Harlan Sprague Dawley, Indianapolis, IN) were infected intranasally with a sublethal dose (lo3 colony forming units [CFUs]) of F. tularensis, LVS (ATCC, Rockville, MD). On Day 7 postinfection, the mice were sacrificed, and target organs for the infection (lungs, liver, and spleen) were removed and processed aseptically. These organs were placed in 15 ml gellsaline media and homogenized in a sterile Ten Broek homogenizer. Tissue homogenates were serially diluted in phosphate-buffered saline (PBS) and plated on cystine heart agar plates with Isovitalex (Difco, Detroit, MI) and Fe phosphate (Sigma, St. Louis, MO) in triplicate. Plates were incubated at 3 7 ' ~ 72 hours and colonies for counted to assess bacterial numbers in infected tissues. Tissue from two additional LVS-infected mice was homogenized in a 500 ml Wedgewood mortar with 60 mg of Lactose, U.S.P. (Pfanstiehl Lab., Waukegan, IL), for 1 hour and mixed with the liquid portion of tissue homogenate. Six SAD preparations were made from the mixed tissue homogenate in the following manner: 1. 0.1 ml of infected tissue homogenate was mixed with 0.9 ml of 70% ethanol in sterile, glass, 5 ml tubes. This mixture was agitated (as described below) for 30 seconds. This tube was labeled FtISAD-ld (F. tularensislagitated ultra-high dilution - 1st loglO[decimal] dilution). 2. 0.1 ml of FtISAD-ld then was mixed with 0.9 ml of 70% ethanol in a sterile 5 ml glass tube. This mixture was agitated again for 30 seconds. This tube was labeled FtISAD-2d. 3. The above procedure was repeated until serial agitated dilutions were prepared at the 3d, 7d, and 14d levels, each number representing the loglodilution of the original preparation. 4. Three more preparations were made using a homogenate-to-solute ratio of 1:100. This was done by mixing 0.01 ml of infected tissue homogenate with 0.99 ml of 70% ethanol in sterile, glass, 5-ml tubes and agitating for 30 seconds. This tube was labeled FtISAD-lc (F: tularensislagitated ultra-high dilution-1st log loo [centesimal] dilution). This procedure then was repeated as in the log 10 preparations until the 30c, 200c, and lOOOc levels were obtained. Each number in this series represents the logloodilution of the original preparation. Immunomodulation has been reported with these solutions in vitro, in vivo, and in clinical trials (Bastide, Doucet-Jaboeuf, and Daurat, 1985; Bastide et al., 1987; Benveniste et al., 1991 ; Carriere, Dorfman, and Bastide, 1988; Daurat et SAD Prophylaxis of Infection 39 al., 1986; Daurat, Dorfman, and Bastide, 1988; Davenas et al., 1988; Ferley et al., 1989; Harisch and Kretschmer, 1988; Linde et al., 1997; Papp et al., 1998; Poitevin, Davenas, and Benveniste, 1988; Reilly et al., 1994; Sainte-Laudy and Belon, 1996b). All agitation was performed manually (except the lOOOc preparation) by vertically shaking the solution vial vigorously onto a firm surface 60 times at the rate of two succusions per second. The 1000c preparation was prepared commercially using the 3c as a starting dilution (Quinn Pharmaceuticals, Berkeley, CA). Control groups received 70% alcohol agitated 60 times in an identical fashion. Analysis of SAD and Control Preparations Each SAD preparation was divided into three portions for analysis of bacterial count, protein content, and Proton-Nuclear Magnetic Resonance (IH NMR) spectroscopy. Bacterial count. We determined bacterial count (CFU) on the original homogenate as described under SAD Preparation. Each of the subsequent dilutions made with 70% alcohol had no viable bacteria on culture. Estimated nonviable bacterial concentration was made from the original homogenate. Protein analysis. Protein concentration was estimated by the Lowry spectrophotometric analysis method (Markwell et al., 1978). The sensitivity of this method is 0.0 1 mglml of protein. ' H ~ ~ ~ - s p e c t r o s cSamples were analyzed for 'H NMR-spectral differop~. ences in an attempt to distinguish between active and control solutions and to look for possible paramagnetic contaminants. NMR Spectroscopy was done at the Department of Chemistry of the U.S. Naval Academy, Annapolis, Maryland. Three cleaned and dried NMR sample tubes (Wilmad 528-PP, 5mm, Wilmad Glass Co., Buena, NJ) were filled with 0.8 ml of the test SAD preparation or control alcohol solution. The alcohol analyzed was the same 70% ethanol solution used to prepare the SADs. A single, clean, coaxial inner cell (Wilmad WGS-5BL) was filled with 0.2 ml of deuterium oxide (Isotec, Inc., Miamisburg, OH) and inserted into the sample tubes to provide the locking solvent for external referencing with each spectra. Recording was done with a General Electric NMR QE-300 spectrometer. Sample tubes were washed three times between each analysis with the solution to be evaluated. The central peak of CH3 was defined at 1.5 parts per million (ppm). The spectral areas of interest for this study were the H 2 0 and OH signals between 4.5 and 6 ppm, where evidence of increased proton exchange in the hydroxyl region of the solution was detected (Cooper, 1980). Spectral broadening was quantified by calculating the ratio of OH curve height to width at midaxis for controls and by subtracting the same ratio from each sample. Animal handling. All experiments were conducted at the barrier animal facility of the Department of Cellular Immunology, Walter Reed Army Institute of Research, Rockville, Maryland. The barrier facility is designed to inhibit 40 W. B. Jonas & D. K. Dillner spread of infectious agents from one animal cage to another, as well as from investigators to animals and from animals to personnel. Animal handling was performed in a laminar flow hood to ensure safety and prevent contamination of the animals. All animals received uniform handling, identical feed, water and environmentally controlled conditions, including constant temperature, humidity, and a 12-hour light-dark cycle. Randomization and matching. The animals were age-matched, 5- to 6week-old, male, C3HlHeN mice (Harlan Sprague Dawley). Seven SAD preparations were tested in a total of 15 trials using 8-12 mice per trial. For each trial, animals were randomly divided by an animal caretaker into two groups of 4-6 mice each and placed in separate cages. These cages then were randomly assigned to either the experimental (SAD) or control (70% ethanol) group. All animals were treated in a uniform way receiving the preparation (SAD or ethanol) at a dose of 0.03 ml per mouse given orally through a sterile pipette. All preparations (including controls) were shaken for 10 seconds before administered. Blinding. Expectation control of the experiment was done over 2 months by having research assistants who were completely blind to the nature of the experiment and to the hypothesized "active" and "inactive" preparations administer the SADs. Blinding of the tularemia challenge (the only maneuver theoretically possible to influence the outcome) was done by having a third party cover all cage labels for the animals to be challenged and having a single individual (W.J.) deliver all challenges in a uniform way on randomly selected cages. In this way, both topical and group blinding was maintained for the critical elements of the experiment. Prophylaxis of animals. The prechallenge dosage regime was one dose (0.03 ml), twice per day for 3 days followed by one dose every 3 days for 10 doses. After challenge, mice were given the SAD or control preparation twice a day for 20 days. Total number of doses was approximately 16 per mouse delivered over the 30 days prior to challenge and 40 doses delivered in the 20 days after challenge. The animals readily took this amount orally, obviating the need for direct gastric delivery, anesthesia, or multiple injections. Challenge procedure. On Day 30 from the start of prophylaxis (3 days after the previous SAD or control dose), all mice were challenged intranasally (in.) with either a LD50 or LD75 dose of LVS (lo3 or lo4 CFU, respectively). This dose will kill 50 to 75%, of untreated mice within 20 days. Animals were anesthetized with 0.06 mg of Telazol (Robbins, Richmond, VA), i.m., and then inoculated i.n. with LVS diluted in 0.05 ml PBS to the desired concentration. Bacterial counts were done on the challenge solution in duplicate both preand postchallenge. Two trials were conducted for each SAD preparation (three for FtISAD-14d) using a high (LD75) and low (LD5O) challenge dose. In addition, two trials were conducted in mice given a mixed SAD prophylaxis regime that dosed animals with ascending dilutions (FtISAD-14d, 30c, 200c, lOOOc each given twice for 1 week). These groups were called the ascending groups. SAD Prophylaxis of Infection Outcome Measures Antibody screening. Before starting SAD prophylaxis and again before infectious challenge, blood was drawn from each group through the lateral tail vein, pooled, centrifuged, and the serum frozen at -20°C. Serum was analyzed using enzyme-linked immunosorbent assay (ELISA) for antibody titers to whole organism LVS using the following procedure. Immulon plates were coated with 5 x lo6organisms and incubated overnight at 4OC. Plates then were washed three times with saline-Tween solution, blocked with 200 ul of 10% fetal bovine serum and rewashed. Each serum batch was screened for the presence of antibodies by adding serial 10-fold dilutions of serum (100 pllwell) and incubated for 90 minutes at 37°C and washed. One tenth pglwell of peroxidase labeled goat antimouse Ig antibody was added and incubated for 90 minutes at 37OC. Substrate (KPL solution A and B, 1:1) was then added at 100 p1 per well and color read at 410 nm in 30 minutes. A reading was considered positive if optical density of the test serum was equal to or greater than positive control serum produced by LVS vaccinated mice. Mortality and time to death. The two primary outcome measures were time to death and mortality. Animals were checked twice daily after infectious challenge and the death day for each animal recorded. The experiment was terminated 30 days postchallenge. Data analysis. Data were entered into a spreadsheet format (Microsoft Excel, Microsoft Corp., Redmond, WA) and analyzed by standard statistical software packages (Statview 11, SuperAnova; Abacus Concepts, Berkeley, CA; and JMP, SAS Institute, Cary, NC). To control for the influence of censured data from mice that survived challenge, both parametric and nonparametric methods were used. The difference between harmonic mean death day (the mean reciprocal death time, MRDT) of matched groups was evaluated using a two-tailed, paired t test and the Wilcoxon sign rank test. The MRDT was the sum of one over each death day divided by the total number of mice in the group. This calculation gave an estimate of resistance to infection in the group based on time to death and overall mortality while controlling for censured and outlying data from surviving animals. The value obtained was between 0 and 1, in which 0 = total resistance and 1 = death of all animals on the first day. Death-day data were also analyzed without this transformation using the Mann-Whitney U nonparametric test. Mortality ratios were analyzed with standard chi-square methods, and stratified mortality was evaluated using the Mantel-Haenszel rate ratio method. One-way analysis of variance (ANOVA) was used to evaluate the influence of SAD treatment effect while controlling for challenge dose with both death day and MRDT as outcome measures. 42 W. B. Jonas & D. K. Dillner Results Bacterial Count Bacterial counts done on the stock homogenate showed these tissues harbored lo7 CFU of LVS. No viable bacteria were found in any of the SAD preparations nor in the two alcohol control preparations that were cultured. Estimates of the number of nonviable bacteria in the SADs indicated that only the FtISAD-3d and 7d preparations contained bacteria: lo4and 1 organism, respectively (Table 1, column 2). Protein Analysis Protein concentration was 0.55 mglml in the initial homogenate, 0.098 mglml in the FtISAD-3d preparation and less than 0.001 mglml in all other SAD preparations and the two controls tested. Estimated protein content for the FtISAD-7d and 14d preparations were 1 nglml and 0.1 fglml, respectively (Table 1, column 3). The calculated amount of protein given to each mouse over the 45-day experiment was 16.5 micrograms, 1.68 nanograms, and 0.17 fentograms for the 3d, 7d, and 14d levels, respectively (Table 1, column 4). Preparations in the centesimal series had no remaining protein, either detectable or theoretical. NMR Spectroscopy The 'H NMR-spectra of the SAD groups were easily distinguished from control groups regardless of the dilution level (Table 2). All SAD preparations had a moderate to large amount of OH spectral broadening compared with control groups. The average height to width ratio of SAD preparations was 3.4 (range 1.8-4.8) compared with control preparations where the height-to-width TABLE 1 Analysis of SAD Preparations Preparation Homogenate 3d 7d 14d 30c 200c 1oooc Control 1 Control 2 CFU count Protein 550 (glml) 9.8 (glml) (1 nglml) (0.1 fglml) 0 0 0 0 0 Proteinlmouse NA 16.5 (g) (1.68 ng) (0.17 fg) 0 0 0 0 0 Note: CFU = colony forming units. Analysis is of CFU, protein concentration, and total protein received per mouse during prophylaxis. Numbers in parenthesis are calculations based on analysis of the stock homogenate. SAD Prophylaxis of Infection TABLE 2 NMR Spectra of SAD Preparations Preparation 3d 7d 14d 30c 200c lO0Oc Mean Control Height = width ratio OH curvea broadening 43 Note: Results of NMR spectroscopy evaluating OH spectral curve broadening between 4.5 and 6 PPm. a Height=width ratio of control preparation minus the height=width ratio of SAD preparation. Higher numbers indicate greater broadening. ratio was 11. The 30c and 200c preparations had height to width ratios of 4.8 and 3.0, respectively. NMR spectroscopy was not done on the commercial SAD preparation because the alcohol concentration of this preparation was not identical to other SADs (50% vs 70% ETOH). Evaluation of Prophylaxis Antibody screening. Serum from the 30c- and 200c-treated mice was positive for anti-tularensis antibodies after Week 4 of SAD treatment. Anti-tularensis antibodies were not found in other SAD-treated groups. Titer endpoints were 1: 100 or less before prophylaxis and greater than 1: 10,000 just before challenge in the 30c- and 200c-treated mice and in the LVS-vaccinated TABLE 3 Anti LVS Antibodies in SAD Groups Preparation 3d 7d 14d 30c 200c 1oooc Ascending Positive control Negative Control Note: LVS = live virus strain; AUHD = agitated ultra-high dilutions; ND = Not done. Endpoint anti-tularensis antibody titers of pooled serum from mice before SAD treatments and just prior to LVS challenge. a This sample was drawn 3 days after start of SAD treatment. This sample was drawn 15 days before LVS challenge. Pre = AUHD treatment Pre = LVS challenge ~ 44 W. B. Jonas & D. K. Dillner mice (positive controls). Titer endpoints were always 1:100 or less in the 3d, 7d, 14d, 1000c, and ascending groups and in the normal (untreated) mice (Table 3). MRDT. The MRDT was calculated for SAD-treated and control mice for each matched trial. MRDT ranged from 0 to 0.16 for all trials and, as expected, was higher in the trials with a higher challenge dose. SAD-treated mice consistently showed reduced MRDT (increased resistance to infection) over their Mean Reciprocal Death Times SAD Level Fig. 1. Mean reciprocal death times (MRDT) of SAD-treated mice (first columns) and their matched control groups (second columns) over 15 trials. Lower MRDT scores indicate increased resistance to infection. First trial listed for each SAD level is the LDS0challenge. Second and third trial listed is the L D 7 challenge. ~ SAD Prophylaxis of Infection TABLE 4 Analysis of Mortality After SAD Treatments Mortality Preparation 3d 7d 14d 30c 200c 1oooc Ascending Dilutions(3d-14d) Solutions(30-As) All Groups a Relative No. in group SAD Control Relative riska Note: Mortality of mice treated with each SAD preparation and matched control group. risk after stratification. ant el-~aenszel chi-square for all groups (95%CI 1.23-2.92, p c .005). matched control group (mean difference = .03). All but two of the 15 trials showed a reduced MRDT in the SAD group (Wilcoxon sign rank test, z = 3.408, p = .0007, paired t test, t = 7.518, p = .0001, two-tailed). There was no relationship between level of dilution in the SAD preparations and protection. All preparations except FtISAD- 1000c demonstrated protection (Figure 1). These results were checked directly against actual death day using the raw data for all SAD groups with Day 30 as the cut off. Mean time to death for SAD-treated mice was 18.6 (range 12.9-25.6) and for controls was 13.7 (Range: 11.6-15.6). The Mann Whitney U test using death day for all groups was 3.087, ( p = .002), consistent with the MRDT evaluation. SAD treatment delayed death by approximately 5 days. One way ANOVA showed that SAD treatment had a significant impact on death day (F = 11.5 1, p = .0009) and MRDT (F= 4.96, p = .037) even after controlling for challenge dose (Table 5). Mortality. Each group was monitored daily for deaths up to 30 days postchallenge. Relative risk was above 1 in all groups except FtISAD-1000c. The Mantel-Haenszel rate ratio (after stratification into treatment groups) was not different than the crude rate ratio for all groups, indicating that no individual preparation disproportionally influenced the overall result (M-H RR 1.89; 95%CI 1.23-2.92, two-tailed p = .0037; Table 4). Overall mortality was 53% in the SAD group and 75% in the matched control group with little difference between solution or dilution groups. Chi-square analysis on the proportion of deaths in each group was 8.41 ( p = .0037). SAD treatment prevented 22% of deaths. The number of organisms required to kill 50% of the animals (LDSO) in the control group was 6,000 CFU compared with 20,000 CFU in the SADtreated group (Table 5). 46 W. B. Jonas & D. K. Dillner TABLE 5 Summary Evaluation of SAD Effects Outcome SAD Control 13.7 0.10 75% 7300 6000 Evaluation p value 18.6 MDT (days) 0.07 MRDT By nonparametric test Controlled for challenge dose Mortality 53% Prevented fraction of deaths by SAD preparations = 22% Challenge dose (CFU) 7300 LD5o (CFU) 20,000 z = 3.08 t=7.518 z = 3.41 F = 4.96 = 8.41 x2 .002a .0001 .0007' .037~ .0037 Note: CFU = colony forming units; MDT = mean time to death; MRDT = mean reciprocal death time. Summary evaluations of mortality (MDT) and MRDT in all groups. a Mann-Whitney U test. Students two-tailed t-test. Wilcoxon sign rank test. way ANOVA on mean time to death after controlling for challenge dose. Conclusion The results in this series of experiments did not confirm the null hypothesis that ultra-low, SAD dilutions prepared from infected tissue acted like chemically identical controls in their ability to protect against subsequent bacterial challenge by the same organism. SADs reduced overall mortality by 22%, delayed death by 5 days, and tripled the number of organisms required to kill 50% of the animals. Two SADs (30c and 200c) stimulated the production of specific anti-tularensis antibodies in animals after 1 month of exposure. No protein was detected in SAD preparations, but they could be distinguished from control solutions with 'H NMR-spectroscopy. All SAD preparations, except the 1000c, consistently imparted some protective effect. Environmental exposures while in transit or differences in the method of preparation of the commercial SAD may have altered the lOOOc solution, rendering it different than those produced in our lab. Data from studies using the SADs made at our lab suggest protection was not related to level of dilution, number of times vortexed, or absence of bacteria or protein in the preparations. The lack of relationship between response and dilution level is probably a reflection of the wide gap between dilutions selected for testing. In vitro work on SADs in which multiple serial, 10-fold dilutions are examined usually report a quadratic or sinusoidal pattern (rather than a linear or logarithmic) doseresponse relationship (Boxenbaum, Neafsey, and Fournier, 1988; Davenas, Poitevin, and Benveniste, 1987; Davenas et al., 1988; Poitevin, Davenas, and Benveniste, 1998; Schulte and Endler, 1998). Peaks and valleys of effect are found when multiple and closely spaced dilutions are tested. We tested only six widely spaced dilutions and so may have missed an existent dose-response relationship. SAD Prophylaxis of Infection 47 SAD solutions could be distinguished from control solutions using 'H NMR-spectroscopy; however, the relevance of these findings in unclear. Broadened hydroxyl band width can occur from a variety of causes, including contamination with paramagnetic particles or increased proton exchange in the OH bonding region. This study was not designed to control for these possibilities, and therefore, no conclusions can be made about this finding. Reports in the literature testing the SAD prophylaxis hypothesis in laboratory models or in veterinary populations have yielded mixed results. Taylor reported no prophylactic effect from a SAD preparation derived from the husk of the lungworm Dictyocaulus vivparus in experimental lung infection in cows after there were promising reports from veterinarians (Taylor, Mallon, and Green, 1989). The experiment used the 30c preparation, dosed the animals four times with the SAD over a month, and gave no doses after challenge. Study power was low. Oberbaum, Weismann, and Bentwich (1989) reported on a study on a retroviral infection with a SAD preparation. This study used SAD preparations (range 6d-30d) derived from C type B-trophic retrovirus (LP-BM5 MuLV), which produces a murine AIDS-like syndrome (MAIDS). Groups of mice were treated either orally or by injection (i.p.) three times a week, but, in contrast to Taylor, SAD preparations were given only after the infectious challenge. Animals given the 12d SAD prepared from stock containing less than 5 infectious units of virus had significant reductions in spleen weights (the primary marker of the disease) compared with saline-alcohol treated animals. No other level of SAD demonstrated an effect (Oberbaum, Weismann, and Bentwich, 1989). Day reported that a nosode prepared from parvovirus stopped an epidemic of kennel cough in dogs (Day, 1987). No control group was included, however, and therefore, no comparison with the natural course of the epidemic could be made. Apparently, the use of nosodes and other SAD and homeopathic preparations are popular among European animal farmers for the control of epidemic outbreaks. Several reports of SAD use for control of bovine mastitis have been reported (Day, 1986; Merck, Sonnenwald, and Rollwage, 1989). Langer reported on a series of field tests and controlled experiments on bovine mastitis using various mixtures of nosodes, SADs, and homeopathic remedies that were both protective and therapeutic (Langer, 1990). What is the mechanism of SAD-induced protection? Besides artifact or experimental error, the most likely explanation is that SAD preparations of LVSinfected tissue contain sufficient immunogenic signals to stimulate an immune response in recipients. Protection does not seem dependent on humoral immunity because anti-tularemia antibodies were detected only in serum from 30cand 200c-treated mice (Table 3). This is not unusual in tularemia infection, however, in which the primary protective mechanism is thought to be cellular rather than humoral (Tarnvik, 1989). Specific antibody induced by SAD treatment, then, cannot account for protection seen in all groups treated with SAD treatment. Induction of specific antibody response to SAD treatments on other 48 W. B. Jonas & D. K. Dillner systems is also variable. Davies reported no antibody production 2 weeks after three doses of a SAD preparation from influenza A2 (Davies, 1971). Weisman and colleagues, on the other hand, have reported highly significant modulation of specific antibody generation in mice using SAD preparations of K-lipid hemocyanin when given three times a week for 8 weeks (Weisman et al., 1997). Neither of these investigators included cytokines, tissue protein, or other adjuvants reported in vitro to modulate immune function when prepared as SADs. If ultra-low dose SADs do have specific effects, one can hardly expect those effects to be large or induced after only a few doses over a short time period. No research has been done to systematically investigate these questions. What is the active agent in the final SAD solutions? Of the many speculative hypotheses previously discussed, one is that stable isotopic positional correlations (IPCs) of minority oxygen singlets are produced around molecules in the stock solutions and subsequently "locked in" during agitation by the polaronic self-stabilization properties of nonlinear media (Berezin, 1990). IPCs, then, might mimic the original molecules through interaction with receptors and enzyme systems on or in the cell. This hypothesis takes into consideration current information about SAD action and is testable by manipulating the concentration of oxygen isotopes in the solution as it is prepared. In addition, the molecular organization of these solutions could be further evaluated using techniques such as thin layer crystallography, NMR, infrared, Raman, or other spectrographic methods (del Giudice, Preparata, and Vitiello, 1988; Popp, Li, and Mei, 1989). One currently popular hypothesis is that water can hold and deliver specific electromagnetic signals (with coherent wavelengths similar to a laser) to which biological systems respond (Benveniste et al., 1997). Others claim these observed effects are all attributable to experimental error and artifact (Maddox, 1988; Metzger, 1988; Plasterek, 1988; Seagrave, 1988). We currently are attempting to replicate these findings under more stringent conditions (e.g., total blinding and using uniform aerosol dosing methods, additional control groups, etc.) Whatever the mechanism, protection produced by SAD preparations in this study was unexpected yet consistent over 15 trials. If these effects represent a real and general phenomenon, SAD preparations might provide a simple and rapid method of protecting against infectious agents for which we do not yet have adequate prophylactic or therapeutic regimes or to agents with emerging resistance to current treatments. This is especially important as the world becomes increasingly mobile and the chance for rapid spread of unusual or resistant infectious pathogens increases. Further investigation into this phenomenon is warranted. Acknowledgments The authors thank Terri Western, Department of Cellular Immunology, for her help in blinding, randomization, and SAD administration; Robert Burge, Division of Biometrics, for his help with statistical analyses of our data; Au- SAD Prophylaxis of Infection 49 gust Salvado, past Director of WRAIR, and Ann Fortier and Carol A. Nacy, previously at the Department of Cellular Immunology, for their support and critical analysis of the project. References Anagnostatos, G. S., Pissis, P., & Viras, K. (1995). Possible water cluster formation by dilution and succussions. In Anagnostatos, G. S., and von Oertzen, W. (Eds.), Atomic and nuclear clusters (pp. 215-217). Heidelberg, Germany: Springer-Verlag. Anderson, D., Fisher, P., Francis, A. J., Philipps, B. J., & Jenkinson, P. C. (1990). Studies of the adaptive repair response in cultured cells. In Proceedings of the First International Congress on Ultra Low Doses, 1, A4.2. Bordeaux, France: University of Bordeaux. Bascands, J. L., Girolami, J. P., Cabos, C., Pecher, C., Bompart, G., Suc, J. M., & Manuel, Y. (1987). Low dose of cadmium (Cd) protects rat glomerular cell culture against the direct toxic effect of cadmium. Heavy Metals in the Environment, 2, 118-120. Bastide, M. (1997). Signals and images. Dordrecht, The Netherlands: Kluwer Academic Publishers. Bastide, M., Daurat, V., Doucet-Jeboeuf, M., Pelegrin, A., & Dorfman, P. (1987). Immunomodulator activity of very low doses of thymulin in mice. International Journal of Immunotherapy, 3, 191-200. Bastide, M., Doucet-Jaboeuf, M., & Daurat, V. (1985). Activity and chronophannacology of very low doses of physiological immune inducers. Immunology Today, 6, 234-235. Bastide, M., & Lagache, A. (1997). A communication process: A new paradigm applied to highdilution effects on the living body. Alternative Therapies in Health and Medicine, 3, 35-39. Benveniste, J., Davenas, E., Ducot, G., Cornillet, B., Poitevin, B., & Spira, A. (1991). L'agitation de solutions hautement kiluees n'indiut pas d'activite specifique. Comptes Rendus de 1 'Academie des Sciences Paris, 312, 461-466. Benveniste, J., Jurgens, P., Hsueh, W., & Aissa, J. (1997). Transatlantic transfer of digitized antigen signal by telephone link. Journal of Allergy and Clinical Immunology [abstract], 99, S175. Berezin, A. A. (1990). Isotopical positional correlations as a possible model for Benveniste experiments. Medical Hypotheses, 31,43-45. Boenninghausen, C. (1887). Prophylaxis. Homoeopathic Physician, 7,63-64. Boissel, J . P., Cucherat, M., Haugh, M., & Gauthier, E. (1996). Critical literature review on the effectiveness of homoeopathy: Overview of data from homoeopathic medicine trials. Report to the European Commission, Brussels. Homoeopathic Medicine Research Group, Sussex, UK. Boxenbaum, H., Neafsey, P. J., & Fournier, D. J. (1988). Hormesis, Gompertz functions, and risk assessment. Drug Metabolism Reviews, 19, 195-229. Calabrese, E. J., & Baldwin, L. A. (1997). The dose determines the stimulation (and poison): development of a chemical hormesis database. International Journal of Toxicology, 16, 545-559. Calabrese, E. J., McCarthy, M. E., & Kenyon, E. (1987). The occurrence of chemically induced hormesis. Health Physics, 52, 531-41. Campbell, N. (1909). Reasons for having faith in the nosodes. Medical Advances, 47L, 633-644. Carriere, V., Dorfman, P., & Bastide, M. (1988). Evaluation of various factors influencing the action of mouse a,B interferon on the chemiluminescence of mouse peritoneal macrophages. Annual Review of Chronophamaco1ogy, 5,9-12. Castro, D., & Nogueira, G. (1975). Use of the nosode meningococcinum as a preventive against meningitis. Journal of the American Institute of Homeopathy, 68, 2 11-219. Cazin, J. C., Cazin, M., Chaoui, A., & Belon, P. (1991). Influence of several physical factors on the activity of ultra low doses. In Doutremepuich, C. (Ed.), Ultra low doses (pp. 69-80). Washington, DC: Taylor & Francis. Cazin, J. C., Cazin, J. L., Gaborit, A., Chaoui, J., Boiron, J., Belon, P., Chenuault, Y., & Papapanayotou, C. (1987). A study of the effect of decimal and centesimal dilutions of arsenic on the retention and mobilization of arsenic in the rat. Human Toxicology, 6, 315-320. Chavanon, P. (1932). La diphterie. Niort, France: Imprimerrie St.-Denis. Cooper, J. W. (1980). Spectroscopic techniques for organic chemists. New York: John Wiley & Sons. Daurat, V., Carriere, V., Douylliez, C., & Bastide, M. (1986). Immunomodulatory activity of thy- W. B. Jonas & D. K. Dillner mulin and a,D interferon on the specific and the nonspecific cellular response of C57BLl6 and NZB mice. Zmmunobiology, 173, 188. Daurat, V., Dorfman, P., & Bastide, M. (1988). Immunomodulatory activity of low doses of interferon a,D in mice. Biomedicine and Pharmacotherapy, 42, 197-206. Davenas, E., Beauvais, J., Oberbaum, M., Robinzon, B., Miadonna, A., Tedeschi, A., Pomeranz, B., Fortner, P., Belon, P., Sainte-Laudy, J., Poitevin, B., & Benveniste, J. (1988). Human basophil degranulation triggered by very dilute antiserum against IgE. Nature, 333, 816-818. Davenas, E., Poitevin, B., & Benveniste, J. (1987). Effect on mouse peritoneal macrophages of orally administered very high dilutions of silica. European Journal of Pharmacology, 135, 313-319. Davies, A. E. (1971). Clinical investigations into the action of potencies. British Homoeopathic Journal, 60, 37-41. Davis, J. J. (1904). Some experience in the prophylaxis of variola. Medical Advances, 42, 225-227. Day, C. (1986). Clinical trial in bovine mastitis. British Homoeopathic Journal, 75, 11-14. Day, C. E. I. (1987). Isopathic prevention of kennel cough-Is vaccination justified? International Journal of Veterinary Homeopathy, 2,45-50. del Giudice, E., Preparata, G., & Vitiello, G. (1988). Water as a free electric dipole laser. Physical Review Letters, 61, 1085-1088. Demangeat, J. L., Gries, P., & Poitevin, B. (1997). Modification of 4 MHz N.M.R. water proton relaxation times in very high diluted aqueous solutions. In Bastide, M. (Ed.), Signals and images (pp. 95-110). Dordrecht, The Netherlands: Kluwer Academic Publishers. Endler, P. C., Pongratz, W., Smith, C. W., Schulte, J., Senekowitsch, F., & Citro, M. (1997). Nonmolecular information transfer from thyroxine to frogs. In Bastide, M. (Ed.), Signals and images (pp. 149-160). Dordrecht, The Netherlands: Kluwer Academic Publishers. Endler, P. C., & Schulte, J. (1994). Ultra high dilution: Physiology and physics. Dordrecht, The Netherlands: Kluwer Academic Publishers. English, J. M. (1987). Pertussin 30-Preventive for whooping cough? British Homoeopathic Journal, 76, 61-65. Ferley, J. P. D., Zmirou, D., D7Adhemar,D., & Balducci, F. (1989). A controlled evaluation of a homeoeopathic preparation in influenza-like syndromes. British Journal of Clinical Pharmacology, 27, 329-335. Fisher, P., & Capel, I. (1982). The treatment of experimantal lead intoxication in rats by penicillinamine and plumbum metallicum. Journal of the Society of Ultramolecular Medicine, I , 30-31. Fisher, P., House, I., Belon, P., & Turner, P. (1987). The influence of the homoeopathic remedy plumbum metallicum on the excretion kinetics of lead in rats. Human Toxicology, 6, 321-324. Fortier, A. H., Slayter, M. V., Ziemba, R., Meltzer, M. S., & Nacy, C. A. (1991). Live vaccine strain of Francisella tularensis infection and immunity in mice. Infection and Immunity, 59, 2922. Fox, A. D. (1987, April). Whooping cough prophylaxis with pertussin 30. British Homoeopathic Journal, 69-70. Furst, A. (1987). Hormetic Effects in pharmacology: Pharmacological inversions as prototypes for hormesis. Health Physics, 52, 527-530. Gibson, D. M. (1958). Nosodes and prophylaxis. Homoeopathy, 8, 11 1-124. Hadji, L., Amoux, B., & Benveniste, J. (199 1). Effect of dilute histamine on coromary flow of guinea-pig isolated heart. Inhibition by a magnetic field. Federation of Associations and Society of Experimental Biology Jounal, 5, 7040. Harisch, G., & Kretschmer, M. (1988). Smallest zinc quantities affect the histamine release from pertoneal mast cells of the rat. Experientia, 44, 761-762. Harisch, G., & Kretschmer, M. (1989). Histamine release from rat peritoneal mast cells after oral doses of homeopathically prepared minerals and disodium cromoglycate. Journal of Applied Nutrition, 41, 45-49. Harisch, V. G., Kretschmer, M., & von-Kries, U. (1987). Beitrag zum histamin release aus peritonealen mastzellen von mannlichen wistar-ratten. DTW Deutsche Tierarztliche Wochenschrijt, 94,497-540. Hirst, S. J., Hayes, N. A., Burridge, J., Pearce, F. L., & Foreman, J. C. (1993). Human basophil degranulation is not tri gg ered by very dilute anti-serum against human IGE. ~ a t u i e 366, , 525-527. SAD Prophylaxis of Infection 51 Jonas, W. B., &Jacobs, J. (1996). Healing with homeopathy. New York: Warner. Kleijnen, J., Knipschild, P., & ter Riet, G. (1991). Clinical trials of homoeopathy. British Medical Journal, 302, 316-323. Krishnamurty, P. S. (1970). Report on the use of influenzinum during the outbreak of epidemic in India in 1968. Hahnemanian Gleanings, 37, 225-226. Langer, P. H. (1 990). Summary of evidence for the efficacy of Para Plus in the treatment of mastitis. Ontario, Canada: Dorchester Laboratories. Larue, F., Dorian, C., Cal, J. C., Guillemain, J., & Cambar, J. (1985). Influence de pretraitment de dilutions infinitesimales de mercurius corrosivus sur la mortalite induite par le chlorure mercurique. Nephrologie, 6, 86. Linde, K., Clausius, N., Ramirez, G., Melchart, D., Eitel, F., Hedges, L. V., & Jonas, W. B. (1997). Are the clinical effects of homoepathy placebo effects? A meta-analysis of placebo-controlled trials. Lancet, 350, 834-843. Linde, K., Jonas, W. B., Melchart, D., Worku, F., Wagner, H., & Eitel, F. (1994). Critical review and meta-analysis of serial agitated dilutions in experimental toxicology. Human and Experimental Toxicology, 13,481-492. Linn, A. M. (1904). Variolinum, a prophylaxis against smallpox. Medical Advances, 42, 431-436. Luckey, T. D. (1980). Hormesis with ionizing radiation. Boca Raton, FL: CRC Press. Maddox, J., Randi, J., & Stewart, W. W. (1988). "High-dilution" experiments a delusion. Nature, 334,287-290. Markwell, M. A. K., Hass, S. M., Bieber, L. L., & Torbert, N. E. (1978). Modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Analytical Biochemistry, 87, 206-210. Merck, C., Sonnenwald, B., & Rollwage, H. (1989). The administration of homeopathic drugs for the treatment of acute mastitis in cattle. Berliner und Miinchener Tierarztliche Wochenschrift, 102,266-272. Metzger, H. (1988). Only the smile is left. Nature, 334, 375-376. Morgan, W. L. (1899). Epidemics, endemics, and contagions. Homoeopathic Physician, 19, 479-488. Neafsey, P. J. (1 990). Longevity hormesis. A review. Mechanisms ofAgeing and Development, 51, 1-31. Oberbaum, M., Weismann, Z., & Bentwich, Z. (1989). Treatment of experimentally-induced AIDS in mice by very high dilutions of virus, a preliminary study. In Bastide, M. (Ed.), Signals and images (pp. 127-129). Paris, France: Atelier Alpha Bleue. Ovelgonne, J. H., Bol, A., Hop, W., & van Wijk, R. (1992). Mechanical agitation of very dilute antiserum against IgE has no effect on basophil staining properties. Experientia, 48, 504-508. Papp, R., Schuback, G., Beck, E., Burkard, G., Bengel, J., Lehrl, S., & Belon, P. (1998). Oscillococcinum in patients with influenza-like syndromes: a placebo-controlled, double-blind evaluation. British Homoeopathic Journal, 87, 69-76. Paterson, J., & Boyd, W. E. (1941). Potency action. A preliminary study of the alteration of the Schick-test by a homoeopathic potency. British Homoeopathic Journal, 31, 301-309. Plasterek, R. (1988). Explanation of Benveniste. Nature, 334, 285-286. Poitevin, B., Davenas, E., & Benveniste, J. (1988). In vitro immunological degranulation of human basophils is modulated by lung histamine and Apis mellifica. British Journal of Clinical Pharmacology, 25,439-444. Popp, F. A., Li, K. H., & Mei, W. P. (1989). Physical aspects of biophotons. Experientia, 44, 576-685. Rastogi, D., & Sharma, V. (1992). Study of homoeopathic drugs in encephalitis epidemic (1991) in Uttar Pradesh (India). Central Council for Research on Homeopathy Quarterly Bulletin, 14, 1-11. Reilly, D., Taylor, M. A., Beattie, N., Cambell, J. H., McSharry, C., Aitchison, T., Carter, R., & Stevenson, R. D. (1 994). Is evidence for homeopathy reproducible? Lancet, 344, 1601-1606. Sainte-Laudy, J., & Belon, P. (1996a). Analysis of immunosuppressive activity of serial dilutions of histamine on human basophil activation by flow cytometry. Injlammation Research, 45, S33-S34. Sainte-Laudy, J., & Belon, P. (1996b). Inhibition of basophil activation by high dilutions of histamine. Agents Actions, Supplements, 38, C245-C247. Sainte-Laudy, J., Haynes, D., & Gerswin, G. (1986). Inhibition effects of whole blood dilutions on basophil degranulation. International Journal of Zmmunotherapy, 2, 247-250. 52 W. B. Jonas & D. K. Dillner Schulte, J., & Endler, P. C. (1998). Fundamental research in ultra-high dilutions and homeopathy. Dordrecht, The Netherlands: Kluwer Academic Publishers. Seagrave, J. C. (1988). Evidence of non-reproducibility. Nature, 334, 559-559. Shephard, D. (1967). Homeopathy in epidemic diseases. Robinson, G. E. (Ed.). Rusington, England: Health Science Press. Stebbing, A. R. (1987). Growth hormesis: A by-product of control. Health Physics, 52, 543-547. Stebbing, A. R. D. (1982). Hormesis-The stimulation of growth by low levels of inhibitors. Science of the Total Environment, 22, 2 13-234. Tarnvik, A. (1989). Nature of protective immunity to Franciscella tularensis. Review of lnfectious Disease, 11,440-451. Taylor, S. M., Mallon, T. R., & Green, W. P. (1989). Efficacy of a homoeopathic prophylaxis against expeirmental infection of calves by the bovine lungworm Dictyocaulus viviparus. Veterinaly Record, 124, 15-17. Taylor-Smith, A. (1950). Poliomyelitis and prophylaxis. British Homoeopathic Journal, 40, 65-77. van Wijk, R., & Wiegant, F. A. C. (1997). The similia principle in surviving stress: Mammalian cells in homoepathy research. Utrecht, The Netherlands: Homeopathy International. Wagner, H., Kreher, B., & Jurcic, K. (1988). In vitro stimulation of human granulocytes and lymphocytes by pico- and fentogram quantities of cytostatic agents. Arzneimittel-Forschung (Drug Research), 38,273-275. Weingartner, 0 . (1989). NMR-Spektren von sulfur-potenzen. Therapeutikon, 3, 438-442. Weisman, Z., Oberbaum, M., Topper, R., Harpaz, N., & Bentwich, Z. (1997). High dilutions of antigens modulate the immune response to KLH. In Bastide, M. (Ed.), Signals and images (pp. 179-190). Dordrecht, The Netherlands: Kluwer Academic Publishers. Journal of Scientific Exploration, Vol. 14, No. 1, pp. 53-72,2000 0892-33 10/00 O 2000 Society for Scientific Exploration The Correlation of the Gradient of Shannon Entropy and Anomalous Cognition: Toward an AC Sensory System The Laboratories for Fundamental Research 330 Cowper Street, Suite 300 Palo Alto, CA 94301 email; may@ P o r g Abstract- In this study, we hoped to replicate earlier findings that have demonstrated strong evidence for anomalous cognition (AC), as well as a significant correlation between the quality of the AC with the gradient of Shannon entropy, but not with the entropy itself. We created a new target pool and a more sensitive analytical system compared with those of earlier studies. We then invited five experienced receivers (i.e., experiment participants) to contribute 15 trials each. In addition to the usual rank-order analysis, two other methods were used to assess the quality of the AC. The first of these was a 0 to 7 rating scale that has been used in the earlier studies. The second, a figure of merit, was based on a fuzzy-set encoding of the targets and responses. The primary hypotheses were (a) that a significant correlation would be seen between the figure of merit quality assessment and the gradient of Shannon entropy for the associated target and (b) that the correlation using the rating assessment would be consistent with earlier findings. A secondary hypothesis was that the figure of merit quality would not correlate with the entropy of the associated target. All hypotheses were confirmed. Our results are part of the growing evidence that AC is mediated through a sensory channel. Keywords: anomalous cognition - pattern analysis -entropy gradient Introduction Lantz, Luke, and May (1994) reported on two experiments, the first of which was conducted in 1992, to test sender condition and target type in an anomalous cognition (AC) experiment. The hypotheses in these studies addressed whether a sender is necessary for AC information transfer and whether AC performance differs when the targets are static photographs or dynamic material, such as videotape. Lantz, Luke, and May (1994) found that a sender is not a necessary component in a successful AC experiment and that the data supported, but not significantly, a target-type preference in favor of static material. Because there were no significant interactions in a 2 x 2 analysis of variance (ANOVA), the data were combined across the sender versus no-sender condition. Blind ranking achieved a sum-of-ranks for the static targets of 265 in which the chance expectation was 300, leading to an effect size of 0.248 andp = .007. The analysis 54 E. May, J. Spottiswoode, & L. Faith of the 100-trial dynamic targets led to a sum of ranks of 300, an effect size of 0.000, andp = .500. A second experiment was conducted one year later and was also reported in Lantz, Luke, and May (1994). In that study, a sender was not used, and the protocol differed considerably from the first experiment. Four participants contributed a total of 45 trials in two target-type conditions, leading to a combined effect size of 0.550 for the static targets and the same value for the dynamic ones. Lantz, Luke, and May (1994) discussed the apparent contradiction between the results of their two studies. They speculated that the static targets were better in their first study because of a lack of content parity between the static and dynamic targets. Nonetheless, this does not explain the similarity between static and dynamic targets in their second study. In addition, their results are inconsistent with those of some of the Ganzfeld research regarding static versus dynamic targets. Bem and Honorton (1994) found that dynamic targets produced better results in the Ganzfeld experiment than did static targets. The data from both of Lantz, Luke, and May (1994) studies were analyzed to investigate whether AC performance depended on the gradient of Shannon entropy of the targets (May, Spotiswoode, and James, 1994a). This idea arose from our laboratory's anecdotal evidence that AC functioned particularly well when targets were especially dynamic. That is, when targets involved large changes of energy-entropy, such as underground nuclear explosions, particle accelerators, or rocket launches. In several instances, AC was outstanding when targets underwent massive changes in energy or entropy in a very short period of time during the session. Bem and Honorton's finding is also suggestive that an entropic change in the target might lead to better results. As a possible explanation for these observations, consider that AC may be mediated through a specialized sensorial system and that this system might behave similarly to the five known sensorial systems. We might reasonably expect, then, that AC would correlate positively with the changes of the sensor-input signal and correlate less well with the level of the sensor-input signal itself. In vision, for example, the system is sensitive to changes in brightness across the field, but relatively insensitive to the absolute level of illumination. Analogously, we hypothesized that the AC system might be sensitive to changes in the level of information content across a target, but insensitive to the absolute level of that measure. In the first experiment, May, Spotiswoode, and James (1994a) found a significant correlation between the gradient of the Shannon entropy of the target and the quality of AC (Spearman rank-order correlation coefficient, r,, of .452, df = 26, t = 2.58, p = 7.0 (10) for static photographic targets. Unfortunately, with dynamic targets, there was little evidence of AC and a resulting small correlation with the gradient of the entropy. In the second experiment, they found strong evidence for AC in both static and dynamic targets and for the two target types combined, the correlation between AC performance and entropy gra- AC Sensory System 55 dient r, was .337, df = 31, t = 1.99, p = .028. As predicted, the correlation with the entropy itself was considerably smaller (r, = .234, t = 1.34, d = 31, p = f .095). The correlation for the combined static targets from both studies was r, = .16l, d = 41, p = 152. Because of the different target systems and protocols f in these studies, the results remain somewhat ambiguous. This report provides a detailed description of an experiment to replicate May, Spotiswoode, and James (1994a) entropy findings. Hypotheses The primary hypotheses were: A significant correlation exists between the quality of AC, as measured by a fuzzy-set technique, and the gradient of Shannon entropy of its associated target. The correlation of the gradient with the quality of AC as measured by the upper half of the rating scale shown below in Figure 2 will be consistent with the static target correlation seen in the earlier experiments (May, Spotiswoode, and James, 1994a). The concept of fuzzy sets was first applied to the analysis of AC data by May et al. (1990). A fuzzy-set definition of a target is similar to the commonly used descriptor lists in which an analyst is asked to ascribe the presence or absence of each element in a list of items. Instead of a forced yes or no to the presence of an element, such as water, a fuzzy approach allows for a quantitative coding of a subjective impression. For example, water might be 30% visually impacting in a target and therefore is coded as 0.3 rather than either 1 or 0. A response is coded in a similar way. Three quantities are defined from the fuzzy-set representation of a target and a response. The accuracy is defined as the percent of the target that was described correctly; the reliability is defined as the percent of the response that was correct; and the figure of merit is defined as the product of the two. Because the fuzzy-set measure is less granular than a rating measure, being a rational rather than an ordinal scale, we chose it as the primary measure. The rating scale correlation was included as a historical link to our earlier experiments. A more detailed description of the technique can be found in the AC Data Analysis Section, below. We also hypothesized that the correlation of the figure of merit with the total entropy of the target would be much less than the correlation with the gradient of the entropy. Experiment Protocol In contrast with the majority of our earlier AC studies, we designed a protocol in which the receivers were physically located from 5 to 4,500 km from the 56 E. May, J. Spottiswoode, & L. Faith laboratory. In addition, many aspects of the experiment were handled automatically by two separate computers. Target Pool Construction For this experiment, we developed a completely new target pool based exclusively on the Corel Stock Photo Library of Professional Photographs. This library of copyright-free images is provided in digital form and comprises 100 images on each of 200 CD-ROMs. Each image is approximately 18 MB in size, which corresponds to a landscape format picture of 3200 x 1875 pixels in 24-bit color. Corel also publishes a booklet of thumbnail images of the complete set. Selection Criteria The first stage in constructing the target pool consisted of creating a design specification of the type of photographs that would qualify as a potential AC target. Based on earlier experience (May et al., 1990), we adopted a series of a guidelines. First, the photographs had to possess common properties. Thematic coherence: Each photograph had to be a real scene, as opposed to a collage. Where possible, the photographs also possessed elements that could be sketched easily. Size homogeneity: The photographs did not contain any surprises with regard to size. For example, there would not be a photograph of a brick followed by one of a mountain range. Pool coherence: All of the photographs would depict outdoor scenes. The following elements would not be included in the pool by construction or by photographic editing: People Transportation devices (e.g., boats, cars, etc.) Small human artifacts (e.g., tools, toys, etc.) We made every effort to remove these kinds of items, although they may have been present in some photographs. If so, they were difficult to see and were insignificant relative to the rest of the scene. Finally, we would not allow odd camera angles, unusual or distorted perspectives, or odd or unusual lighting conditions. Aside from the above restrictions, the target pool photographs could show any scene at any location. Following these guidelines, we rejected approximately half the original set of 20,000 photographs by visual inspection of the thumbnail images. Our long-standing earlier target pool consisted of 100 photographs divided into 20 packets of five dissimilar images each. In that pool, a target for a trial was determined by first choosing a random integer between one and 20 to se- AC Sensory System 57 lect a packet and then choosing a random integer between one and five to select a target. The remaining four targets within the selected pack then served as decoys for a blind analytical assessment by rank ordering. For the development of this new pool, we chose a different approach. Namely, the analysis decoy target images would be determined after the AC trial was complete. To assure that we could do this in a blind and algorithmic fashion, we adopted a hierarchical design of groups, categories, and images. A group consisted of five categories, and each category contained five images. The images within a category would be as much alike each other as possible, although they must be of different scenes. Differing perspectives of the same scene were not included. Thus, a single category of "waterfalls," for example, would contain five similar, but different, waterfalls. In contrast, we made every attempt to choose categories within a group to be as different from one another as possible, to make them orthogonal in other words. For example, we would not have a "river" category in the same group as a "waterfall" category. The number of different groups was determined by the remaining 10,000 images that survived the first cut. Two laboratory personnel examined all 10,000 images on a high-resolution computer display, and approximately 800 candidate photographs met the above acceptance criteria. After some digital editing, we identified from this set of 800 photographs 12 groups of 25 images for a total of 300 targets. Table 1 shows the categories that were identified for each of the 12 target groups. No attempt was made to force the categories to be orthogonal across groups. Figure 1 shows an example of the digital editing of an image that was not selected as part of the pool to illustrate the capability to modify an image to conform to the construction guidelines. In the temple scene, nearly all the people were removed by making reasonable guesses as to what the image would have TABLE 1 Categories for Each Target Group Category Group ID 1 2 3 4 5 6 7 8 9 10 11 12 1 Bridges Bridges Bridges Bridges Bridges Fields Cities Coasts Buildings Buildings Fields Coasts 2 Canyons Cities Lakes Mosques Churches Islands Coasts Fields Coasts Coasts Structures Mountains 3 4 5 Waterfalls Structures Towns Waterfalls Pyramids Waterfalls Windmills Rivers Waterfalls Rocks Streets Towns 1 Cities Fields Mountains Mountains Deserts Roads Deserts Lighthouses Pyramids Fences Rivers Roads Structures Mountains Structures Roads Mountains Ruins Waterfalls Mountains Vineyards Lakes Ruins Ruins Note: All Structures in the table represent Asian structures 58 E. May, J. Spottiswoode, & L. Faith Fig. 1. An example of digital editing. AC Sensory System TABLE 2 Universal Set of Elements (USE) Buildings Villages, towns, cities Ruins Roads Pyramids Windmills Lighthouses Bridges Coliseums Hills, cliffs, valleys Mountains Land-water interface Lakes, ponds Rivers, streams Coastlines Waterfalls Glaciers, ice, snow Vegetation Deserts Natural Manmade Prominent, central Textured Repeat Motif been behind each individual. As the final step in preparing an image for the target pool, the picture was cropped, if necessary, and resized to 800 x 600 pixels, each having 24 bits of color information. Fuzzy -Set Encoding To facilitate subsequent computer analysis of AC trials, the images were encoded using a system of descriptive elements. Each element was assigned a fuzzy-set membership value for each image. We created a universal set of elements (USE), including 50 elements that we selected from the original set of 131 elements used in our earlier work (May et al., 1990). We also added elements for features that were unique to this particular set of photographs. Six individuals each coded all 300 images against this USE. As in earlier work, the coding criterion was the degree to which each element was visually impacting to the general scene. The range of visual impact ran from 0 to 1 in steps of 0.1. For example, in the bottom image in Figure 1, we might code 0.6 for "buildings" and 0.3 for "repeat motif." The principal investigator selected 24 elements out of the 50 and qualitatively condensed the scorings from the six coders to a single "consen~us'~ fuzzy-set representation of the targets. These 24 elements were selected on the basis of extensive experience, as well as on the formal analysis of a single study. The principal criterion used in the selection was that the elements should not be too "low level" such as lines and geometric shapes, nor should they be too "high level" such as and office building. These 24 elements were an attempt to strike a compromise between these two extremes. Table 2 shows the 24 elements that comprised the final fuzzy-set USE.' Receiver Selection Five experienced receivers participated in this experiment. They were chosen on the basis of their availability, their willingness to participate in a 'A detailed description of the target pool construction and its associated fuzzy-set encoding is currently being considered as a separate paper. I 60 E. May, J. Spottiswoode, & L. Faith lengthy AC study, and especially on their previous and sustained superior performance. Number of Trials The total number of trials for this study was 75 (i.e., 15 for each receiver) and was determined, in advance, by receivers' availability and statistical power considerations. We used the average effect size of 0.550 from a previous similar experiment (Lantz, Luke, and May, 1994) to compute a statistical power of 68% to reach significance (i.e., p = .05) for a single receiver and a power of 99% to reach a significant study. Trial Protocol We designate Experimenter 1 and Experimenter 2 as E l and E2, respectively. E l was located in the laboratory in Palo Alto, California, and E2 was located in a laboratory in Los Angeles. The complete target pool was independently installed on E l and E2's computers. Note that all communication between E l and E2 occurred only by e-mail. At a prearranged scheduled time, the following events took place in the order shown: E l requested that E2 generate a target for the upcoming trial. E2 invoked a computer program that first randomly selected one of the 12 groups, randomly selected one of the five available categories in that group, and then randomly selected a target image from within that category of five images. The program saved its choice to a binary file and did not notify E2 about any aspect of the selection. E2 notified E l that the selection process was complete. E l telephoned the scheduled receiver and acted as a monitor for an AC session lasting from 5 to 15 minutes. The receiver drew and wrote the impressions and faxed them to E l at the end of the session. E l requested that E2 generate a decoy set. E2 invoked a second computer program that read the binary file containing the target information and randomly selected a target image from each of the four remaining categories from within the selected group. The four decoy target numbers and the intended target number were randomly ordered and then automatically e-mailed to E l . E l analyzed the session and e-mailed the results to E2. At this point, nobody was aware of the selected target, and the analysis was complete before the receiver obtained feedback. E2 invoked a third program that read the original binary file and e-mailed the actual target number to E l . AC Sensory System 61 E l posted the target photograph on a Web-site to which only the receiver had access and then telephoned the receiver to provide verbal feedback and to prompt the receiver to access the Web-site for visual feedback. All transactions were logged, and session and analysis details were automatically stored in a database. Typically, such a trial would be complete in 30-60 minutes. Furthermore, in contrast to our earlier studies, the analysis was completed on each trial before anyone was aware of the intended target. AC Data Analysis We had decided to perform three separate analyses on the AC data. The first of these was a standard rank ordering of the target pack, which consisted of four decoys and the intended target. E l was presented with the words and drawings along with the target pack associated with the trial, and the task was to rank order the targets from the best to worst matches to the response. After all N trials were analyzed for a single receiver, a continuity-corrected effect size omputed as: where R,, is the average rank over the N trials, and the last term in the numerator is a continuity correction for small N. The z score associated with this effect size is given by: The rank-order analysis was designated, in advance, as the primary indicator of AC in the study. Because, the primary goal of the experiment was to explore the relationship between the gradient of Shannon entropy and the quality of the remote viewing, we performed two additional analyses. Lantz, Luke, and May (1994) showed that assessing AC performance by rank ordering was not optimal for correlation studies for two reasons. First, the rank number is strongly dependent on the degree to which the photographs in the analysis pack are different from one another. Second, the ranking method discards information about the absolute quality of the match; it only describes the relative closeness of the match in comparison to the decoys. Consequently, a perfect match between a response and a target would be assigned the same first-place rank as a response that corresponded far less closely but nonetheless was sufficient to allow the analyst to assign a first-palace match. For historical reasons and for comparison with earlier entropy experiments, we used a slightly modified version of the 0 to 7 rating scale. Figure 2 shows a screen capture image of the scale that was presented to E l during the analysis. To be assigned a given assessment value, the correspondence between target and response must meet one of the criteria shown in Figure 2. As before (May, 62 E. May, J. Spottiswoode, & L. Faith Fig. 2. The 0 to 7 Assessment Scale. Spotiswoode, and James, 1994a), the scale was divided into two sections; an assessment of four and above indicating possible AC contact with the target and three and below indicating no contact. We recognized a number of difficulties with the rating scale. The assessment values are granular, that is, they are integers with no possibility of values in between. More important, the scale does not account for the amount of material in the response. For example, the response could be simply the word "city" and receive a value of 7 for the match to a city target, even though there might be many elements, such as a river, a bridge, and mountain background, in addition to the city in the target. Because of its extensive previous use, this rating scale was included in the analysis but defined only as a secondary measure to be used in the entropy correlation analysis. The primary measure of the absolute correspondence of a response to its intended target was the fuzzy-set based figure of merit; whereas the rank-order statistic was used as the primary measure for overall AC. May et al. (1990) provided a partial solution to the problem associated with the rating scale through the use of fuzzy sets. We used a fuzzy-set measure for an assessment of the degree of correspondence between a response and a target. We defined the figure of merit (FM) as the accuracy times the reliability. The accuracy is the percentage of the target image elements that were described correctly, and reliability is the percentage of the response elements that were correct. Although neither accuracy nor reliability alone is a suffi- AC Sensory System 63 cient measure of AC, the product of the two is. Formally, the accuracy is defined by: N j=l N accuracy = , (0 5 accuracy 5 1) where N is the number of elements in the USE. Similarly, reliability is defined by: C min(l;, Rj) reliability = j=1 N , (0 5 reliability 5 1) where Tj and Rj are the fuzzy-set membership values for the target and response, respectively. Min(Tj Rj) means the minimum of the two quantities. The figure of merit (FM) is the accuracy x reliability. The fuzzy-set analysis for each trial occurred as follows. After the response was received by fax and while blind to the target, E l scored each element in Table 2 as to the degree to which that element was contained in the response. If the response contained the word "waterfall," then by definition, the waterfall element would receive a score of one. If, however, there was a vague sketch that might look slightly like a waterfall, then that element might only be scored as 0.3. Thus the entire USE was scored before E l was shown the analysis target pack. E l then displayed five targets for the trial, performed the rank ordering, the 0- to 7-point scale assessment, and finally entered the two target-dependent fuzzy-set elements. All results were inserted into a database for subsequent analysis. To summarize, the fuzzy-set elements for the targets were assigned, before the experiment, to represent the degree to which each element was visually impacting in the scene. The response elements were scored as to the degree to which the element was contained in the response. At this time, we have little evidence that a receiver is capable of not only recognizing that an element is in a target but also capable of determining its visual impact. For example, we rarely received a statement such as, "There is a river in the target but it is hardly noticeable." Thus, the target fuzzy-set encoding contained more information than could be obtained easily with AC. At this stage of our understanding of AC, we must be content with a simple recognition on the part of the receiver as to the presence or absence of a particular element. Therefore, before the calculation of the accuracy, reliability, and FM, we converted the target fuzzy set to a crisp set, containing only 1 for presence and 0 for absence for the membership values of the elements in the USE. 64 E. May, J. Spottiswoode, & L. Faith This process is called an alpha cut in fuzzy-set parlance. That is, we specify a threshold for the fuzzy-set membership value so that if an element is equal to or above that threshold, it is converted to a 1 or set to 0, otherwise. We adopted the threshold value of 0.2 to remove some of the noise "clutter" of 0.1-encoded elements. This value was empirically determined as a reasonable value (May et al., 1990). An alpha cut was not applied to the response because the fuzzy element represented the degree to which the analyst felt that the given element was represented in the response. Finally, we added two additional elements, which were independently scored for each target in the analysis pack, to the USE shown in Table 2. The element "visual" was an assessment of the degree to which the drawings, independent of the labels or other written material, matched a target image. The element "analytic" was the degree to which the written material, independent of the drawings, matched a target image. By definition, these elements were scored as 1 for all targets and were added to the consensus-scored fuzzy-set representation of the targets. Thus, in the equations for accuracy and reliability, N is equal to 26, with 24 elements coming from Table 2 and two coming from these additional elements. Entropy Analysis An entropic analysis of a photographic image is an assessment only of intensity patterns and does not include any cognitive information. In this context, the gradient means transitions between light and dark regions. The details of how such an analysis is conducted can be found in May, Spotiswoode, and James (1994a). We will, however, summarize the approach here. The Shannon entropy for a single color plane with a depth of eight bits is given by: where pi if the probability of observing an intensity value of j. The following discussion holds separately for each of the three color planes. The total entropy is the sum of the three color entropies. We computed this entropy for all targets in the following way: Each image was divided into rn x n patches where we constrained the patch size to be evenly divisible into 800 and 600, the standardized target size. The patch sizes chosen were 4, 8, 20, 40, and 100 pixels square. For a given size, we computed the entropy for each patch across the photograph. For example, using a patch size of 20, we would compute the entropy for each of the 40 x 30 different patches. The pis were determined by the empirical values contained in each patch. Finally, using standard numerical techniques, we computed the average absolute magnitude of the gadient2 in this 2-dimensional entropy 2 ~ hgradient is a formal measure of the "steepness" of the "hills" and "valleys" in the entropy space. e AC Sensory System 65 Fig. 3.1,ow and high entropy gradlent images (top). Entropy per patch for two images (bottom). space. Figure 3 shows images that have low entropy gradient, such as a pyramid and those that have high entropy gradient, such as a bridge. Both entropy plots have the same vertical scaling of 20 bits.' The steeper gradient of the "hills" and "valleys ' in the bridge plot results in that image having 365% greater entropy gradient than the pyramid irnage has. The entropy gradient calculation was performed for all 01 the patch sizes shown above for all targets. Additionally, we calculated what we call the total entropy in which we computed a single valile for the entire picture. That is, the single patch size was 800 x 600. All results were stored in the database for later analysis. 7 Correlation Analysi~ To closely approximate the patch size that was used in our earlier studies (May, Spotiswoode, and James, 1994a), we adopted a patch size of'20 x 20 as 'The rnax~nrl~l~n entropy 15 24 brts, given that thc \urn I, over all three 8-btt color plane\. 66 E. May, J. Spottiswoode, & L. Faith the primary value for the correlation calculations. We did, however, examine any effects as a function of patch size. Similarly, we divided the assessment scale in half and used only the upper half for the correlation calculations, but we examined any effects as a function of scale division. For the correlation of gradient and entropy with the rating scale, we used the conservative, nonparametric Spearman's r method and converted the observed correlation to a standard normal deviate with Fischer's Z transform. The FM values are more nearly continuous as a consequence of its algorithm. Nonetheless, even in this case, we used the more conservative Spearman's r to compute the correlation. Results The results fall into the two categories of evidence for AC and correlation effects. All p values are quoted as single-tailed. AC Results Table 3 shows the average rank, continuity-corrected effect size, and associatedp value for the five participants' 15 trials. The results, using the rank-order statistic, illustrated in Table 3, show no AC in this study either for individual receivers or overall. The effect size falls below what we have come to expect from this group of receivers. We shall return to this point in the Discussion section. Note that the effect size for the total is not the average of the effect sizes for the individual receivers. This is because, taken as a study with 75 trials, the continuity correction is different. Entropy Results For our primary hypothesis, which requires a correlation test with the figure of merit, we find a Spearman's r for the average magnitude of the gradient of Shannon entropy correlated with the figure of merit of 0.212 with 73 degrees of freedom. This corresponds to Z = 1.83, p = .034. Figure 4 shows the scatter diagram for the gradient versus the figure of merit. Although the sum-of-rank statistic did not show significant evidence for AC, the primary hypothesis was confirmed. That is, the quality of the AC as TABLE 3 AC Results Receiver - Average Rank Effect Size p value 8 127 221 497 937 Totals AC Sensory System 67 Fig. 4. Correlation of FM with entropy and its gradient. measured by the figure of merit significantly correlated with the gradient of Shannon entropy for a patch size of 20. The points for the gradient are shown as A, and the regression line is shown as the solid line in Figure 4. Next, we examined the correlation of the gradient of Shannon entropy with the assessment scale with values greater than three. With a patch size of 20, this correlation most closely replicates the earlier work. The second hypothesis was also confirmed. The combined static target results from the previous two studies produced a strong correlation (r, = .161, df = 41, p = .152). In this study, as measured by the upper half of the rating scale as shown in Figure 2, the gradient of Shannon entropy did correlate with the quality of AC ( r , = .146, df = 23, p = .246) at nearly the same level as before. Finally the secondary hypothesis was confirmed as well. The correlation between the total entropy and the figure of merit was small ( r , = .042, df = 73, p = .362). The points for the entropy are shown as x in Figure 4, and the regression line is shown as dashed. Discussion AC Result Utts (1995) has shown that our experienced receivers exhibit a consistency of AC effect size over time, a result consistent with our own observations. Be- 68 E. May, J. Spottiswoode, & L. Faith cause of this, we have come to expect significant results when, as in this study, we have sufficient statistical power to observe a significant result. When a study fails to exhibit significant evidence for AC, we are usually suspicious that some aspect of the protocol was responsible for the decrease in study effect size. There were several protocol differences between this study and our usual method: a new target pool was used, the primary analysis was completed before feedback was given, and the subjects were physically remote from the lab. We do not expect that the first two points are a factor, but the last probably is. This is not to suggest that distance between target and receiver is a modulating variable. Rather, we suggest that it is a matter of attention to the task. In the past, we have invited our receivers to the laboratory for their sessions. Many times, this involved flying them across the United States for a weeklong visit on two or three separate occasions. In these cases, when a receiver was present in the laboratory, they had our full attention for the trial. All activity in the laboratory was focused on that single trial. In this study, the monitor called each receiver at a prearranged time and conducted a short session by telephone. The trials therefore amounted to relatively brief interludes in the otherwise busy schedules of both the receivers and the experimenters. These are psychological conditions unlike the intense focus during trials in our earlier studies. There are numerous laboratory anecdotes about excellent AC performance under high attention. The "Put to the Test" AC trial that was shown on national U.S. television is just one example.4 In this example, approximately 10 people had their full attention on a trial that cost an estimated $100,000; the result was near perfect correspondence between responses and targets. These kinds of arguments can only be speculation, of course. One of the benefits of working with the same receivers over a protracted period of time is that our observed individual performance consistency allows such speculation. In this case, all of the receivers have been participating in experiments of this nature for more than 15 years. Further studies in which the receivers are in the laboratory will test this possible explanation. Entropy Result Figure of merit assessment of the quality of the AC. As shown in Figure 4, we observed a significant correlation between the gradient of Shannon entropy and the quality of the AC as measured by the figure of merit. This correlation, however, was observed at a patch size of 20 x 20 pixels. A question arises about possible dependency of the correlation on patch size. Table 4 shows the patch size, Spearman's r (df = 73), its associated Z score, and p value tested against r, = .O. Except for patch sizes of 40 and 100, we see a consistent correlation as a 4~~~~ Productions, Sherman Oaks, CA; November 28,1995. AC Sensory System TABLE 4 Patch-Size Dependence Patch Spearman's r Z Score p Value function of the patch size. Perhaps the decrease for the larger patches is because the details of the intensity features are lost as they become an increasing fraction of the picture. For example, consider an 800 x 600 pixel image. The patch of size of 40 and 100 correspond to 0.3% and 2.1%, respectively, of the total area. These numbers intimately depend on the details of the target pictures in the study and do not generalize. A more important consideration, however, is to determine what other circumstances might induce an apparent correlation between the entropy gradient and the figure of merit. Because the targets were chosen randomly, the probability of matching a given response to the intended target is 20%, regardless of response bias on the part of the receiver or judging bias on the part of the analyst. In particular, analyst bias cannot systematically affect the figure of merit values because of this blind assessment. Thus, a number of potential artifacts are eliminated because of the differential match and the random selection of the target. Nonetheless, it might be that there is some variable that correlates independently both with the gradient of the entropy and the figure of merit. One such candidate is the cognitive complexity of the target. If the gradient of the entropy correlated significantly with some measure of cognitive complexity, and the figure of merit did so as well, then the observed correlation of the gradient of the entropy with the figure of merit would contain an artifact. As May et al. (1990) showed, a reasonable estimate for the cognitive complexity is the fuzzy-set sigma count for each target. The sigma count is simply the sum of the membership elements in the fuzzy-set representation of the target. The USE as shown in Table 2, represents high-level cognitive elements, whereas the USE that has been used in the past contained a large number of nonobject features such as ambiance, color, and low-level linear features. May, Spottiswoode, and James (1994a) reported a small and nonsignificant correlation of target sigma count and the gradient (r, = -.028, df = 98, p = .609). In our current USE, however, the elements are all features that might contain significant intensity patterns and thus might show an overall correlation of gradient with sigma count. As expected, therefore, for all 300 targets in the pool, we observed a significant correlation between the gradient of Shannon entropy, computed for a 70 E. May, J. Spottiswoode, & L. Faith For patch size of 20, and the sigma count (r, = .199, df = 297, p = 2.59 x the 75 targets that were selected in the study, the correlation is larger (r, = .359, df = 72, p = 7.10 x The correlation of the sigma count with the figure of merit is small, however (r, = .0017, df = 72, p = .494). To determine the impact of these two correlations on the correlation of the gradient with the figure of merit, we consider a general case. Suppose that there is a significant correlation between variables X and Y, r(X, Y). Suppose further that X and Y both independently correlate with a third variable, 2. We must determine the conditions for the magnitude of these independent correlations such that the observed r(X, Y ) would be an artifact. Assume that the r(Y, Z ) is unity (i.e., completely correlated). We then can replace Z with Y and consider r(X, Z ) as r(X, Y). In this case, r(X, Y) is completely determined by r(X, 2).If r(Y, Z ) is less than unity, than the contribution to r(X, Y)from the independent correlation with Z will be smaller than in it is the unity case. In our case, the correlation of the figure of merit with the sigma count is r, = .OO17 and the correlation of the gradient with the sigma count is r, = .217 < 1. The contribution to the observed correlation of the gradient with the figure of merit for this potential artifact is therefore less then .OO17. Rating assessment of the quality of the AC. For historical and replication reasons, we examined the correlation of the gradient with the upper half of the blind rating scale shown in Figure 2. The Spearman's r was .I46 (df = 23, p = .246), which was consistent with the combined correlation of r, = .161, df = 41, p = .I52 for the static targets in the earlier two studies (May, Spotiswoode, and James, 1994a). Conclusions The primary and secondary hypotheses were confirmed. That is, the gradient of Shannon entropy of the target appeared to correlate with the quality of AC, whereas the quality did not correlate with the entropy itself-a result that is suggestive of a sensory system. We legitimately might ask how it is possible to see no AC in the study as defined by the accepted rank-order technique yet see a significant correlation with the gradient of the entropy. One way to understand this apparent contradiction is to examine closely the underlying assumptions of the two AC measurements that are involved, rank order and figure of merit. It is clear that the rank-order technique is a relative measure, which is strongly dependent on the orthogonality of the set of photographs in a judging pack. As an example, let us assume that the target is a small cabin next to a stream in the woods and that a minimal response includes flowing water but does not include the cabin or the woods. In the best case scenario, suppose the pack orthogonality was such that only one picture contained any water at all. In this case, with a modest amount of AC Sensory System 71 AC, an analyst would have no trouble making a first-place match. In the worst case scenario, suppose the pack contained all five pictures with flowing water of various types, but only one contained a cabin in the woods as well. In this case, the analyst would, on average, end up with a third-place match. Thus, we can see that the rank statistic for the same response strongly depends on the photographs in the judging pack. In the figure of merit analysis of this same example, it might be that the scores for all the photographs in the worst case scenario might be identical, say 0.15, because the response matches each target with about the same level of small correspondence. But in the best case scenario, the figure of merit analysis will give the same score for the stream-cabin-in-the-woods target (i.e., 0.15), but all the decoy targets will be lower. The point is that the figure of merit for the intended target is independent from the content of the judging pack. Many trials in the best case scenario would likely yield a significant rankorder statistic, whereas in the worst case, the rank order is exactly at chance. For the small amount of AC that was assumed in the example, one might come to each of these conclusions, depending on the judging pack orthogonality. In our case, there is no question that the AC is far below what we have come to expect from our established receivers. Second, by the nature of our target pool bandwidth (May, Spotiswoode, and James, 1994b), it is difficult to assure strong orthogonality. In the past, these types of targets have done well when the AC functioning is strong; however, when the functioning is weak, as in this experiment, a rough threshold of AC is needed to produce a significant rank-order statistic. As we have illustrated from the example, it is unlikely that the threshold is zero. That is, small amounts of AC might still produce a rank-order statistic near chance. A correlation is algebraically not sensitive to the absolute level of one or both of its variables. We could add a large constant to either the gradient or the figure of merit and would find exactly the same correlation. Therefore, we believe that we have replicated the earlier finding in which the quality of AC is correlated with the gradient of Shannon entropy and not with the entropy itself. This result is part of the growing and compelling evidence that AC is mediated through a sensory channel. This might be either some combination of the known senses or an additional one. Functional brain imaging may resolve this question by allowing us to directly observe neural functioning during anomalous cognition. Acknowledgements We thank Maggie Blackman, Bob Bourgeois, Nicola Kerr, and Lisa Woods of the Rhine Research Center for their heroic and tireless effort at coding the 50 fuzzy elements for each of the 300 targets. We also thank Laura Faith at the Cognitive Sciences Laboratory for her significant contribution in the selection of the target pool and for her coding expertise. Finally, we are deeply apprecia- 72 E. May, J. Spottiswoode, & L. Faith tive because this work would not have been possible without the generous support of the Fundagiio Bial of Porto, Portugal. References Bem, D. J., & Honorton, C. (1994). Does psi exist? Replicable evidence for an anomalous process of information transfer. Psychological Bulletin, 115, 4-18. Lantz, N. D., Luke, W. L. W., & May, E. C. (1994). Target and sender dependencies in anomalous cognition experiments. The Journal of Parapsychology, 58, 285-302. May, E. C., Spottiswoode, S. J. P, & James, C. L. (1994a). Shannon entropy: A possible intrinsic target property. Journal of Parapsychology, 58, 384-401. May, E. C., Spottiswoode, S. J. P, & James, C. L. (1994b). Managing the target-pool bandwidth: Possible noise reduction for anomalous cognition experiments. Journal of Parapsychology, 58,303-313. May, E. C., Utts, J. M., Humphrey, B. S., Luke, W. L. W., Frivold, T. J., & Trask, V. V. (1990). Advances in remote-viewing analysis. Journal of Parapsychology, 54, 193-228. Utts, J. (1995). In Mumford, M. D., Rose, A. M., & Goslin, D. A. (Eds.), An evaluation of remote viewing research and applications. The American Institute for Research. Report, September 29. Journal of Scient$c Exploration, Vol. 14, No. 1, pp. 73-89,2000 0892-33 1 O/OO O 2000 Society for Scientific Exploration Contributions to Variance in REG Experiments: ANOVA Models and Specialized Subsidiary Analyses Princeton Engineering Anomalies Research School of Engineering/Applied Science Princeton University, Princeton NJ 08544 Abstract- Judicious application of a complementary set of sophisticated analytic techniques to large databases from humanlmachine anomalous interaction experiments can extract subtle structural features that might elude more simplistic analyses. The combination of a multi-factor analysis of variance (ANOVA) with various subsidiary, a d hoc approaches suggested by the ANOVA or directly by the data, can establish an instructive hierarchy of salient physical and subjective parameters and illuminate some of their specific details. In this particular study, the dominant finding is a significant correlation of anomalous effects with prescribed intentions of the human operators, compounded of small contributions from many individuals across many experimental conditions. The grand concatenation, which includes all combinations of successful and unsuccessful parameters or conditions, shows a chance probability for this correlation with intention on the order of The effect apparently is confined to non-deterministic devices; i.e., deterministic pseudo-random sources show no overall effect. The correlation with intention for non-deterministic sources alone has a chance probability of 106 . Beyond operator intention, most of the other technical, procedural, and subjective parameters explored show unimpressive contributions to the overall variance, with a few notable exceptions that are clarified in the subsidiary analyses. For example, individual differences among operators are indicated, but there is a relatively normal distribution of effect sizes, within which a few participants are distinguished by consistent achievement over large databases. The temporal development of effect sizes shows a consistent pattern of initial success that declines but then recovers. There is essentially no evidence for a dependence of effect size on spatial or temporal separation, supporting other indications that ordinary physical variables have little impact on the anomalous interactions. In sum, although the composite ANOVA models explain less than 1% of total variance, implying very small and subtle effects, the analysis provides strong evidence that the anomalies are statistically robust; they are not due to chance fluctuations, but are demonstrably correlated with definable subjective factors. Keywords: anomalies - ANOVA - consciousness - electronic random event generator - mindlmachine interactions - models -REG - RNG Introduction Beginning about four decades ago, electronic random event generators or random number generators (REG or RNG) have been used in a wide range of 74 Nelson et al. laboratory experiments designed to test the hypothesis that human consciousness might interact directly with labile physical systems. The results provide clear statistical evidence that the behavior of these devices deviates from chance expectation in correlation with the pre-defined intentions of participants in the experiments (Radin and Nelson, 1989). In 1979, the Princeton Engineering Anomalies Research Laboratory (PEAR) began collecting large databases in REG experiments using particularly rigorous controls and a variety of optional parameters to assess the character and replicability of such anomalous mindlmachine interactions. Over a 12-year period of primary investigation, ten physical and psychological conditions were examined as possible mediating variables in the experimental results. A number of extensions and variations of the basic protocol were explored, using different REG sources as well as a selection of other physical systems, the performance of which was dependent on some form of random process. The experiments had accompanying calibrations which confirmed that the random sources were of high quality, producing data that consistently conformed to theoretical expectations in non-experimental conditions. In the active experiments, however, the data sequences and distributions were significantly correlated with the experimental variables, especially the operators' intentions to shift the means of the REG output distributions, and showed structure that could not be accounted for by chance fluctuations. This paper presents a compact summary of a formal analysis of variance (ANOVA), and a number of subsidiary ad hoc analyses of the results from these REG studies, comprising 1338 replications of the experiment, including a small set of variations on the basic protocol. Equipment The PEAR program has used three generations of random event generators, utilizing different primary sources of white noise but maintaining important common features of design. An original benchmark experiment employed a commercial random source sold by Elgenco, Inc., the core of which is proprietary. Elgenco's engineering staff describe this module as solid state junctions with precision preamplifiers, implying processes that rely on quantum tunneling to produce an unpredictable, broad-spectrum white noise in the form of low-amplitude voltage fluctuations (Nelson, Bradish, and Dobyns, 1989). A much simpler and more compact portable REG was based on Johnson noise in resistors, or so-called thermal noise, which also is a quantum-level phenomenon that produces a well-behaved, broad-spectrum fluctuation (Nelson, Bradish, and Dobyns, 1992). A later-generation device called the PEAR Micro-REG used a field effect transistor (FET) for the primary noise source, again relying on quantum tunneling to provide uncorrelated fundamental events that compound to an unpredictable voltage fluctuation. In all cases, the design begins with some white-noise frequency distribution. For example, the benchmark REG presents a flat spectrum + 1 db from Variance in REG Experiments: ANOVA Models 75 50 Hz to 20 kHz. A low-end cutoff at 1000 Hz attenuates frequencies at and below the data-sampling rate. This filtering, followed by appropriate amplification and clipping, produces an approximately rectangular wave train with unpredictable temporal spacing. Gated sampling, typically at a constant 1kHz rate, yields a regularly spaced sequence of random bits, suitable for rapid counting. Other sources have been constructed that allow higher sampling rates, up to 2 MHz (Ibison, 1998), but this paper summarizes data from the standard unit only. Analog and digital processes are isolated by temporally alternating these operations to exclude contamination of the analog noise train by the digital pulses. To eliminate biases of the mean that might arise from such environmental stresses as temperature change or component aging, an exclusive or (XOR) mask is applied to the digital data stream. This is either a regularly alternating 110 pattern or a more complex mask comprising a randomly ordered array of all 8-bit bytes with equal occurrence of 110. The latter procedure also excludes all short-lag bit-to-bit and byte-to-byte auto-correlations. Finally, data for the experiments are presented and recorded in "trials" that are the sum of N samples (typically 200 bits) from the primary sequence, thus further mitigating any residual short-lag auto-correlations. The final output of the benchmark REG thus is a sequence of conditioned bits, and in the later devices, of bytes presented to the computer's serial port, which then are formed into a sequence of trials, usually generated at approximately 1 per second. Calibrations on all of the devices closely conform to statistical expectations for the mean, variance, skew, and kurtosis of the accumulated count distributions, and expectations for time-series of independent events (Nelson, Bradish and Dobyns, 1989; Nelson, 1993; Nelson et al., 1997). Experimental Design The basic experimental designs also embody protocol-level protection against artifactual sources of apparent effect. Following a "tripolar" protocol, participants generate data under three conditions of pre-specified intention, namely to achieve high (HI) or low (LO) mean values, or to generate baseline (BL) data. With the exception of the intention held in the participant's mind, which is pre-recorded in computer files, these three conditions are otherwise the same; all potentially influential variables are maintained constant within an experimental session or series. In addition to the primary variable of tripolar intention, a number of secondary parameters are available as options that can be explored in separate sessions, and assessed as factors that may contribute to the experimental outcomes. These include: 1 . Human variables such as the identity of the individual operators (participants), their gender, and whether they are "prolific," i.e., have performed sufficient replications of the experiment to permit robust comparisons. In this category we also include the replication number or , 76 Nelson et al. serial position of the session as a factor that reflects operator experience. 2. Physical variables such as the different noise sources, including not only the true random sources described earlier, but also various hardware and algorithmic pseudo-random generators, designated as non-deterministic and deterministic sources, respectively. 3. Operational variables, including spatial separation of the operator from the machine (up to thousands of miles) and separations in time (up to several hours or in a few cases, one to four days); information density (bits per second); the number of trials in automatically sequenced "runs;" the instruction mode (volitional or instructed); and the type of feedback provided to the operator. Analysis of Variance The benchmark REG database was accumulated over a period of 12 years, with contributions from 108 individual operators, 30 of whom met our criterion for prolific operators by generating a minimum of 10,000 trials per condition (the equivalent of 10 experimental series or replications). The primary database of 1262 independent replications, comprising a total of 5.6 million 200-bit trials, was analyzed by a regression-based ANOVA, previously reported (Nelson et al., 199 1 ) . Since this phase was completed, two smaller data sets have been added, for a total of 1338 replications, and new versions of two factors (the random source variable is simplified; operator identification is now enhanced with gender specification) included in the original analysis have been defined to help understand the results of the original analysis and those of other special-purpose, detailed analyses. Although they are defined a posteriori, they provide legitimate assessments of the questions they represent, within the context of the full complex of potential influences in this experiment. The discussions that follow thus rely on the original analysis, supplemented by additional information derived from the expanded assessment using the new factor definitions. The REG database is complex, involving nine analytical factors with two to five levels, and a tenth (operators) with more than 100. The analytical matrix has unbalanced cells, requiring that the ANOVA be based on multiple regression modeling, employing a "model-comparison" procedure for partitioning the regression sum of squares. In addition to the comprehensive model and a similar analysis addressing only the contributions of prolific operators, a number of smaller models have been used to examine the details of variance contributions in particular subsets and in individual operator databases. Details of these may be found in the earlier technical report (Nelson et al., 1991). The formal models indicate that anomalous effects appear as small statistical signals in a background of noise; in most cases the amount of variance ex- Variance in REG Experiments: ANOVA Models 77 plained is on the order of one percent or less. Hence, it is only through the accumulation of large databases that these small effects can be examined. Full Database Analysis The tables, figures, and comments in this section summarize the major results of the original 1991 analysis of variance (Nelson et al., 199 1) based on models for all (1262) replications, the prolific operator subset, and a selection of restricted subsets relevant to particular questions. Each regression model is evaluated in terms of its sum of squares (SSR), degrees of freedom (dB, associated p-value (p), and proportion of total variance explained by the parameters used in the model (R* Following this, a partitioning of the regression sum of ). squares reveals the contributions due to intention and to each of the secondary parameters. The latter combine the main effect and the interaction with intention for each parameter, and include an "unaccounted" entry that indicates the average over- or underestimation of factor contributions resulting from the model-comparison approach to the breakdown of the regression sum of squares. This is on the order of 1% or less of the total residual variance and is not significant. Tables 1-6 all have the same format, showing the composite sum of squares, degrees of freedom, F-ratio, and a p-value indicating the significance of each factor's contribution, where "p-value" refers to the chance of an outcome more extreme than the observed one, assuming the null hypothesis. In subsequent discussions, the terms "suggestive" or "marginal" refer to p-values between 0.10 and 0.05, and "significant" to those less than 0.05. Figures 1-4 illustrate the overall effect of parameters, usually by displaying the mean shift, or effect size, as a function of intention, with one-sigma error bars. The unit for the effect size is the equivalent number of bits per 50 trials, and corresponds to the number of excess bits per 10,000 binary events. The model for the entire data set, with the decomposition of the total sum of squares due to the experimental factors, is presented in Table 1. The corresponding overall mean shifts are shown in Figure 1. All data and parameters TABLE 1 All Data (All Parameters Except Operators) Model: SSR = 199220; df = 50, 113703; p = 0.0050; ' = 0.00070 R Parameter Intention Device Location Protocol Series Runlength Assignment Control Feedback Unaccounted Sum of Squares df F-Ratio Probability Nelson et al. Intentions Fig. 1. Correlation of mean shift with intention. All data are from the comprehensive model including all factors except operators. Units for the mean shift are the number of bits per 50 trials, which is equivalent to the number of bits per 10,000. Point estimates (circles) with one-sigma error bars. are included, with the exception of the operators factor, whose interactions with intention comprise more than 300 degrees of freedom, entailing calculations that exceed the computational system limits. This factor is addressed later in the prolific operators model. The regression model for the full database across all operators and parameters is significant, clearly indicating that the combined influence of the experimental parameters adds information to the nominally random distribution. The model explains less than 0.1% of the variance, however, which is consistent with the very small ratio of anomalous effect to stochastic "noise." We will see in subsequent analyses that certain well-defined subsets have an R~an order of magnitude larger, but none of the models explain much more than 1% of variance. As predicted, intention is the primary contributor to the regression, with a mean squared deviation (and F-ratio) four times that of the next largest parameter, and, as the figure shows, the relationship between intention and meanshift is consistent with the experimental hypothesis. While HI and LO are properly displaced relative to BL, the meanshift relative to theoretical expectation is greater for HI than LO. This imbalance appears consistently in various subsets but the differences are not statistically significant. The BL data also show a tendency for high deviations that are fairly consistent, although not significant, in curious contrast to the calibrations, which properly do not exhibit any consistent trends (Nelson, 1993). "Device" (type of random source) and "location" are significant and marginal secondary parameters, respectively, suggesting a need for separate examination of the various levels of Variance in REG Experiments: ANOVA Models 79 each. Briefly, such subsidiary analyses show that results with the diode source and with a non-deterministic shift-register-based device are similar to each other, but differ from those with the deterministic or algorithmic pseudo-random devices. The shift-register device originally was classified as pseudorandom, but subsequently was determined to be a combination of pseudo-random and truly random components, hence qualifying as a non-deterministic source (Jahn et al., 1997). With respect to "location" the subsidiary analysis shows a second-order interaction: local and remote data are similar, while the "B" location (operator in the next room, delayed feedback) differs, but the latter effect is driven by a confounding interaction with the device type. This will be discussed further, but we should note in this context that the database for the "B" location is comparatively small and hence any inferences must be tentative. To examine the effect of large spatial separations on the anomalous effects more thoroughly, a special-purpose analysis using linear regression was applied. The results show no trend in scoring associated with increasing separation of operator from machine, up to several thousands of miles (Dunne and Jahn, 1992). In the full model, "assignment" (whether the instruction was random or volitionally chosen) and "control" (automatic vs. manual sequencing of trials) are not represented precisely because 16 early series mixed these parameters. When these series are excluded, "assignment" is associated with a p-value of 0.52, and "control" with ap-value of 0.29, hence neither factor is a contributor in the grand concatenation. The "protocol" (the experiment's length and purpose, particular questions, etc.), "series" (series position), "runlength," and "feedback factors are all non-significant in the main analysis. For "series," however, other detailed assessments indicate a clear, albeit complex, non-linear structure (Dunne et al., 1994) which will be discussed later. In qualification tests of the multiple regression modeling procedure, the corresponding analysis applied to arbitrarily assigned calibration data in the full database model yielded a non-significant regression, and parameter contributions that were well within chance variation: one factor had a marginal pvalue (0.063), but none reached the nominal significance level (Nelson et al., 1991). Thus, the appearance of structure in the experimental database cannot be attributed to artifacts of the modeling procedure. Prolific Operator Results Table 2 and Figure 2 show the results for a subset of the full original database consisting of 1060 series produced by the 30 "prolific" operators, each of whom generated 10,000 or more trials per intention. In this model, which constitutes approximately 80% of the full database, the operator factor can be assessed, and because the various parameter levels are represented more evenly, the results may be interpreted with higher confidence. This prolific operator model is highly significant and explains more than twice as much of the variance (0.2%) as the full model, suggesting that the inclusion of both multiple 80 Nelson et al. TABLE 2 Prolific Operators (All Data and All Parameters) Model: SSR = 499986; df = 137,97135; = 4 . 4 ~ 1 0 - R = .0020 ;' ~ Parameter Intention Operators Device Location Protocol Series Runlength Assignment Control Feedback Unaccounted Sum of Squares df F-Ratio Probability Meanshift (Bits150 trl) 1 I Intentions Fig. 2. Correlation of mean shift with intention. Data are from the prolific operators model including all factors. Mean shift units are the number of bits per 50 trials, which is equivalent to the number of bits per 10,000. Point estimates (circles) with one-sigma error bars. replications by individuals and a factor that represents individual differences may serve to clarify the effects of the parameters. Again, intention is highly significant and is the primary contributor to the regression. The operator parameter also is significant, indicating individual differences in performance and the need to consider each operator's data separately to assess individual responses to parameters. Such analyses have been detailed in the earlier technical report (Nelson et al., 1991), and will be considered further in discussion of the updated ANOVA. Figure 3 shows the results of a cluster analysis of the intention-linked performance of the 30 prolific operators. There are three well-defined clusters: one for success in the direction of intention (HI - LO), one for a small number of operators with Variance in REG Experiments: ANOVA Models 81 Delta (HI-LO) SS Intentions 1 SS Total Fig. 3. Clustering of prolific operator effect sizes. The difference between HI and LO mean shift (arbitrary units) is plotted against the proportion of the sum of squares attributable to intention. Operator numbers are printed in a point size proportional to the operator's database size. The computed clusters are indicated by dotted lines around a group with large effect sizes in the intended direction (upper group) and opposite to intention (lower group). The remaining operators did not generate effects correlated with intention. large effects in the direction opposite to intention, and a third cluster for whom intention was not a significant contributor to the regression model. Both device and location appear as important secondary parameters, similar to the grand concatenation shown in Table 1, but for location we must again apply the previously mentioned caveats concerning the device interaction stemming from the small "B" data set. Feedback shows a marginal contribution, and examination of the subset means suggests, surprisingly, that the simple digital feedback and non-feedback conditions produce somewhat higher scores than the more engaging and informative graphic feedback mode while detailed examination shows this to be driven largely by early trials, which had larger effect sizes but which were done only with digital feedback. Subsequent studies suggest that feedback may indeed be an important parameter. Although the series position contribution to the multiple regression is nonsignificant, the assessment of subset means reveals a complex pattern of results as a function of series or replication number. As shown in Figure 4, the highly significant composite result in the first series declines to non-significance, and then recovers to a significant effect in later series. The figure shows that this non-linear progression of effect size is superimposed on a weakly defined linear trend, thus explaining why the contribution of this parameter to the simple regression model is modest. This pattern has been the subject of a more specific and detailed regression analysis (Dunne et al., 1994) that confirms and extends the indication of a strong influence of series position (corresponding to developing operator 82 Nelson et al. Delta (HI - LO) Series Number Fig. 4. Sequential development of effect size as a function of series position. Units of effect size are bits per 50 trials, and the differences between the HI and LO intentions for each series are plotted as point estimates (circles), with one-sigma error bars. The point labeled 5+ includes series 5 and all subsequent series. experience), despite the modest contribution from this factor in the primary ANOVA model. The subsidiary analysis reveals that a similar pattern of strong early performance followed by a decline and subsequent recovery obtains in both the high and low intentions. The quadratic component of the regression is significant ( p = 0.016) for the high-low difference, while the linear component is negligible. No such pattern is evident in the baseline data. Most of the other secondary parameters in the ANOVA model are not overall contributors across operators, but a number of individuals respond differently to the experimental variations. Detailed analyses of the associated patterns are beyond the scope of this paper, but may be found in the technical report (Nelson et al., 1991). The unaccounted variance in the prolific operator model is negative, indicating average overestimation of contributions, but this is inconsequential in magnitude. Again a corresponding analysis using arbitrarily assigned calibration data yields a non-significant regression and no evidence of non-chance variation. One parameter has a marginal p-value (0.086), but it is a factor different from that with a marginal p-value in the full database model, giving further evidence that calibration data show only normal chance fluctuations, and that the modeling procedure does not produce artifactual indications of parameter effects. ANOVA Update The specialized analyses of the REG database using the a d hoc tools and perspectives mentioned above revealed some influential variables not includ- Variance in REG Experiments: ANOVA Models 83 ed in the original, 1991 analysis, and prompted the development of corresponding parameters to be included in an update of the analysis of variance. Of course this is an a posteriori procedure, in the sense that the questions arose out of specific analyses of the data, but the new factors are reasonable extensions or modifications of those in the original ANOVA. In this updated analysis, a factor called "gender" is represented with three levels: male (59 operators), female (50 operators), and co-operators (18 pairs). Although Dunne found instructive, significant differences in the average effects of men and women (Dunne, 1991), this factor is not a significant contributor in either the new overall model (see Table 3) or the corresponding prolific operators model. The finding is consistent with Dunne's observation that the compounded effects (composite 2-scores) do not differ significantly for men and women, although their average effect sizes do differ (Dunne, 1998). This is explained by the disproportionate contributions in very large databases by three female operators whose results differ significantly from the other forty-seven. The co-operator data were not included in the 199 1 analysis, but they have been assessed directly in a specialized, ad hoe analysis (Dunne, 1991). In particular, the effect sizes have been compared with those of single operators and appear to be larger by a factor of two or more, depending on the composition of the pairs. Despite the comparatively large mean shift for the co-operators, this database is too small relative to those for the individual male and female operators to produce a significant difference across the gender factor. We made no attempt to confirm the striking "bonded-pair" results found by Dunne because this would require a further dilution of cell populations in the co-operator subset of the database. The redefinition of the device parameter as a deterministic or non-deterministic "source," and the inclusion of "gender," have two other notable effects. The previous indication that location was a significant or marginal TABLE 3 Revised Model, New Parameter Definitions: Source (Random vs. Deterministic), Gender (Female, Male, Co-operators) Model: SSR = 204391 ; df = 56,120987; p = 0.0147; R = 0.00067 Parameter Intention Source Location Protocol Series Runlength Assignment Control Feedback Gender Unaccounted Sum of Squares Probability 84 Nelson et al. parameter is mitigated here, as a result of clarifying the device distinction. Secondly, due to clearer definition of the device influences, as well as possible effects of gender differences, the series position parameter now is marginally significant. A separate, special-purpose definition of the gender factor using four levels includes, in addition, a special category that segregates the contributions of three high-performing female operators with moderate to large effect sizes and extremely large databases, resulting in disproportionate contributions to the overall effect. Their operator numbers (10, 78, 80) are notable in Figure 3, where the font size is proportional to the database size. This four-level "gender" factor is obviously defined a posteriori, and should be regarded as a means to confirm findings of other, independent analyses, (Jahn et al., 1997), within the context of the full analysis of variance. The full ANOVA model using this four-level parameter shows it to be a highly significant contributor @ = 3.4 x indicating that results for the small, selected group of operators clearly differ from the general pattern (see Table 4 and Figure 3). Although this conclusion must be tempered by the a posteriori nature of the analysis, it is clear that the combination of very large databases with moderate, positive effects contributes powerfully to the experimental outcome. Understanding the contributions of this source of variance may help to interpret other aspects of the experiment; it underscores the importance of large individual databases, where interactions with other variables are not confounded by significant individual differences. The updated regression models also include a newly defined device factor, called "source," with two levels: non-deterministic and deterministic. The former comprises the Elgenco-based REG and the first version of a hardware "pseudo-random" source, which was, in fact, non-deterministic because it employed randomly varying shift-register steps. A revised hardware pseudo-random source subsequently was developed, as well as an algorithmic pseudoTABLE 4 Revised Model, All Data: Contribution of High Performers Model: SSR = 278012; df = 59,120984; p = 0.00005 1; R~= 0.00092 Parameter Intention Source Location Protocol Series Runlength Assignment Control Feedback Gender+hi-perf Unaccounted Sum of Squares df F-Ratio Probability Variance in REG Experiments: ANOVA Models 85 random source. The latter two are properly deterministic, and data generated using them are compared with those from the truly random, non-deterministic devices by means of the new source factor. The updated models show this parameter to be significant, with ap-value of 0.020 in the grand, overall regression, and p = 0.007 1 in the prolific operator model. When the non-deterministic sources, also shown in other analyses to yield no anomalous effect (Jahn et al., 1997), are excluded (see Table 5 ) , the explanatory power of the model increases greatly to nearly 1% of the variance, and the significance of the intention factor increases to No other factor except that representing series position is a prominent contributor to this model; without the relatively noisy deterministic source data, the series parameter becomes significant. It is also worth noting that in this model location shows no suggestion of differentiation, confirming other evidence that anomalous effects are not a function of spatial separation. Table 6 presents a model restricted to the deterministic sources alone, where we find no indication that the intention factor contributes to an TABLE 5 Revised Model, Non-Deterministic Sources Only Model: SSR = 197238; df = 53,90210; p = 0.012; R2= 0.00087 Parameter Intention Location Protocol Series Runlength Assignment Control Feedback Gender Unaccounted Sum of Squares df F-Ratio Probability 8253 3454 957 6 6 0.549 0.230 0.77 0.97 TABLE 6 Revised Model, Deterministic Sources Only Model: SSR = 128858; df = 47,30732; p = 0.313; R2 = 0.0017 Parameter Intention Location Series Runlength Assignment Control Feedback Gender Unaccounted Sum of Squares df F-Ratio Probability 86 Nelson et al. explanation of variance. Only the location factor is significant in this case, while the regression model itself is non-significant. Thus, the indications of differentiation by location can be attributed to the large, though inconsistent effects in the small "B" and "C" subsets of the deterministic data. Combining these two specialized questions in a single analysis, we generate a model that includes the exploratory parameter segregating the three highperforming operators, and excludes data from deterministic sources (which show no effect). The variance explained in this model increases to well over 1% (see Table 7). Intention is the primary contributor, followed by the fourlevel gender factor, the explanatory power of which is about one-third that of intention. Of the remaining parameters, only series position appears as a marginally significant contributor. Conclusions The most important finding in both the original and the updated analyses is a significant correlation of outcome with the pre-assigned intentions, compounded largely of small contributions from many individual operators across most of the experimental conditions, but with a disproportionate contribution from the three high-performing operators. Depending upon the particular subset of the large and complex database, the statistical significance ranges up to a few parts per million, with the grand concatenation, which includes all combinations of successful and unsuccessful parameters or conditions, showing a probability for the correlation with intention on the order of 2 x These comprehensive ANOVA models provide compact summaries of the major findings of the REG experiments, with the combined effect of all measured parameters taken into account. In conjunction with the complementary analyses addressing specific questions, this approach leads to a number of well-defined conclusions: TABLE 7 Revised Model: Segregation of Non-Deterministic Sources and Contribution of High Performers Model: SSR = 291671; df = 56,90207; p = 4 . 0 ~ 1 0 - R~= 0.0013 ;~ Parameter Intention Location Protocol Series Runlength Assignment Control Feedback Gend+hi-perf Unaccounted Sum of Squares df F-Ratio Probability Variance in REG Experiments: ANOVA Models 87 1. Overall, the correlation with intention is a small-magnitude effect, equivalent to a distribution mean shift of about 1 part in 10,000, ranging up to an order of magnitude larger in certain subsets. This finding differs little between the original and updated analyses ( p = 0.00015, p = 0.00024, respectively). This result also is similar to that in Jahn et al. (1997) for the benchmark REG ( p = 0.00007), although a direct comparison is not appropriate since the ANOVA result considers the structuring effect of all three intentions, while the standard analysis regards only the differential between the HI and LO conditions. 2. The broad generality of the finding across most of the different combinations of parameters suggests a mechanism that operates at a very fundamental level. For example, a separate, specialized analysis indicates that the anomalous effect can be modeled most simply as an alteration of the fundamental binary probability of the random events (Jahn, Dobyns, and Dunne, 199 1). 3. The effect is apparently confined to non-deterministic devices. The updated version of the ANOVA confirms that deterministic pseudo-random sources do not change behavior in correlation with the operators' intentions, while non-deterministic random sources incontrovertibly do so. The full regression model indicates a significant contribution of the "source" parameter, and a model limited to data taken with non-deterministic sources is highly significant (4.0 x with the intention parameter also significant (1.0 x l ~ - ~ ) , the corresponding model for while data from deterministic sources alone is not significant, nor is the intention parameter. (We note, however, that location is significant and control is marginal in the latter model [ p = 0.002 and 0.061, respectively], and since other researchers have reported effects with pseudo-random sources (Radin and Nelson, 1989), a strong conclusion discounting deterministic sources would be premature.) 4. Although there are individual differences, there is a relatively normal distribution of effect sizes across individuals, with no indications of outliers indicating special performance in the sense that only certain "gifted" individuals can produce the anomalous effect. Nonetheless, consistent positive achievement over large databases does distinguish a few individuals. While the clear differentiation for the selected group of three operators is predetermined by the selection process, their distinctive performance is instructive. Inspection of their effect sizes shows them to fall within the range of the full distribution of operators, indicating that the differential is driven mainly by the large database sizes. 5. The gender variable, necessarily represented in the ANOVA by composite rather than average operator scores, does not show a significant contribution to the variance. Specialized analyses addressing the average effect size reveal that there are gender differences, however, comprising small variability and regular correlation with intention for men, and 88 Nelson et al. larger variability and less consistent correlations for women (Dunne, 1998). 6. The temporal development of effect sizes shows a consistent pattern, with initial significant success that declines but then recovers. A specialized analysis reveals that this is a broadly distributed pattern in the intentional conditions, which does not appear in baseline data (Dunne et al., 1994). 7 . The overall findings show essentially no evidence for a dependence on spatial or temporal separation, complementing other indications that ordinary physical variables have little impact on these anomalous interactions. A specialized analysis has supplemented this conclusion within REG and other databases (Dunne and Jahn, 1992). In summary, we conclude that both the comprehensive ANOVA technique and the more sharply focused a d hoe analyses can play important roles in the assessment of complicated databases of this sort. On the one hand, analysis of variance allows an examination of large and complex data sets as a whole, with a perspective that displays the relative strength of effects from all the measured variables in the experiment. But ANOVA has limitations, especially in a database where the cell populations in the analytical matrix vary as much as do those in the REG ANOVA. The practical requirements of the experimental program have dictated an a d lib accumulation of data under various combinations of parameters in order to explore the engineering questions motivating the experiment, while also maintaining a viable psychological context for the human operators. In the course of the experimental development, emphases have changed to include a progressively more incisive examination of subjective factors. Thus, the REG experiment, although based on an unchanging fundamental design, has a complex and unbalanced, non-orthogonal set of variables. This prevents easy assessment of higher-order interactions, even when it is apparent that interactions among the various conditions may be important contributors to the explanation of variance. It thus follows that specialized, detailed analyses can help reveal the structure of the data in delimited subsets of the database, without compromising the integrity of interpretations. For example, the implications of spatial separation are critical to modeling the anomalous effects, and an analysis that examines both the linear and higher-order regression of effect size on distance provides essential information. Likewise, the serial position effects must be important indicators of psychological factors bearing on the results, and these can be detailed only in a d hoc formats. In this context, ANOVA not only provides guidance for focused assessments, showing, for example, that a specialized analysis must address the non-deterministic and deterministic sources separately, but it also confirms the legitimacy of the a d hoc studies. Thus, in these and other cases previously discussed, we see a complementary balance between the global perspective provided by ANOVA and the sharply focused, incisive answers provided by well-posed supplementary analyses. Variance in REG Experiments: ANOVA Models Acknowledgments The Princeton Engineering Anomalies Research program has been supported by a number of foundations and individuals, including the Institut fiir Grenzgebiete der Psychologie und Psychohygiene, the Lifebridge Foundation, the Ohrstrom Foundation, Mr. Richard Adams, Mr. Laurance Rockefeller, and Mr. Donald Webster. We acknowledge also the contributions of the volunteer operators, and the editorial assistance of Elissa Hoeger. References Dunne, B. J. (1991). Co-operator experiments with an REG device. PEAR Technical Report 91005. Princeton Engineering Anomalies Research, Princeton, NJ. Dunne, B. J. (1998). Gender differences in humanlmachine anomalies. Journal of Scientific Exploration, 12, 1, 3-55. Dunne, B. J., Dobyns, Y. H., Jahn, R. G., & Nelson, R. D. (1994). Series position effects in random event generator experiments, with appendix by Angela Thompson. Journal of Scientific Exploration, 8, 2, 197-216. Dunne, B. J., & Jahn, R. G. (1992). Experiments in remote humanlmachine interaction. Journal of Scientific Exploration, 6, 4 , 31 1-332. Ibison, M. (1998). Evidence that anomalous statistical influence depends on the details of the random process. Journal of Scientific Exploration, 12, 3,407-424. Jahn, R. G., Dobyns, Y. H., & Dunne, B. J. (1991). Count population profiles in engineering anomalies experiments. Journal ofscientific Exploration, 5, 2,205-232. Jahn, R. G., Dunne, B. J., Nelson, R. D., Dobyns, Y. H., & Bradish, G . J. (1997). Correlations of random binary sequences with prestated operator intention: A review of a 12-year program. Journal of Scientific Exploration, 11, 3,345-367. Nelson, R. D., Bradish, G. J., & Dobyns, Y. H. (1989). Random event generator qualification, calibration and analysis. PEAR Technical Report 89001. Princeton Engineering Anomalies Research, Princeton, NJ. Nelson, R. D., Bradish, G. J., & Dobyns, Y. H. (1992). The portable PEAR REG: Hardware and software documentation. PEAR Internal Document 92-1. Princeton Engineering Anomalies Research, Princeton, NJ. Nelson, R. D. (1993). CONTCAL: Continuous automatic calibrations, pseudo-intentions, and active experiments. PEAR Internal Document 93-4. Princeton Engineering Anomalies Research, Princeton, NJ. Nelson, R. D., Jahn, R. G., Dunne, B. J., Dobyns, Y. H., & Bradish, G. J. (1997). FieldREG 11: Consciousness field effects, replications and explorations. Journal of Scientific Exploration, 12, 3,425-454. Nelson, R. D., Dobyns, Y. H., Dunne, B . J., & Jahn, R. G. (1991). Analysis of variance of REG experiments: Operator intention, secondary parameters, database structure. PEAR Technical Report 91004. Princeton Engineering Anomalies Research, Princeton, NJ. Radin, D. I., & Nelson, R. D. (1989). Consciousness-related effects in random physical systems. Foundations of Physics, 19, 12, 1499-15 14. Journal of Scientific Exploration, Vol. 14, No. 1 , pp. 9 1-106,2000 0892-33 10/00 O 2000 Society for Scientific Exploration Publication Bias: The "File-Drawer" Problem in Scientific Inference Space Science Division National Aeronautics and Space Administration Ames Research Center MS 245-3, Moffett Field, CA 94035-1000 e-mail: [email protected] It is human nature for "the affirmative or active to effect more than the negative or privative. So that a few times hitting, or presence, countervails ofttimes failing or absence." -Francis Bacon, The Advancement of Learning Abstract- Publication bias arises whenever the probability that a study is published depends on the statistical significance of its results. This bias, often called the file-drawer effect because the unpublished results are imagined to be tucked away in researchers' file cabinets, is a potentially severe impediment to combining the statistical results of studies collected from the literature. With almost any reasonable quantitative model for publication bias, only a small number of studies lost in the file drawer will produce a significant bias. This result contradicts the well-known fail-safe file-drawer (FSFD) method for setting limits on the potential harm of publication bias, widely used in social, medical, and psychic research. This method incorrectly treats the file drawer as unbiased and almost always misestimates the seriousness of publication bias. A large body of not only psychic research, but medical and social science studies as well, has mistakenly relied on this method to validate claimed discoveries. Statistical combination can be trusted only if it is known with certainty that all studies that have been carried out are included. Such certainty is virtually impossible to achieve in literature surveys. Key words: publication bias statistics - meta-analysis - file drawer effect - 1. Introduction: Combined Studies The goal of many studies in science, medicine, and engineering is the measurement of a quantity to detect a suspected effect or to gain information about a known one. Observational errors and other noise sources make this a statistical endeavor in which one obtains repeated measurements to average out these fluctuations. If individual studies are not conclusive, improvement is possible by combining the results of different measurements of the same effect. The idea is to 92 J. D. Scargle perform statistical analysis on relevant data collected from the literature1 to improve the signal-to-noise ratio (on the assumption that the noise averages to zero). (For clarity and consistency with most of the literature, individually published results will be called studies throughout this paper, and the term analysis will refer to the endeavor to combine two or more studies.) Two problems arise in such analyses. First, experimenters often publish only statistical summaries of their studies, not the actual data. The analyst then is faced with combining the summaries, a nontrivial technical problem (e.g., Rosenthal, 1978, 1995). Modern communication technology should circumvent these problems by making even large data arrays accessible to other researchers. Reproducible research (Buckheit and Donoho, 1995; Claerbout, Schwab, and Karrenbach, 1999) is a discipline for doing this and more, but even this methodology does not solve the other problem, which is publication bias. 2. Publication Bias The second problem facing combined analysis is that studies collected from the literature often are not a representative sample. Erroneous statistical conclusions may result from a prejudiced collection process or if the literature is itself a biased sample of all relevant studies. The latter, publication bias, is the subject of this paper. The bibliography contains a sample of the rather large literature on the file-drawer effect and publication bias. The essence of this effect, as described by nearly all authors, can be expressed in statistical language as follows: A publication bias exists if the probability that a study reaches the literature, and is thus available for combined analysis, depends on the results of the study. What matters is whether the experimental results are actually used in the combined analysis, not just the question of publication. That is, the relevant process is the following sequence, in its entirety, following the initiation of the study: Study is carried out to a predefined stop point. All data are permanently recorded. Data are reduced and analyzed. A paper is written. The paper is submitted to a journal. The journal agrees to consider the paper. The referee and author negotiate revisions. The referee accepts the paper. In some fields, this is called rneta-analysis, from the Greek meta, meaning behind, after, higher or beyond, and often denoting change. Its usage here presumably refers to new issues arising in the statistical analysis of combined data, such as the file-drawer effect itself. It is used, mostly in scientific terminology, to imply a kind of superior or oversight status, as in metaphysics. I prefer the more straightforward term combined analysis. ~ 9. 10. 1 1. 12. 13. 14. Publication Bias The editor accepts the paper for publication. The author still wishes to publish the paper. The author's institution agrees to pay page charges. The paper is published. The paper is located during a literature search. Data from the paper are included in combined analysis. I refer to this concatenated process loosely as publication, but it is obviously more complex than what is usually meant by the term. Some of the steps may seem trivial, but all are relevant. Each step involves human decisions and so may be influenced by the result of the study. Publication probability may depend on the specific c o n c l ~ s i o non the size of the effect measured, and on the ,~ statistical confidence in the result. Here is the prototype for the analyses treated here: Each of the publishable studies consists of repeated measurements of a quantity, say x. The number of measurements is, in general, different in each such study. The null hypothesis is that x is normally distributed, for example: this notation means that the errors in x are normally distributed with mean p and variance o. The results of the study are reported in terms of a shifted (i.e., p is subtracted) and renormalized (i. e . , ois divided out) standard normal deviate Z = (x - p)lo. The null hypothesis-usually that p is zero, one half, or some other specific value-yields I This normalization removes the dependence on the number of repeated measurements in the studies. In this step, it is important that o be well estimated, often a tricky business. A common procedure for evaluating such studies is to obtain a p value from the probability distribution P(Z) and interpret it as providing the statistical significance of the result. The discussion here is confined to this approach because it has been adopted by most researchers in the relevant fields. Nonetheless, as noted by Sturrock (Sturrock, 1994, 1997) and others (Jefferys, 1990, 1995; Matthews, 1999), this procedure may yield incorrect conclusions-usually overestimating the significance of putative anomalous results. The Bayesian methodology is probably the best way to treat publication bias (Big- 1 The literature of social sciences contains horror stories of journal editors and others who consider a study worthwhile only if it reaches a statistically significant, positive conclusion; that is, an equally significant rejection of a hypothesis is not considered worthwhile. 94 J. D. Scargle gerstaff, 1995; Biggerstaff, Tweedie, and Mengersen, 1994; Givens, Smith, and Tweedie, 1995, 1997; Tweedie et al., 1996). This section concludes with some historical notes. Science has long known the problems of biased samples. An interesting historical note (Petticrew, 1998) nominates an utterance by Diagoras of Melos in 500 BC as the first historical mention of publication bias.3 See also Dickersin and Min (1993) for other early examples of awareness of publication bias. Publication bias is an important problem in medical studies (e.g., Allison, Faith, and Gorman, 1996; Dickersin, 1997; Earleywine, 1993; Faber and Galloe, 1994; Helfenstein and Steiner, 1994; Kleijnen and Knipschild, 1992; LaFleur et al., 1997; Laupacis, 1997; Persaud, 1996; Sacks et al., 1983), as well as in other fields (see Bauchau, 1997; Fiske, Rintamaki, and Karvonen, 1998; Riniolo, 1997; Silvertown and McConway, 1997; and Stanton and Shadish, 1997, for examples). The negative conclusions we shall soon reach about the commonly used procedure to deal with this problem yield a discouraging picture of the usefulness of combined analysis in all of these contexts. On the other hand, the application of modern, sophisticated methods such as those listed above (see also Taylor, 2000) is encouraging. 3. The "Fail Safe File-Drawer" Calculation Rosenthal's influential work (Rosenthal, 1979, 1984, 1990, 199 1, 1995) is widely used to set limits on the possibility that the file-drawer effect is causing a spurious result. One of the clearest descriptions of the overall problem, and certainly the most influential in the social sciences, is researchers and statisticians have long suspected that the studies published in the behavioral sciences are a biased sample of the studies that are actually carried out.. . . The extreme view of this problem, the "file drawer problem," is that the journals are filled with the 5% of the studies that show Type I errors, while the file drawers back at the lab are filled with the 95% of the studies that show nonsignificant (e.g., p > .05) results. (Rosenthal, 1979, p. 638) A Type I error is rejection of a true null hypothesis. (Type I1 is failing to reject a false one.) This lucid description of the problem is followed by a proposed solution: In the past, there was very little we could do to assess the net effect of studies tucked away in file drawers that did not make the magic .05 level.. . Now, however, although Early Greek sailors who escaped from shipwrecks or were saved from drowning at sea displayed portraits of themselves in a votive temple on the Aegean island of Samothrace, in thanks to Neptune. Answering a claim that these portraits are sure proof that the gods really do intervene in human affairs, Diagoras replied "Yea, but ... where are they painted that are drowned?" Publication Bias 95 no definitive solution to the problem is available, we can establish reasonable boundaries on the problem and estimate the degree of damage to any research conclusions that could be done by the file drawer problem. The fundamental idea in coping with the file drawer problem is simply to calculate the number of studies averaging null results that must be in the file drawers before the overall probability of a Type I error can be just brought to any desired level of significance, s a y p = .05. This number of filed studies, or the tolerance for future null results, is then evaluated for whether such a tolerance level is small enough to threaten the overall conclusion drawn by the reviewer. If the overall level of significance of the research review will be brought down to the level of just significant by the addition of just a few more null results, the finding is not resistant to the file drawer threat. (Rosenthal, 1984, p. 108) [The italic emphasis is original; I have indicated what I believe is the fundamental flaw in reasoning with boldface.] By its very definition, the file drawer is a biased sample. In the nominal example given, it is the 95% of the studies that have 5% or greater chance of being statistical fluctuations. The mean Z in this subsample is It is not zero. As we will see below, Rosenthal's analysis explicitly assumed that 2 = 0 for the file-drawer sample. Because this assumption contradicts the essence of the file-drawer effect, the quantitative results are incorrect. I now recapitulate the analysis given in Rosenthal(1984), using slightly different notation. For convenience and consistency with the literature, we refer to this as the fail-safe file-drawer, or FSFD, analysis. The basic context is a specific collection of published studies having a combined Z that is deemed to be significant, that is, the probability that the Z value is attributable to a statistical fluctuation is below some threshold, say 0.05. The question Rosenthal sought to answer is, How many members of a hypothetical set of unpublished studies have to be added to this collection to bring the mean Z down to a level considered insignificant? As argued elsewhere, this does not mirror publication bias, but this is the problem addressed. Let Npubbe the number studies combined in the analysis of the published literature (Rosenthal's K), and Nfiledbe the number of studies that are unpublished, for whatever reason (Rosenthal's X). Then N = Npub+ Nfiledis the total number of studies carried out. The basic relation (Equation 5.16 of Rosenthal, [1984]) is, This is presumably derived from i.e., an application of his method of adding weighted 2's as in Equation 5.5 of 96 J. D. Scargle Rosenthal, by setting the standard normal deviate of the file drawer, Zfiled 0. = This is incorrect for a biased file drawer. Equation 4 can be rearranged to give which is the equation used widely in the literature, and throughout this paper, to compute FSFD estimates. What fundamentally went wrong in this analysis, and why has it survived uncriticized for so long? First, it is simply the case that the notion of null, or insignificant results is easily confused with z = 0. While the latter implies the former, the former (which is true, in a sense, for the file drawer) does not imply the latter. Second, the logic behind the FSFD is seductive. Pouring a flood of insignificant studies-with normally distributed 2s-into the published sample until the putative effect submerges into insignificance is a great idea. What does it mean, though? It is, indeed, a possible measure of the statistical fragility of the result obtained from the published sample. On the other hand, there are much better and more direct ways of assessing statistical significance; I do not believe that FSFD should be used even in this fashion. Third, I believe some workers have mechanically calculated FSFD results, found that it justified their analysis or otherwise confirmed their beliefs, and were therefore not inclined to look for errors or examine the method critically. A simple thought experiment makes the threat of a publication bias with Nfiled on the same order as Npub clear: Construct a putative file-drawer sample by multiplying all published Z values by -1; then the total sample then has Z exactly zero, no matter what. The two-sided case often is raised in this context. That is to say, publication bias might mean that studies with either Z > 0 or 2 < 0 are published in pref> < erence to those with small IZI. This situation could be discussed, but it is a red herring in the current context (e.g., Iyengar and Greenhouse, 1988). Further, none of the empirical distributions I have seen published show any hint of the bimodality that would result from this effect, including those in Radin (1997). In addition, I first was puzzled by statements such as, FSFD analysis assesses "tolerance for future null results." The file-drawer effect is something that has already happened by the time one is preparing a combined analysis. I concluded that such expressions are a kind of metaphor for what would happen if the studies in the file drawers were suddenly made available for analysis. But even if this imagined event were to happen, one would still be left to explain how a biased sample was culled from the total sample. It seems to me that whether the result of opening this Pandora's file drawer is to dilute the putative effect into insignificance is of no explanatory value in this context-even if the calculation were correct. In any case, the bottom line is that the widespread use of the FSFD to conclude that various statistical analyses are robust against the file-drawer effect Publication Bias 97 is wrong because the underlying calculation is based on an inapplicable premise. Other critical comments have been published, including Berlin (1998) and Sharpe (1997). Most important is the paper by Iyengar and Greenhouse (1988), which makes the same point but further provides an analysis that explicitly accounts for the bias (for the case of the truncated selection functions discussed in the next section). These authors note that their formulation "always yields a smaller estimate of the fail-safe sample size than does" Rosenthal's. 4. Statistical Models This section presents a specific model for combined analysis and the filedrawer effect operating on it. Figure 1 shows the distribution of N = 1,000 samples from a normal distribution with zero mean and unit variance, namely Publication bias means that the probability of completion of the entire Null Hypothesis -- 1000 Studies p-value: 1.86-48 Filed: a> -0.1 12 = Published: Fig. 1. Histogram corresponding to the null hypothesis: normal distribution of Z values from 1,000 independent studies. Those 5% with Z 2 1.645 are published (open bars), the remainder are "tucked away in file drawers," i.e., unpublished (solid bars). Empirical and exact values of Z (cf. Equation 14) are indicated for the filed and published sets. 98 J . D. Scargle process detailed in Section 2 is a function of the study's reported Z. Note that I am not assuming this is the only thing that it depends on. I use the notation S(Z) = publication probability (8) for the selection function, where 0 _< S(Z) _< 1. Note S(Z) is not a probability distribution over Z ; for example, its Z integral is not constrained to be 1. This model captures what almost all authors describe as publication bias (e.g., Hedges, 1992; Iyengar and Greenhouse, 1988). In view of the complexity of the full publication process (see Section 2), it is unlikely that its psychological and sociological factors can be understood and accurately modeled. I therefore regard the function S as unknowable. The approach taken here is to study the generic behavior of plausible quantitative models, with no claim to having accurate or detailed representations of actual selection functions. Taken with the distribution G(Z)in Equation 7, an assumed S(Z) immediately yields three useful quantities: the number of studies published the number consigned to the file drawer and the expected value of the standard normal deviate Z, averaged over the published values The denominator in the last equation takes into account the reduced size of the collection of published studies (Npub)relative to the set of all studies performed (N). Absent any publication bias [i.e., S(Z)= 11 2 = 0 from the symmetry of the Gaussian, as expected. Cut Of Selection Functions Consider first the following choice for the selection function: S(Z)= 0 for Z < Zo 1 for Z > Zo, where, in principle, Zo can have any value, but small positive numbers are of most practical interest. That is to say, studies that find an effect at or above the significance level corresponding to Zo are always published, whereas those that do not are never published. Putting this into the above general expressions Equations 9 and 11 gives Publication Bias 99 where erfc is the complementary error function. The mean Z for these published studies is Consider the special value Zo = 1.645, corresponding to the classical 95% confidence level and obtained by solving for Z the equation 1 p = - erfc 2 (2) (with p = .05). This case frequently is taken as a prototype for the file drawer (Iyengar and Greenhouse, 1988; Rosenthal, 1979). Equation 13 gives Npub = 0.05N; that is (by design) 5% of the studies are published and 95% are not published. Equation 14 gives 2 = 2.0622. Let the total number of studies, be N = 1,000, i.e., 1,000 studies, of which 50 are published and 950 are not. We have chosen this large value for N, here and in Figure 1, for illustration, not realism. For the 50 published studies, the combined Z, namely m , Equation 15 gives an infinitesimal p value, = 2 in highly supportive of rejecting the null hypothesis. The FSFD estimate of the ratio of filed to published experiments (see equation 6) is about 78 for this case, an overestimate by a factor of around 4 of the true value of 19. The formula of Iyengar and Greenhouse (1988) discussed above gives 11.4 for the same ratio, an underestimate by a factor of about 1.7. Finally, Figure 2 shows the behavior of the filed and published fractions, as a function of Zo, including the fraction of the studies in the file drawer predicted by FSFD analysis, given by Equation 6 above. Two things are evident from this comparison: The FSFD prediction is a strong function of N, whereas the true filed fraction is independent of N. The FSFD values are far from the truth, except accidentally at a few special values of N and Zo. Step Selection Functions Because it is zero over a significant range, the selection function considered in the previous section may be too extreme. The following generalization allows any value between zero and one for small Z: S(Z) = So for Z < Zo 1 for Z > Zo (0 i So 5 1). J. D. Scargle Published (solid) Filed (dashed) Fig. 2. Plot of the fraction of the total number of studies that are published (thick solid line) and filed (thick dashed line) in the model given by Equation 12. The FSFD predictions for the filed fraction are shown as a series of thin dashed lines, labeled by the value of N, the total number of studies carried out. Publication Bias Much as before, direct integration yields SO (1 - SO) erfc - + 2 and Equation 17 directly determines the ratio of the number of filed studies to the number of those published under the given selection function. The value of 2 is the quantity that would be (incorrectly) used to reject the hypothesis that there is no effect present and the sample is drawn from a normal distribution. Figure 3 shows this function. The basic result is that almost all of the rele- Contours of log ( NtikdI Npub ) Fig. 3. This figure shows the dependence of R = Nfiled INpub the two parameters of the step seon lection function. Contours of logloR are indicated. R is large only for a tiny region at the bottom right of this diagram. 102 J. D. Scargle vant part of the So-Zo plane corresponds to a rather small number of unpublished studies. In particular, R > 1 in only a small region, namely where simul> taneously So 0 and Zo > 0. The behavior of 3 is simple: roughly speaking > - where g(Zo) is a function on the order of unity or larger for all Zo > 0. Hence, the bias brought about by even a small file drawer is large unless So 1 (to be expected, because for such values of So the selection function is almost flat). Smooth Selection Functions It might be objected that the selection functions considered here are unreal in that they have a discrete step, at Zo. What matters here, however, are integrals over 2, which do not have any pathologic behavior in the presence of such steps. Nevertheless, I experimented with some smooth selection functions and found results that are completely consistent with the conclusions reached in the previous sections for step function choices. 5. Conclusions Based on the models considered here, I conclude the following: Apparently significant, but actually spurious, results can arise from publication bias, with only a modest number of unpublished studies. The widely used fail-safe file-drawer (FSFD) analysis is irrelevant because it treats the inherently biased file drawer as unbiased and gives grossly wrong estimates of the size of the file drawer. Statistical combinations of studies from the literature can be trusted to be unbiased only if there is reason to believe that there are essentially no unpublished studies (almost never the case!). It is hoped that these results will discourage combined ("meta") analyses based on selection from published literature but encourage methodology to control publication bias, such as the establishment of registries (to try to render the probability of publication unity once a study is proposed and accepted in the registry). The best prospects for establishing conditions under which combined analysis might be reasonable even in the face of possible publication bias seem to lie in a fully Bayesian treatment of this problem (Givens, Smith, and Tweedie, 1997; Smith, Givens, and Tweedie, 1999; Sturrock, 1994, 1997). It is possible that the approach discussed in Radin and Nelson (1989) can lead to improved treatment of publication bias, but one must be cautious when experimenting with ad hoc distributions and be aware that small errors in fitting the tail of a distribution can be multiplied by extrapolation to the body of the distribution. Publication Bias Acknowledgements I am especially grateful to Peter Sturrock for guidance and comments in the course of this work. This work was presented at the Peter A. Sturrock Symposium, held on March 20, 1999, at Stanford University. I thank Kevin Zahnle, Aaron Barnes, Ingram Olkin, Ed May, and Bill Jefferys for valuable suggestions and Jordan Gruber for calling my attention to the book The Conscious Universe (Radin, 1997), and its author, Dean Radin, for helpful discussions. None of these statements are meant to imply that any of the persons mentioned agrees with the conclusions expressed here. References Allison, D. B., Faith, M. S., & Gorman, B. S. ( 1996). Publication bias in obesity treatment trials? International Journal of Obesity, 20, 931-937. Discussion of publication bias in obesity treatment studies, mentioning the FSFD method. Bauchau, V. (1997). Is there a "file drawer problem" in biological research? OIKOS, 19,407-409. Concludes that the extent of the file-drawer problem in biology is an open question. Berlin, J. A. (1998). Publication bias in clinical research: Outcome of projects submitted to ethics committees. In Perman, J. A., & Rey, J., (Eds.), Clinical Trials in Infant Nutrition. Nest16 Nutrition Workshop Series, Vol. 40. Philadelphia: Vevet-Lippincott-Raven Publishers. Interesting overview of the subject of publication bias. "The existence of publication bias is clearly demonstrated by the studies described." (That is, various health and nutrition studies.) A clear restatement of the FSFD analysis is followed by the statement that "this approach is of limited utility for two reasons: First, because it uses only Z statistics and ignores quantitative estimates of effects (for example odds ratios), and second, because the assumption that all the unpublished studies have a Z statistic of exactly zero is unrealistic." The paper is followed by an interesting, free-wheeling discussion that touches on publication in obscure journals (something that could be added to the discussion in Section 2; something called reference bias), electronic journals, and registries. Biggerstaff, B. J. (1995). Random effects methods in meta-analysis with application in epidemiology. Colorado State University, Unpublished doctoral dissertation, Ft. Collins. Comprehensive discussion of a number of issues in combined studies, including random effects and Bayesian methods with application to environmental tobacco smoke and lung-cancer studies. Biggerstaff, B. J., Tweedie, R. L., & Mengersen, K. L. (1994). Passive smoking in the workplace: Classical and bayesian meta-analyses. International Archives o Occupational and Environf mental Health, 66, 269-277. Discussion and comparison of classical and Bayesian approaches to combined analysis. Buckheit, J., & Donoho, D. (1995). WaveLab and reproducible research. In Antoniadis, A. & Oppenheim, G. (eds.), Wavelets and statistics. Lecture Notes in Statistics No. 103. Springer-Verlag. Claerbout, J., Schwab, M., & Karrenbach, M. (1999). Various documents at http://sepwww. stanford.edu/research/redoc/. Csada, R. D., James, P. C., & Espie, R. H. M. (1996). The "file drawer problem" of non-significant results: Does it apply to biological research? OIKOS, 76,591-593. An empirical study that quantitatively supports the suspicion that publication bias is important in biology. Dickersin, K. (1997). How important is publication bias? A synthesis of available data. AIDS Education and Prevention, 9A, 15-21. Empirical study that reaches the conclusion that "publication is dramatically influenced by the direction and strength of research findings." Found that editors were less enthusiastic about including unpublished studies in combined analysis. Dickersin, K., & Min, Y. (1993). Publication bias: The problem that won't go away. Ann. NY Acad. Sci., 703, 135-146. 104 J. D. Scargle Earleywine, M. (1 993). The file drawer problem in the meta-analysis of subjective responses to alcohol. American Journal of Psychiatry, 150, 1435-1436. Applies the FSFD calculation to a combined analysis of 10 published studies by V. Pollock of subjective responses to alcohol and risk for alcoholism. The resulting FSFD N is 26. In a response, Pollock agrees with the (dubious) conclusion that this means the analysis can be considered resistant to the file-drawer problem. Faber, J., & Galloe, A. M. (1994). Changes in bone mass during prolonged subclinical hyperthyroidism due to L-thyroxine treatment: A meta-analysis. European Journal of Endocrinology, 130,350-356. Analysis combining 13 studies of a thyroid treatment; the FSFD calculation yielded 18 as the "fail-safe N," which seems to be taken as evidence of the resistance of the results of the analysis to publication bias. Fiske, P., Rintamaki, P. T., & Karvonen, E. (1998). Mating success in lekking males: A metaanalysis. Behavioral Ecology, 9, 328-338. Combined analysis of zoological data. "When we used this (the FSFD) method on our significant effects, the number of 'hidden studies' needed to change our results ranged from 15 ... to 3183 . .., showing that most of our results are quite robust." Robust against including a set of unbiased unpublished results, but not against the relevant biased set. Givens, G. H., Smith, D. D., & Tweedie, R. L. (1995). Estimating and adjusting for publication bias using data augmentation in bayesian meta-analysis (Technical Report 95131). Ft. Collins, CO: Colorado State University, Department of Statistics. A Bayesian analysis of publication bias, using the data augmentation principle, with simulations and application to data from 35 studies of the relation between lung cancer and spousal exposure to environmental tobacco smoke. Givens, G. H., Smith, D. D., & Tweedie, R. L. (1997). Publication bias in meta-analysis: A Bayesian data-augmentation approach to account for issues exemplified in the passive smoking debate. Statistical Science, 12, 221-250. A Bayesian analysis of publication bias, based on choosing a model for the selection function and marginalizing over the parameters of the model. LaFleur, B., Taylor, S., Smith, D. D., & Tweedie, R. L. (1 997). Bayesian assessment of publication bias in meta-analysis of cervical cancer and oral contraceptives. University of Colorado reprint? Application of a Bayesian method for assessing publication bias to studies of the possible effects of oral contraceptive use on the incidence of cervical cancer. They conclude that publication bias, probably caused by the explicit disregard of "low quality" studies, yielded a spurious statistical connection between oral contraceptive use and the incidence of cervical cancer in a previous study. Hedges, L. V. (1992). Modeling publication selection effects in meta-analysis. Statistical Science, 7,246-255. Interesting overview, with comments on the nature of the publishing process and results that are "not statistically significant." Introduces a quantitative model for publication bias that is much like that given here and offers a quantitative test for the presence of publication bias. Helfenstein, U., & Steiner, M. (1994). Fluoride varnishes (Duraphat): A meta-analysis. Community Dentistry and Oral Epidemiology, 22, 1-5. The authors apply the FSFD method to a combined analysis of studies designed to detect the cavity-preventive effect of a fluoride varnish called Duraphat. Application of the FSFD method yields the conclusion (unwarranted, based on my conclusions) that "It is very unlikely that underreporting of non-significant results could reverse the conclusion into an overall nullresult." Iyengar, S., & Greenhouse, J. B. (1988). Selection models and the file-drawer problem. Statistical Science, 3, 109-135. An excellent overview and nearly definitive study using much the same selection function approach as used here, reaching more or less the same critical conclusions. In addition, these authors' Equation 4 offers a presumably improved basis for the FSFD estimate. The paper is followed by extensive comments by Larry Hedges; Robert Rosenthal and Donald Rubin; Nan Laird, G . Patil and C. Taillie; M. Bayarri; C. Radhakrishna Rao; and William DuMouchel-all followed by a rejoinder from Igengar and Greenhouse. Much of this discussion seems to evade the simple issue raised here. Publication Bias 105 Jefferys, W. H. (1990). Bayesian analysis of random event generator data. Journal of Scientific Exploration, 4, 153-169. This excellent paper criticizes the classical use and interpretation of "p values" in the analysis of data from psychic random number generator experiments and argues for a Bayesian alternative. Jefferys, W. H. (1995). Letter to the editor. Journal of Scientific Exploration, 9, 121-122, 595-597. Incisive comments on the meaning of "p values." Kleijnen, J., & Knipschild, P. (1992). Review articles and publication bias. Arzneimittel Forschung/Drug Research, 42, 587-591. Overview of publication bias in drug research. Laupacis, A. (1997). Methodological studies of systematic reviews: Is there publication bias? Archives of Internal Medicine, 157, 357-358. Brief suggestion that publication bias should be considered in medical research. Matthews, Robert A. J. (1999). Significance levels for the assessment of anomalous phenomena. Journal of Scientific Exploration, 13, 1-7. Persaud, R. (1 996). Studies of ventricular enlargement. Archives of General Psychiatry, 53, 1165. Criticism of a combined study of ventricular enlargement in patients with mood disorders for misinterpretation of the results of a FSFD calculation. In the case at issue, a "fail-safe N" of 10 results for an analysis with 11 studies. The closing statement, "A few studies that are highly significant, even when their combined P value is significant, may well be misleading because only a few unpublished, or yet to be published, studies could change the combined significant result to a nonsignificant result." seems to capture what people must have in mind for the significance of the FSFD computation. Petticrew, M. (1998). Diagoras of Melos (500 BC): An early analyst of publication bias. The Lancet, 352, 1558. Radin, D. 1. (1997). The conscious universe: The scientific truth ofpsychic phenomena. NewYork: HarperEdge. A number of combined studies are displayed and offered as evidence for the reality of psychic phenomena. The possibility that these results are spurious and attributable to publication bias is considered and then rejected because the FSFD calculation yields huge values for the putative file drawer. It is my opinion that, for the reasons described here, these values are meaningless and that publication bias may well be responsible for the positive results derived from combined studies. Radin, D. I., & Nelson, R. D. (1989). Evidence for consciousness-related anomalies in random physical systems. Foundations of Physics, 19, 1499-15 14. Combined study of 597 experimental random number generator studies. An innovative method of dealing with the file drawer effect is used, based on fitting the high-end tail of the Z distribution. I question the use of an exponential model for this tail, justified by the comment that it is used "to simulate the effect of skew or kurtosis in producing the disproportionately long positive tail." Riniolo, T. C. (1997). Publication bias: A computer-assisted demonstration of excluding nonsignificant results from research interpretation. Teaching of Psychology, 24, 279-282. Describes an educational software system that allows students to perform statistical experiments to study the effects of publication bias. Rosenthal, R. (1978). Combining results of independent studies. Psychological Bulletin, 85, 185-193. Comprehensive and influential (158 references) analysis of the statistics of combined studies. Rosenthal, R. (1 979). The "file drawer problem" and tolerance for null results. Psychological Bulletin, 86, 638-641. This is the standard reference, cited in almost all applied research in which the file-drawer effect is at issue. Rosenthal, R. (1984). Applied Social Research Methods Series, Vol. 6. Meta-analytic procedures for social research. Newbury Park, CA: Sage Publications. The treatment of the file-drawer effect (Chapter 5, Section 1I.B) is essentially identical to that in Rosenthal(1979). Rosenthal, R. (1990). Replication in behavioral research. In Handbook of replication research in the behavioral and social sciences [Special issue]. Journal of Social Behavior and Personality, 5, 1-30. J. D. Scargle Overview of the issue of replication, including a brief discussion of the FSFD approach. Rosenthal, R. (1995). Writing meta-analytic reviews. Psychological Bulletin, 118, 83-192. A broad overview of combined studies. Rosenthal, R. (1991). Meta-analysis: A review. Psychosomatic Medicine, 53, 247-271. Comprehensive overview of various procedural and statistical issues in analysis of combined data. There is a section on the file-drawer problem (p. 260), more or less repeating the analysis in Rosenthal(1979) with an illustrative example. Sacks, H. S., Reitman, D., Chalmers. T. C., & Smith, H., Jr. (1983). The effect of unpublished studies (the file drawer problem) on decisions about therapeutic efficacy. Clinical Research [abstract], 31, 236. This abstract applies the FSFD calculation to a number of medical studies, dealing with anticoagulants for acute myocardial infarction and other treatments. The computed file-drawer sizes are not large in most cases, but the concluding statement, "This is a useful measure of the strength of the published evidence on any question," is wrong for reasons given in this paper. Sharpe, D. (1997). Of apples and oranges, file drawers and garbage: Why validity issues in metaanalysis will not go away. Clinical Psychology Review, 17, 881-901. This is an extensive review of a number of problems in combined analysis and includes an extensive critical discussion of the FSFD approach, commenting that the rule behind it is arbitrary, but falling short of saying that the method is wrong. Silvertown, J., & McConway, K. J. (1997). Does "publication bias" lead to biased science? OIKOS, 79, 167-168. Short commentary on publication bias in biology. The authors mention several techniques for studying the presence and scope of publication bias and generally downplay the seriousness of the problem. Smith, D. D., Givens, G. H., & Tweedie, R. L. (1999). Adjustment for publication and quality bias in Bayesian meta-analysis. In Berry, D., & Stangl, D. (Eds.), Meta-analysis in medicine and health policy (Chaper 12). New York: Marcel Dekker. Development of a data augmentation technique to assess potential publication bias. Stanton, M. D., & Shadish, W. R. (1997). Outcome, attrition, and family-couples treatment for drug abuse: A meta-analysis and review of the controlled, comparative studies. Psychological Bulletin, 122, 170-191. Combined analysis of a large number of psychological studies; mentions the FSFD problem but does not appear to do anything about it. Sturrock, P. A. (1994). Applied scientific inference. Journal o Scientific Exploration, 8,491-508. f A clear and incisive discussion of a Bayesian formalism for combining judgements in the evaluation of scientific hypotheses from empirical data, such as in investigations of anomalous phenomena. Sturrock, P. A. (1997). A Bayesian maximum-entropy approach to hypothesis testing, for application to RNG and similar experiments. Journal of Scientific Exploration, 11, x-xx. Cogent critique of the standard " p value" test in applied scientific inference. Shows that the Bayesian formalism, based on a clear statement and enumeration of the relevant hypotheses, is superior to and different from " p value" analysis. Taylor, S. (2000). Several papers available at http://www.stat.colostate.edu/-taylorl http:// and www.stat.colostate.edu/-duvall. Taylor, S., & Tweedie, R. (1999). Practical estimates of the effect of publication bias in metaanalysis. U. Colorado preprint. Tweedie, R. L., Scott, D. J., Biggerstaff, B. J., & Mengersen, K. L. (1996). "Bayesian meta-analysis, with application to studies of ETS and lung cancer. Lung Cancer, 14(Suppl. l), S171-194. A comprehensive analysis of publication bias, comparing a Markov chain Monte Carlo technique for implementing Bayesian hierarchical models with random effects models and other classical methods. Journal of Scientific Exploration, Vol. 14, NO. 1, pp. 107-114,2000 0892-33 10/00 O 2000 Society for Scientific Exploration Remote Viewing in a Group Setting Interval Research Corporation Palo Alto, CA 94304 e-mail: [email protected] 1010 Harriet St. Palo Alto, CA 94301 Abstract- Remote viewing (RV) is a perceptual ability whereby individuals are able to describe and experience objects, pictures, and locations that are blocked from ordinary perception, either by distance, shielding, or time. RV is usually carried out as a team effort, consisting of a viewer who is attempting to describe a target, and an interviewer who assists the viewer in exacting images and sensations from his of her subconscious process. We report a RV experiment carried out at a conference in Arco, Northern Italy, with a class of 24 participants, many of whom were healers and "energy workers." Based on previous work of the authors, great attention was given to creating a feeling of community and coherence of intention within the group during the threeday class. In the fourth of the five sessions of the class, a formal, RV experiment was conducted with class members working in pairs, wherein each person served alternately as viewer and interviewer. Viewers were asked to describe a picture of an outdoor scene, encased in an opaque, sealed envelope, which they would be shown immediately after the session. The interviewer then was directed to take the viewer's sketches and written impressions to the front of the room and rank order the material (from 1 to 4) against the four possible pictures from a preset target package. In this blind-ranking protocol, 6 first-place matches would be expected by chance from the 24 viewers. Instead, 14 first-place matches were achieved. The binomial probability of this outcome is 5 x with an effect size ~ l ( l \ r ) " ~0.64 = Keywords: remote viewing-psi-ESP Introduction 1 The remote viewing (RV) protocol that was developed in 1972 by scientists at Stanford Research Institute has now been in the public domain for more than 25 years. This perceptual processing technique pertains to the acquisition and description by mental means of verifiable information about the physical universe that is blocked from ordinary sensory perception by distance or shielding (Puthoff and Targ, 1976). The authors have many years of experience conducting RV studies, in which effect sizes Z I ( N ) ' ' ~ 0.6 and greater = 108 R. Targ & J. E. Katra are not unusual. We often have attributed this degree of success to the energy and positive expectation that the experimenters bring to each session. This experimental ambiance and communicated expectation was described in detail in a 1990 Parapsychological Association Conference panel, "Increasing Psychic Reliability" (Targ et al., 1991). ESP experiments in group and classroom settings have traditionally had low effect sizes, 0.2 or less. This is attributable principally to a lack of attention, coherence of feelings, seriousness of purpose, and motivation in the group, combined with the use of unselected and untrained subjects and a lack of trial-by-trial or otherwise timely feedback to the subjects (Honorton and Ferrari, 1989). The purpose of the experiment described here was to determine if we could overcome these obstacles and carry out a successful experiment in a group setting with people who were previously unknown to each other. The Arco Experiment For a phenomenon thought in many circles not to occur (Hyman, 1996), we have learned a great deal about how to increase and decrease the accuracy and reliability of RV. Remote viewers often can contact, experience, and describe a hidden object or a remote natural or architectural site based on the presence of a cooperative person at the location, geographical coordinates, or some other target demarcation, which we call an address. Shape, form, and color are described much more reliably than the target's function or other analytical information. In addition to this vivid visual imagery, viewers sometimes describe associated feelings, sounds, smells, and even electrical or magnetic fields. Blueprint accuracy sometimes can be achieved, and reliability in a series can be as high as 70%. With practice, people become increasingly able to separate out the psychic signal from the mental noise of memory, analysis, and imagination. Targets and target details as small as 1 mm can be perceived. Again and again, we have seen that accuracy and resolution of RV targets are insensitive to variations in distance (Targ and Katra, 1998). With this goal in mind, the authors accepted an invitation to conduct a 15hour RV workshop at the 20th International Astra Meeting, called "Rights of Passage," in Arco, Italy (October 12-15, 1999). Astra publishes a widely read metaphysical magazine in Italy and conducts an annual conference on a variety of esoteric subjects in cooperation with residents and city officials of the town of Arco in the foothills of the Italian Alps. We accepted the invitation to introduce a class of 24 Italian students to spiritual healing and to teach them how to perform RV. Outline of the Workshop We had five 3-hour sessions with our 24 students. Everything that we wished to communicate to our students had to be translated into Italian, sentence by sentence. In the first morning session, we described our proposed program Remote Viewing in a Group Setting 109 and introduced the students to the idea of remote viewing and spiritual healing. An overview of the material was presented, together with numerous slides from previous experiments, showing what can and cannot be expected from RV. We discussed the necessity of separating the so-called psychic signal from mental noise. We shared our belief that RV is a natural and widely distributed ability for which everyone, to a greater or lesser degree, has the potential. The emphasis of this session was on how to do the mental processing for real-time RV with immediate feedback. The session ended with each participant doing RV of a "small, interesting object" that the authors had brought for them to psychically observe and describe. This was, of course, not a double-blind trial because the person guiding the students in their efforts knew the object. The purpose of the exercise was to show the students the variety of questions that an interviewer can ask regarding the shape, texture, size, weight, type of material, color, possible use, and so forth, as he or she leads the viewer to look for surprising mental images. After the trial, the students were each given a small opaque paper bag, asked to put a small object into it, and bring it to the next morning's class. The afternoon session was experiential and dealt with meditation, group coherence, and spiritual healing. There was great attention given to building rapport and trust, both between the experimenters and the students and among the students themselves. To achieve this, we conducted a lengthy, guided meditation with music and a guided experience of "energy sharing" among pairs of students. In the third session, the second morning, the students divided themselves into pairs. They took turns being interviewers and viewers for the objects each had brought to the session. This activity also was not a double-blind trial, but it gave the students another opportunity to look for mental pictures that correspond to something outside their experience. We did not want to use pictures for this training, because we hoped to keep their mental slates "clean" for the pictures we would use in the formal experiment in the next session. The fourth session (held on October 14, 1999, from 3 to 5 p.m.) was a formal experiment, described below. The fifth and final session was carried out the next morning. It included a discussion of the experiment and the spiritual implications of psychic abilities. We asked, what do the spiritual healer, the mystic, and the psychic have in common? We proposed that they are all in touch with their nonlocal, interconnected mind and their community of spirit. In the spirit of the conference, we suggested that as we approach the millenium, in every area of human activity, we are experiencing a climax in which science and religion are finally becoming coherent in the exclamation of a single, unified truth. Recent research in areas as different as distant healing and quantum physics are in agreement with the oldest of spiritual teachings of the sages of India, who taught that "separation is an illusion," suggesting that we have an inner knowledge of time and space. In this final session, we observed that the in-flow of information, which is 110 R. Targ & J. E. Katra the hallmark of RV, and the out-flow of intention, which plays a part in facilitating distant healing, are on either side of the quiet mind and the stillness that can arise between them. Perhaps narrowly focusing on the omniscience of ESP is simply a trap that prevents us from discovering who we really are and how we might direct our life's attention. Whenever any one person demonstrates an ability beyond the ordinary, it can be seen as an inspiration to the rest of us, indicating an immense and still largely undeveloped human potential. Experimental Protocol The formal experiment in the fourth session was a demonstration-of-ability test to determine if the students could actually show some RV capability. Before leaving for Italy, we had prepared 24 file folders, each with four target pictures. The 8 x 10 color pictures were carefully selected from the 20,000 Core1 Professional Photos available on a set of 200 CD ROMs, which were made available to us by Dr. Edwin May of the Laboratories for Fundamental Research. The pictures each contained a central focus, such as a mountain, waterfall, lighthouse, windmill, bridge, tall building, ruins of various descriptions, pyramid, trees, coastline, and so forth. Each group of four pictures was carefully assembled so as to have as few overlapping pictorial elements as possible. One picture from each group then was put into an opaque, tamper-resistant envelope, and then the envelope was sealed. These pictures were selected randomly, and then filtered to provide a representative mixture of possible targets to avoid any accidental stacking that could occur if, for example, we had an overrepresentation of waterfalls, or bridges. We keyed each envelope by number to the target folder to which it belonged. To carry out the experiment, the group again divided themselves into pairs. From each pair, the person who was to be the first interviewer came to the front of the large, dimly lit meeting room and was given a sealed envelope containing a picture. Each interviewer then proceeded to elicit from their partner his or her impressions of the picture that was in the envelope. They also could describe their impression of the same picture because it would be shown to them for feedback right after their session. The interviewers then asked their partners to draw sketches and to write down any key words, both of which were to reflect their mental image of the target picture. When interviewers felt that they had a coherent description from their partner viewer, they brought the remarks and sketches to the front of the room and gave their material to one of the two assistants. The sealed envelope was then carefully opened under the front table, out of sight, and the picture inside was randomized into the folder with the other three pictures of its set. Because these pictures had all been used previously, many of them had little wrinkles around the edges; any wrinkles caused by handling in this experiment was not thought to be a factor. The folder then was given to an assistant, who spread the four pictures out on a table. The interviewer was then asked to rank the Remote Viewing in a Group Setting 111 four pictures from 1 to 4 in accordance with their estimation of best to worst match to their viewer's description. Neither of the assistants working with the interviewer had any knowledge of which of the four pictures was the target picture. After the assignment was made, the correct picture was identified by the independent scientist tracking the target pictures that were selected from the target folders. The interviewer then took the correct target picture (regardless of its rank) back to the viewer for feedback. Results The first group of 12 viewers received eight first-place matches (p = .0028, h = .863). The second group of 12 obtained six first-place matches (p = .0544, h = .52). The overall result of the experiment found 14 first-place matches for the 24 students (p = .0005, h = .69), with a 58.3% hitting rate. There were two second-place matches, four third-place matches, and four fourth place matches. In Figure 1, we show the sketch produced by Viewer 1 to finish the RV task. Viewer 1 was a highly regarded Italian energy healer; his wife was his interviewer, and she was known as a psychic practitioner in her own right. Within 1 minute of the interviewer receiving the target picture in its envelope, she returned with Viewer 1's sketch. After seeing the four possible pictures, it took the interviewer no time at all to identify the correct one, an image with pillars. In the illustration, the word cielo is Italian for sky. It is interesting to note that Viewer 1 was in no way limited in his drawing by the edges of the paper he was given. Figure 2 shows a drawing made by Viewer 2, a psychotherapist, who was interviewed by a good friend. This interviewer also had no difficulty choosing the correct picture from the four offered, which was a picture with the domed buildings and cross-hatched windows. Discussion Teaching RV is one thing, but teaching it entirely though a translator seemed like a daunting task because of our belief in the importance of intimacy and coherence in the process. We present this experiment here to function as a possible aid to other researchers who are called upon to demonstrate or teach psychic abilities in a group setting. We did not carry out a double-blind comparison of this approach with other possible methodologies. Nonetheless, what we describe here reflects many years of success in eliciting psi from inexperienced students. It was the success of this experiment that made us feel it was worthwhile to describe our approach. We believe that the success of this experiment can be attributed to several factors. Perhaps most important, all of the participants were self-selected to take part in a RV training program for which they had to pay in advance. Also, the 20 women and 4 men in the class all considered intuition at least a moder- 112 R. Tirg & J. E. Katra Fig. 1. Sketch by Viewer 1 at top, together with actual target picture (lowcr left) and three decoys. Rernote Viewing in a Group Setting 113 I:ig 2 Sketches by Viewer 2 at top, together with actual target picture (lower right) and three dccoy\. 114 R. Targ & J. E. Katra physicians. We further believe it was helpful to have found a way to give the students practice in RV with an interviewer through the use of small objects, which did not contaminating their mental imagery with pictures resembling their target pictures. Thus, we were able to work with "first timers" who actually had some practice in RV. It is likely that the use of large, clear, colorful, easy-to-describe targets was an additional helpful element. Finally, we wish to point out that the effect sizes seen in this experiment are analogous with effect sizes seen in the recently published future forecasting experiment by the authors (Targ and Katra, 1998), and the 36 trial experiment carried out many years ago with six army volunteers at Stanford Research Institute (Targ, 1994). These intelligence officers achieved an overall effect size of 0.63, comparable to the 0.69 seen in this experiment. We consider these results typical for a wellconducted RV experiment. These experiments differed from many of the usual RV cases in that no one knew the correct answer at the time of the experiment. Therefore, this study would be considered one of the clairvoyance type, with only the final feedback providing a possible precognitive channel. Acknowledgements We wish to sincerely thank Dr. Dean Radin for his thoughtful help in the initial design of the formal experiment described here and also for his technical assistance in preparation of the paper for publication. If any measure of success is to be claimed for our approach to teaching remote viewing, an equal measure of credit must be given to our enormously talented and intuitive translator, Giorgio Cerquetti, who conceived and organized our participation in the conference. We also very gratefully acknowledge the generous support of the Astra Meeting, which underwrote the costs of the entire program. References Honorton, C., & Ferrari, D. C. (1 989). Future-telling: A meta-analysis of forced-choice precognition experiments. Journal of Parapsychology, 53, 281-209. Hyman, R. (1996, MarchIApril). The evidence for psychic functioning: Claims vs. reality. The Skeptical Inquirer. Puthoff, H. E., & Targ, R. (1976). A perceptual channel for information transfer over kilometer distances: Historical perspective and recent research. Proceedings of the Institute of Electrical and Electronical Engineers, 64, 329-254. Targ, R. (1994). Remote viewing replication evaluated by concept analysis. Journal of Parapsychology, 58,271-284. Targ, R., Broad, W., Schlitz, M., Stanford, R., & Honorton, C. (1991). Increasing psychic reliability: A panel discussion presented at the 33rd annual conference of the Parapsychological Association, Chevy Chase, Maryland (August 16-20, 1990). Journal of Parapsychology, 55, 59. Targ, R., & Katra, J., (1998). Miracles of mind: Exploring nonlocal consciousness and spiritual healing. Novato, CA: New World Library. Targ, R., Katra, J., Brown, D., & Wiegand, W. (1995). Viewing the future: A pilot study with an error-detecting protocol. Journal of Scientific Exploration, 9, 367-380. Journal ofScientijic Exploration, Vol. 14, No. 1, pp. 115-120,2000 0892-33 10100 O 2000 Society for Scientific Exploration GUEST COLUMN: THE SOVEREIGNTY OF SCIENCE 430 Kennedy Avenue Pittsburgh, PA 15214 The Pronouncements of Science In today's world, most knowledgeable persons concede the authority of science. Since the findings of science are vast, knowledgeable persons, including scientists, accept the pronouncements of science in lieu of a personal mastery of its evidential structure. The "pronouncements of science" are the consensus generalizations expressed by the leaders of science, either as individuals or through their professional organizations or without dispute in most of the relevant textbooks. The pronouncements of science, thus understood, are accepted by persons who wield power in the First World. Being intelligent but often scientifically illiterate, the movers and shakers have little choice but to accept the sovereignty of science in all those areas where science makes pronouncements. In most areas of science, "pronouncements" are broadly based on empirical observations that have been widely and repeatedly made and that are often embodied in unifying theories of great eloquence. In some cases, however, the pronouncements of science may depend on observations that have been made by only a few specialists whose findings are tentatively accepted by the scientific community because of the credentials of those specialists. In other words, science allows itself some laxity in awarding its approval. The Opposition of Science to Psi With regard to the occurrence of psi (psychic) phenomena, however, science makes a clear pronouncement, namely, that such phenomena do not occur (McConnell and Clark, 1991). What needs explanation is why this negative consensus of science persists despite the volume and quality of evidence favoring the phenomena and the outstanding credentials of some of those who vouch for that evidence (Radin, 1997). What demands attention, however, is not the absence of a consensus that psi phenomena occur but the presence of a consensus that they do not! It isto say the least-unusual that science should make a pronouncement on the nonoccurrence of a phenomenon in any area of current observation (as opposed to theory, e.g., Creationism). Moreover, in view of the supporting evidence and the importance of psi, if real, it might have been expected that some 116 R. A. McConnell 1 of the leaders of science would have disassociated themselves from this knownothing attitude of their colleagues. The only living leader who has done so is Nobel Prize winner Brian D. Josephson, Professor of Physics at Cambridge University (McConnell, 2000, p. 364-365). Scientifically pretentious denunciations of belief in psi as a superstitious pox upon humanity usually end with the justification, "There is insufficient evidence to prove the occurrence of the phenomena." This suggests that the evidence, while currently inadequate, would be accepted if it were not deficient. My intention is to explain why the evidence can never be "adequate" unless, psychologically speaking, the modern world morphs to a very different future state. While positive evidence for psi occasionally is reviewed in a respected journal of science, it is usually done by an incompetent critic and, as a result, is generally scientifically misrepresented-often with ideological fervor (McConnell, 2000, p. 359-364). Any realistic examination of the literature will show that the competent spokesmen for the evidence favoring psi are denied access to the respected journals of popular science, such as Scientific American, and to the leading research journals of general science (Science and Nature), while at the same time there is an international organization devoted to the disparaging of "paranormal phenomena." This organization is called the Committee for the Scientific Investigation of Claims of the Paranormal (CSICOP). Among its 77 listed sponsoring Fellows are five Nobel Prize winners in science. CSICOP publishes The Skeptical Inquirer: The Magazine for Science and Reason with a circulation of 50,000. In the past, this magazine has attacked all psi phenomena. Recently, it has softened its attacks but remains generally unfavorable to parapsychology as the serious study of psi. If The Skeptical Inquirer were to adopt a neutral attitude, e.g., by fairly reviewing Radin (1997), it would offend most of its subscribers and risk economic collapse. To understand the subtle role of this journal in blocking the progress of science in this area, it will be necessary to go back a long way. Religion and Psi From the beginning of consciousness, man has encountered phenomena and situations whose cause or significance was not apparent to him. Fearing the unknown as a threat to his existence, he invented reassuring explanations to calm his mind and allow him to think constructively or even enjoyably about his immediate activities. Many of his feared experiences seemed related, and these he lumped together and explained by a variety of otherworldly myths. This led to organized religion, which reached the peak of its political power among Western European people during the Middle Ages. As men become more sophisticated in their thinking, beginning in the 16th and 17th Centuries, religion was rejected in favor of materialist science, The Sovereignty of Science 117 which, however, was unable to encompass certain of the nonmaterial aspects of experience. These aspects, therefore, remained in the domain of religion. By the time of the 18th Century Enlightment, materialist science was in its ascendancy, while the beliefs of religion were increasingly recognized as fraught with self-contradictions. An intellectual movement was set afoot to dispense with religion. In our own time, the majority of educated people in the First World reject most of the superficial, irrelevant aspects of religion, which however continue to hold sway among the less educated. This has led to the interclass religious conflict in which we are embroiled today. The author believes that the ultimate solution for this dilemma is for the better educated people to recognize the reality of psi phenomena, and, under the guidance of scientific method, to make a place for them in their own belief systems. Then, having regained credibility with the masses, they might hope to eradicate truly superstitious beliefs. This is what your author believes should be the long-range goal of parapsychology. Science and Psi In this paper, the author will explain how the materialists, in their struggle to be free of anything smacking of superstition, have adopted a defensive ideology that denies reality and betrays their own intellectual principles. The author has subscribed to The Skeptical Inquirer for many years and regards CSICOP as the foremost instrument of the arch materialists in their fight against superstition. In their zealotry, they have opposed those few scientists today who are struggling with the difficult, perhaps impossible, task of discovering the nature of nonmaterialist phenomena and of explaining them in the materialist language that we speak. These visionaries seek no less than the reconciling, through the use of the scientific method, of the material and the nonmaterial: of science and religion! Parapsychologists have no quarrel with the antisuperstition program of CSICOP and The Skeptical Inquirer, but they are frustrated by CSICOP's implicit denial of the traces of psi phenomena that are all around us. Spontaneous Psi: Its Characteristics Psi occurs spontaneously, as well as in the laboratory, and the evidence for its spontaneous occurrence (as well as for its laboratory occurrence) overwhelmingly affirms its reality (McConnell, 2000). In this paper, spontaneous psi is where we shall focus our attention. From all accounts, it would appear that spontaneous psi occurs more often among less educated persons, and this is commonly used as an argument that the evidence supporting psi must be fallacious. On the other hand, it is possi- 118 R. A. McConnell but are recognized more easily by the less educated who are not subservient to the sovereignty of science and are under little social pressure to accept the common wisdom of science, according to which, psi does not occur. Spontaneous psi occurs in many forms and with all degrees of intensity and frequency, depending on the person. The most common form of psi is extrasensory perception (ESP) in which a percipient acquires information through other than sensory-motor means. Spontaneous ESP is most obvious when it reveals information from a distance or about the future. How Science Ignores Psi If, indeed, psi occurs equally at all levels of education, why is its reality not obvious to the world? The answer is multifold. For most people, ESP occurs only as a weak and infrequent effect, readily deniable as coincidence if one so wishes or readily recognizable as anomalous if one has low evidential standards. Also, those professional psychics who make a living using psi, which is at best a thoroughly undependable phenomenon, must often supplement their performance by cheating to maintain their business reputations. This leads to the supposition that all exhibitions of spontaneous psi are mistaken or fraudulent. Add to this the treatment of psi as entertainment on television. Powers far beyond those scientifically demonstrated to be possible are daily fare on television, enhancing the disbelief in psi among educated people. Some parapsychologists believe that psi occurs more readily in minds that have a limited ability to think analytically but whose mental forte lies in their right cerebral hemisphere, i.e., in those who are artistically inclined rather than in those who are logical by nature. If so, this merely strengthens the prejudice of scientists against the reality of psi. The real puzzle is why ESP is not universally acknowledged when it occurs, as it occasionally does with some people, as a dramatic and undeniable effect. For high-powered psychics who happen to be highly educated, there is an escape hatch from the embarrassment they might otherwise cause to their professional-class friends. To avoid cognitive dissonance, these individuals usually join their less educated fellow psychics and quietly assign psi to a realm of spiritual belief that is commonly conceded to be outside the realm of science. Thus, as regards psi experience, religion serves two purposes. For those deeply devoted to an organized sect, psi may be considered either as a gift of God or as a manifestation of His presence. For those rare others, seeking escape from their own undeniable psi, religion provides a haven to put psi safely out of mind. In both functions, psi strengthens the legitimacy of religion. Quite aside from those who recognize themselves as psychic, the existence of a nonphysical realm-call it what you will-must ultimately strengthen the The Sovereignty of Science 119 Rewriting the Textbooks There is another and more encompassing explanation for the failure of the Scientific Establishment to examine carefully the empirical evidence for the occurrence of psi. To understand the undying opposition of scientists to psi, one must consider the hypothetical consequences of psi's universal acceptance. All general textbooks of psychology and physics would have to be rewritten. For physics, this might require no more than an acknowledgment that there exists a nonphysical realm with which the physical realm can interact, both spontaneously and experimentally. The exploration of these interactions would attract the interest of theoretical physicists even while experimental physicists might be frustrated in their misguided attempts to demonstrate psi as though it were a new form of electromagnetic radiation. In psychology, the fallout from a universal recognition of the reality of psi would be catastrophic. Most present and past psychological research would be recognized at once as trivial. Much of the rest would lose its claim to validity because of the probable confounding effects of ESP or of psychokinesis (e.g., by the direct mental action of "wishing" on measuring instruments). Experimental psychology as now practiced would be destroyed as a scientific enterprise. Psychiatry, for its part, would have to go back and start from the beginning. A Revisionist Worldview The foregoing practical considerations do not, however, come to grips with the philosophic heart of the matter, namely, the shattering of the materialist worldview of physics. This worldview governs not only the physical sciences, but also psychology and postmodernism, and it justifies the aggressive behavior of much of the First World's ruling class. It is not possible for the average layman to understand what the acceptance of psi would mean to a thoughtful scientist. The acknowledgement of the reality of psi by physics departments, implying as it would, philosophic dualism, would be comparable to the denying by Rome of the reality of God. Any attempt by a thoughtful scientist to reconcile the established facts of parapsychology with his understanding of his philosophic commitment to his profession would encounter an emotional block. Direct, uninferable knowledge of the future, for example, which is accepted matter-of-factly by parapsychologists, would be horrifying to the scientist who accepts the materialist view of reality. Conclusion Since the sovereignty of science extends to most persons who wield power in the First World, including most of those in middle management, its collapse could result in social chaos. The materialist scientist's adamant denial of the 120 R. A. McConnell existence of psi can perhaps best be explained by his fear of the consequences that might follow in the event of its acceptance. References McConnell, R. A. (2000). Jo~rrcle irljr~lzf~l.re-rentrflt stlrd> of the rloo~il\cia\ liicrcrturc.. Washto A ington, DC: Scott-Town\end. McConnell, R. A,, & Clark, T. I .( 1 99 1). National Academy 01 Sciencc4' oplrlion on parapsyS chology. Jounlnl of thc Anrcrrt L ~ HS o c r e t ~for P S Jhicrrl Ke\cjrrrtIr, 8.5, 3331-365. ~ Radin, Dean I. (1997). 7 1 1c~ otzcc rolr s utlrllrr.\cJ: 7 ' 1 ~ ~rc~~ltifl( \c t111t11 of 1)')) hrc / ) / Z P I I O ~ ~ I ~San I . c I~C Francisco: Harpercolllns In 1943, by visiting Harvard's library, Dr. McConnell ascertained that ESP occurs, although beyond explanation by known physics and psychology. After entering parapsychology full time in 1947, he devoted his efforts primarily to the question: "Why do scientists reject ESP?" This paper gives the answer to his 50-year search. To the pursuit of this question, Dr. McConnell brought an unusual breadth of experience. He holds a doctorate in physics. During World War 11, he led a radar development group at M.I.T. He is a life senior member of the Institute of Electrical and Electronics Engi~~cers. In 1957, he was the founding president of the Parapsychological A\sociation, which was admitted to aCfili;ltion with the American Association for thc Advanccrrlent of' Science (AAAS) in 1969. He is a Fellow of the American Psychological Society, Research Professor Eralcn itus tsf Bk~logical C ~ ~ L I C Ca, Fellow of the AAL2S S ~irid Journal of Scientific Exploration, Vol. 14, No. 1 , pp. 121-1 38,2000 0892-33 10100 O 2000 Society for Scientific Exploration BOOK REVIEWS Best UFO Cases-Europe by Illobrand von Ludwiger. Las Vegas, NV: National Institute for Discovery Science, 1998. 173 pp. $19.95, paper. ISBN 09666077-0-8. Available from: The National Institute for Discovery Science, 1515 East Tropicana Avenue, Suite 400, Las Vegas, NV 89 119. Although there is a plethora of books in English about unidentified flying objects (UFOs), the majority deal with the North American scene, with several covering phenomena in the United Kingdom and other English-speaking countries. Best UFO Cases- Europe helps fill a void in the libraries of investigators who are interested in central Europe. The author, Illobrand von Ludwiger, is well qualified to present this wide array of physical evidence, having been employed as a systems analyst at Daimler-Benz Aereospace AG-DASA in Ottobrunn, Germany, since 1964. He has participated in rocket launches (1966) as a member of the European Launcher Development Organization (ELDO) and was the project manager responsible for the simulation of traffic for new transportation systems. He has studied UFO phenomena for almost 30 years, and he founded the MUFON UFO Network, Central European Section, in 1974, now named the Society for the Scientific Exploration of Anomalous Atmospheric and Radar-Phenomena, MUFON-CES, Inc., which has employed groups of technically trained and equipped field investigators. Best UFO Cases- Europe begins with three forewords written by John F. Schuessler, Bruce Maccabee, and Richard F. Haines. Each provides different views of ufology in general and of this volume in particular. As Schuessler points out, this work "is especially valuable because it provides data showing a level of UFO activity comparable to what has been going on over the United States, South America, Australia, Japan and other parts of the world." The professional reader of Best UFO Cases- Europe will find much to analyze, but the book will interest even casual readers. Chapter One provides a historical look at UFO cases in central Europe from the 16th through the 18th centuries, as well as more recent phenomena such as the so-called Foo Fighters of WWII and "Ghost Rockets." There are several previously unpublished cases presented. Chapter Two presents five cases involving well-defined UFO shapes that have been reported, mostly from the 1970s through the 1990s. Readers discover that the same object shape names were reported in central Europe as were reported in America. The huge number of triangular-shaped object sightings merits its own chapter, and thus Chapter Three begins with the Belgian flap of November 1989 through April 199 1, when at least 3,500 separate UFO sightings were reported. All of these were seemingly of the same large, round-comered, triangular object. Another 13 new cases are also reviewed, three witnesses of which were MUFON-CES members. Numerous 122 Book Reviews helpful drawings in both color and black and white are included. In Chapter Four objects with complex structures are reviewed (eight cases). These cases are important because such highly detailed yet unusual objects do not appear very often. They become more persuasive when citizens in one part of the world independently report seeing the same or similarly shaped objects. The author gives several such examples. The two main sections and 10 subsections of Chapter Five are perhaps the most troubling in this book, for they deal not only with landed UFOs but also their alleged "occupants." The author begins with four cases (1914 to 1954) from Germany and France in which none of the witnesses knew each other, yet the UFOs and "entities" still had the same basic shape. These "entities did not conform to the description of UFO occupants as handsome humans with long hair then in vogue." Three more occupant cases are presented in detail, with emphasis given to after effects, various physical and psychological tests given, and artist drawings and photographs of witnesses and locales. Chapter Six is devoted to a single CE-2 nighttime case in Vaddo, Sweden, on November 11, 1956. This case involved two men who saw a silent, flattened, elliptical, metallic sphere approach their car and illuminate the entire surrounding countryside as bright as day. The object seemingly caused their engine to sputter and die and their headlights to go out as it landed directly ahead of them in the roadway. The UFO (estimated at about 8 m in width) was wider than the two-lane road. After some 10 minutes, it took off rapidly. The two men searched the landing area and found that the grass had been flattened on each side of the road and also discovered a small chunk of metal the size of a matchbox. It was still very hot to the touch. Details of the eight separate laboratory chemical analyses that were performed on this sample are given, along with drawings and macrophotographs. It was found to consist of pulverized wolfram (tungsten) carbide and cobalt. Three major photographic cases (Greifswald Lights of August 24, 1990; Guiseppe Lucifora photos of June 19, 1987; Rudi Nagora photos of May 23, 197 1) are discussed in Chapter Seven. Thirty color plates present these mostly daylight disc images. Relatively little technical information is included about them. I consider the major contribution of this book to be found in Chapter Eight, "Traces of Unidentified Flying Objects on Military Radar Devices over Central Europe." Within these 34 pages of text and radar plot diagrams are found numerous examples of discontinuous flight paths, very sharp corners, ultrahigh-velocity flight, and other anomalous data. Von Ludwiger also includes the almost mandatory discussions of false returns and the various kinds of radar in use today. He makes the intriguing statement here: "During daily operating procedures, such points or short tracks are interpreted as disturbances and ignored. Every employee of military airspace control is familiar with them. But nobody can satisfactorily explain what they are. They are considered to be radar or the computer program mistakes or some kind of atmospher- I Book Reviews 123 ic phenomenon. They are not reported because of fear of being reproached by colleagues and superiors for (having an) insufficient working knowledge of the radar system." In short, European air-traffic controllers behave the same way as controllers do elsewhere in this regard. Also provided are almost 30 different radar traces of highly provocative aerial flight paths by UFO and unidentified aerospace vehicles. Nonetheless, not even America's F-117A Stealth Fighter can perform most of the maneuvers shown here. Anyone who claims that no radar traces of highly anomalous flight behavior are available should study this chapter in detail. One drawback is the lack of all background technical information that one would need to fully understand these traces. Of course, each radar contact case could fill a book of this size. In Chapter Nine, we are treated to a brief but informative discussion on magnetism and selected details of a "Fluxgate" magnetometer designed and built by two MUFON-CES members. Sample X, Y, Z channel output tracings are given for two 24-hour periods but are not related to UFO phenomena. The following 14 pages (Chapter Ten) are devoted to various data catalogues and statistical analyses, with special emphasis given to physical interactions between UFOs and the local environment. Results of a study of electromagnetic and "gravity" cases by Adolf Schneider is presented (1,319 total events for the period 1930-1982); these data are broken down in many useful ways for the interested reader. Donald Johnson's work published in the Journal for UFO Studies (1983) is also included for comparison. Three other cases are reviewed in moderate detail as well (August 13, 1970; August 14, 1973; and January 8, 1981). Chapter Eleven is, in my view, the second-most valuable contribution of this book. It presents nine proposed explanatory hypotheses for UFO phenomena, although it can be argued whether some actually nestle together in fewer categories. I suggest that only four basic groups are needed: (1) extraterrestrial origin (ET visitation, parallel universes), (2) intra-cranial origin (psychological, paranormal, psychic projection), (3) environmental origin (tectonic strain, earth lights), and (4) exotic origin (interdimensional, the author's "projector theory," which deals with hyperdimensional, "transcendental" manifestations, time travel). Of course, how one subdivides this theoretical "pie" is actually a matter of interpretation, definitions, and personal taste, rather than hard-andfast science. Nevertheless, such discussions are valuable because they force readers to make explicit what they expect of the data collected about UFO phenomena. In the final chapter, von Ludwiger perhaps may be forgiven his personal interpretations of earlier occupant evidence presented in which he appears either to go beyond the evidence or to disregard still other data. For example, he states, "The fact that the UFO occupants are very human-like speaks for the Earth as their home planet. Sometimes their appearance doesn't differ from that of a European, and witnesses will have heard that these occupants also 124 Book Reviews UFO occupants were from a time that is still to come. In that case they must avoid any close contact to the people who are born in a different time period, because all actions could influence future events and therefore undesired reactions to the chain of events for the occupants" (p. 157). The author seems to have overlooked a large and continually growing number of cases describing (a) nonhumanoid creatures and (b) narrative interviews with people from around the world who claim to hear occupants speak in no earthly language but with other strange utterances. He also invokes the well-known "categorical imperative" from science fiction that one race should not interfere with the development of another, even itself seen from the future. These few difficulties notwithstanding, Best UFO Cases-Europe is a serious and positive contribution to the literature on a variety of UFO phenomena and is likely to remain so for many years to come. Its rather minor problems with English grammar, which is not the author's native language, are far outweighed by its emphasis and clear presentation of hard data from Central Europe. Illobrand von Ludwiger is to be commended on collecting, translating, and presenting these data for the benefit of many others around the Englishspeaking world. The National Institute for Discovery Science is also to be commended for bringing this work into print. Richard I;: Haines 325 Langton Avenue Los Altos, CA 94022 The Last Laugh by Raymond Moody. Charlottesville, VA: Hampton Roads Publishing Company, 1999, paper. 210 pp. This well-written and entertaining book bears the subtitle A New Philosophy of Near-Death Experiences, Apparitions, and the Paranormal. The author proclaims that this book is an obligatory addendum to his original celebrated work Life After Life. As such, it consists of the thoughts that commercial publishers edited out of his works in a 20-year period. Indeed, the author claims that the publishers over the years hacked out so much of his work that he does not recognize his work anymore. He objects, for example, to covers of books stamped with untruthful exclamations such as "Scientific Proof of Life After Death!" Although he objected in various ways to such extravagant claims about his work at the time, he felt it nonetheless was important to have the work published, especially because publishers were not interested in the work without the addition of such hype. These publisher's tactics were a constant headache for Moody-and a continuing source of embarrassment. In fact, the author frequently claims in this book, and elsewhere, that the idea of proving by appeal to paranormal phenomena, scientific or otherwise, any form of life after death is a waste of time because it cannot, in his view, be done. He is emphatic, however, in insisting that it cannot be disproved either. Book Reviews 125 Moody's skepticism on this score embraces the view that one can neither prove nor disprove the existence of life after death by appeal to the paranormal, and this is the story he has not been able heretofore to defend in writing. As he says, "The publishers rake in the cash, and I'm the one left to answer the critic's objections-the very objections I had anticipated and resolved in the passages editors cut out of what I wrote, or altered" (viii). To that end, the author claims that, in issuing this book, he is declaring null and void Life After Life. He will accept responsibility for Life After Life only insofar as it is read and interpreted in the broader context provided by this new book. By way of the implications of the subtitle of the book (i.e., A New Philosophy o f Near-Death Experiences, Apparitions, and the Paranormal), Moody declares early that this book challenges settled thinking about the supernatural and that it is, in fact, a secession from encrusted discussions on the paranormal. On this point, he is intent on dislodging three distinct sects of true believers who have long dominated learned discussion about these things, and he seeks to do as much by developing an alternative theory about the nature of the paranormal. The three distinct types of true believers are parapsychologists, members of CSICOP (i.e., the Committee for the Scientific Investigation of Claims of the Paranormal.), and fundamental Christians. After stating that it has been a grave mistake on the part of the media to typecast him as a parapsychologist, he proceeds to characterize parapsychologists as people who typically: masquerade as scientists, alleging they can prove mind-reading, prophetic abilities or life after death by laboratory techniques or, more generally, by rational procedure. In fact, parapsychologists are pseudo-scientists, which means that they espouse a system of methods and assumptions they erroneously regard as scientific. (ix) He classifies them with rhapsodists, pipe dreamers, lotus eaters, and wool gatherers; they are all pleasant enough, but their basic assumption is seriously in error. Their basic error, of course, is to think that there is some way of scientifically establishing their claims of the paranormal or of life after death based on the paranormal. He also thinks that those groups who think they have scientific proof of the falsity of the paranormal (groups such as those humanists who are members of CSICOP are equally problematic. Of them he says: Many self-styled skeptics about the paranormal join a fringe social movement which advertises itself as a scientific organization while at the same time underhandedly representing itself as a para-law enforcement agency. Let me clarify, parenthetically, that I am not making this up! The members of this social movement call themselves the sigh-cops.. .. For now, suffice it to say first, that sigh cops aren't skeptics but believers in a particular ideology about what knowledge is and how it is acquired, and second, that there are good reasons for believing that their ideology is mistaken. The Last Laugh is an example of philosophical skepticism about the paranormal. I push skepticism far beyond the limits sigh-cops set for their inquiries. By comparison to mine, theirs is a weak-kneed and wimpy approach. (x) 126 Book Reviews In addition to the parapsychologists and the "sigh-cops" (both of whom are wrong for thinking that one can prove or disprove anything at all about the truth of paranormal experience), we have a third group of true believers, the fundamental Christians. This latter group, as one might expect, garner's the lion's share of Moody's mirthful but deadly serious attack on true believers. Of this third bickering batch, he says: They are the goshawful deadfannies, stiffs, bores, nuisances, uptight dogmatists, broken records, and wet blankets, the fundamentalist Christians, Religious Right, Bible Brigade, "JAY-zus"-Sayers, Brimfire and Hellstoners, Swaggartists, Falwellers, Bakker-Boosters, Pat Robertsonians,or whatever you would like to call them. Moody claims that all three groups are full of nonsense and that The Last Laugh demolishes all three of these standard approaches to the paranormal and erects a better, more comprehensive and pragmatic system of thinking in their place. With regard to the latter, his positive approach is predicated on the thesis that while one can neither prove nor disprove any claims about the paranormal or life after death, the value of the paranormal and claims about it fall squarely into the realm of entertainment and that therefore, he advocates the view of "playful paranormalism." Under this new view, the paranormal is not to be made the battleground in metaphysics, but to be enjoyed and played with joyfully for whatever it might inspire. Apart from whether the paranormal has any epistemic significance in seeking the truth about life after death, for Moody there is something inherently charming, entertaining, enchanting, and funny about it all, and we should not look for anything more in it than a source of fun and entertainment. More on this shortly. The chapters of the book include the following: 1. The Experience of Dying; 2. Play and the Paranormal; 3. Breaking up the Logjam: Unriddleing the Controversy about the Paranormal; 4. Miracles, Meanings and Merriment; 5. Believing the Unbelievable Believably; 6. Knowing the Unknowable; 7. Classifying the Paranormal; 8. Justifying the Paranormal or Even the Study of It; 9. The Rhetoric of Dysbelief 10. The Only Way a Serious Study of the Paranormal Will be Legitimatized; 11. A Treasure Chest Waiting to Be Opened; 12. Coming Full Circle: Back to the Subject of Life After Life, 13. Having the Last Laugh. At every turn in the book, Moody criticizes, for what seem to be altogether persuasive reasons, the mental set and practices of the "sigh-cops" and the fundamental Christians. Indeed, it is difficult not to smile when his psychological characterizations hit squarely on the target in expressing his thinly veiled disdain for the epistemic and outdated ideology parading itself as serious science in the practices of the "sigh-cops." A serious critic of the book will note in passing the humorous ad hominems, but then again, Moody's thesis is that ruling against ad hominems seems to presuppose we should take all this very seriously. That is, according to Moody, just the sort of thing we should not do Book Reviews because it suggests that we are committed to a logical analysis of things in the interest of proving something about the metaphysical import of the paranormal. Moody has lost his patience with what he sees as a fundamentally worthless debate about the significance of the paranormal for understanding the nature of human nature. He laments the fact that this debate has actually detracted from a relaxed examination of, and participation in, the paranormal simply as a source of entertainment and edification. For example, when describing the role of the paranormal discourse vis-a-vis its epistemic import for understanding human nature, he declares himself a playful paranormalist, and says: "Dysbelieving" about the paranormal is a pastime for shut-ins and stay-at-homes; it is the paranormal as seen from an armchair. Our society is caught up in what playful paranormalists categorize as a couch-potato model. Viewers sit passively on a sofa and watch television panel discussions about the paranormal. There are a couple of people on the panel who describe their own, personal paranormal experiences of perimortal visions, apparitions of the deceased, dreams that came true, or whatever. Then two or more "dysbelieving" experts slug it out before the cameras, pretending to explain what the paranormal experiences means, i.e. whether or not they "prove" life after death, precognition and so on. Viewers are supposed to make up their minds by evaluating what the panel of experients and "dysbelieving" experts say. "Dysbeliever" experts aren't expected to be able to perform the paranormal, or to be able to enable others to experience the paranormal, as Ancient Greek experts were. Playful paranormalists say, to change the study of the paranormal for the better, change the model of participation. Operating under the unific principle that the paranormal is entertainment, it is possible to symphonize anomalistic psychology, social history, and clinical know-how to enable people to have their own, first hand paranormal experiences. Then they will be in a better position to make up their minds about what such experiences mean. That would be to resolve the pivotal dilemma. (1528.) It appears that Moody believes having one's own paranormal experiences, and doing whatever is required to have them, will provide more illumination on their significance than any objective inquiry. This is not at all unlike the religious mystic's claim that one cannot, by the light of reason alone, prove the existence of God. If one lives as the mystic does, however, then there is a good chance he or she will have the experience of God. In that experience, the need to prove the existence of God dissolves and passes as a juvenile, futile, and worthless activity. Of course, Moody does not say explicitly that one will know or see the metaphysical significance of the paranormal as a legitimate and reliable source of belief in some form of life after death. Nonetheless, I sometimes suspect it is implied in the thesis of playful paranormalism. In addition, it is not far removed from the epistemological thesis that there are some things we know and are justified in believing even when we cannot say precisely what we know or how we are justified in believing it. Whether one should accept this Wittgensteinian thesis or not is a long story not to be taken 128 Book Reviews phers have argued the same basic view-if not about the experience of the paranormal, then at least about experience in general. In a continuing set of comments clearly revealing what playful paranormalism amounts to, he says: Many of those who study the supernatural are chronically distressed because neardeath experiences, ghosts, premonitions and the like are spontaneous and unpredictable happenings that cannot be reproduced under set conditions conducive to scientific examination. So investigators are reduced to sifting through a rubble of reports made sometime after the purportedly paranormal occurrences themselves - retrospective narratives the scientific skeptics belittle as anecdotes. If visionary reunions with the departed, perimortal visions, or apparent foreseeings of the future are what they purport to be, they would involve transaction between ordinary reality and what presumably would be other, alternate levels or dimensions of reality. Science has been extraordinarily successful in part, because of its steadfast and commendable determination to confine its deliberations to this ordinary reality in which we find ourselves. However, the performing arts routinely and reliably effect transactions between ordinary reality and intriguing alternate realities. Playgoers, moviegoers, and concertgoers, for example, regularly are transported, still in their seats, into seemingly different realms of being. They are made to feel that they are in the midst of an entirely different order of things. So by modeling themselves partly on the performing arts, playful paranormalists can entertain the prospect actually of reproducing experiences or phenomena that, when occurring spontaneously, often are deemed paranormal. (p. 153) In the end, some of us are probably a bit more sanguine about scientific or empirical proof for some form of personal survival after death, even if we have not had such paranormal experiences. For that reason, we might remain difficult to persuade that there is no possible way to empirically confirm, via appeal to the paranormal, some form of personal survival after death. In other words, for those of us who agree with Moody's general characterizations of the three classes of true believer, there still may be some lingering suspicion that his new theory of the paranormal may be predicated on an assumption that needs a bit more persuasive defending, namely, that it is a waste of time try to prove empirically in any way that some essential aspect of human personality sometimes survives bodily death. His concept of "proof' could be discussed a bit more fully in this regard as well. Otherwise, the claim that it cannot be done could turn out to be another form of the "true believing" that Moody so richly criticizes. Even so, the interesting epistemological point here may be that there is nothing logically contradictory between knowing that something is so by directly experiencing it and knowing that something is so because one has empirical proof of it. At any rate, readers should see nothing wrong with Moody's suggested playful paranormalism as a possible source of reliable belief formation. Nonetheless, the evidence for the reliability of such beliefs, if they are to be items of public knowledge rather than religious belief, require some empirical confirmation. In the absence of the latter, we could all agree that there are things some people know privately that other people do not know and cannot Book Reviews 129 be expected to know. This is so simply because some people have experiences that others do not have-and perhaps never will. But this is a long story for another time. In sum, this book is a fun read intended for a literate and large audience. It is replete with wonderful insights and on-target assessments from a person whose original playful paranormalism has contributed so much to an important discussion on human nature, a discussion that might never have taken place so forcefully otherwise. For that, we should all be grateful-even if only for the book's entertainment value. Although one could disagree with Moody's major points, there is much to recommend in this book to anybody interested in the topic of the paranormal and the merits of discussions on it. Robert Almeder Department of Philosophy Georgia State University Atlanta, GA 30303 The Discovery of the Cold Fusion Phenomenon by Hideo Kozima. Tokyo, Japan: Ohotake Shuppan, 1998.370 pp. $42.00 (in the USA). ISBN 4-87 186046-2. Hideo Kozima's remarkable book is the first textbook describing cold fusion phenomena. With its more than 400 references and 70 diagrams of experimental results, thorough readers would have difficulty supporting the contention of university physicists that no such phenomenon exists. The author is careful to explain in the introductory chapter how cold fusion is a term based on misunderstandings in early work. Many of the phenomena do not involve fusion, but reactions with neutrons and protons within the solid material in which most of the phenomena occur. The book consists of 18 chapters. Only 10 of them explain experimental work. An usually large number of chapters, four in all, are about the author's theory and its detailed numerical application to the varied phenomena of socalled cold fusion: the neutrons, the tritium, the heat, and the gamma rays. (Incidentally, he does not mention x-ray emission from electrodes, which have been reliably reported.) Kozima's idea-and ideas of this kind are at the cutting edge of present theories of cold fusion-is connected with neutrons inside the solid lattice. His contention is that there are trapped neutrons. They originate in the atmosphere by the interaction of cosmic rays with nitrogen. They arrive on a solid lattice at about lo2 cm-2 sec-l, and thereafter they can take part in a large number of nuclear reactions inside the solid. In the later chapters, Kozima works out what would be the number of neutrons per cc to be consistent with the results that he examines and comes to the conclusion that, by and large, consistency is reached if the concentration of trapped neutrons is between lo8 and 1013 per 130 Book Reviews The first four chapters are introductory and rather light weight. They talk about the infamous ERAB (Energy Research Advisory Board) Report in which the cold fusion researchers 1989 work was interrogated by a number of scientists appointed by DOE (U.S. Dept. of Energy) in a style that would suggest a prosecuting attorney's examination. They touch on the general idea of catalysts and how enzymes react in the body (the appropriateness of some sections here in Chapter 3 seems rather doubtful). Chapter 4 treats nuclear fusion reactions in a classical sense. In Chapter 5, we begin to take off and fly with a discussion of the rediscovery of the cold fusion phenomenon in modern times. The author presents Fleischmann, Pons, and Hawkins (1989) as the discoverers. Nonetheless, the book also makes clear-in later chapters-that several papers reported nuclear reactions in solids before Fleischmann, Pons, and Hawkins. Most remarkable of all, and something new to the reviewer, Kozima quotes a recent book by Kushi (1994) in which a U.S. Army report from the Material Technology Laboratories Report of 1978 is described. This report concerns Energy Development from Elemental Transmutation in Biological Systems. It is said to have validated the work done up to that time as proving the transmutation in the cold and also the production of nuclear energy. Other works carried out before that of Fleischmann and Pons include a study at U.S. National Laboratories of Nuclear Reactions, conducted by passing high currents through wires (producing neutrons), and the work of Borghi in 1943 in which neutrons were produced by a passage of high currents through a klystron. Chapter 6 has the essence of the phenomenological descriptions for systems involving deuterium, and Chapter 7 reports similar material for hydrogen-containing systems, although it is far less in extent. In Chapter 8, the idea of thermal neutrons is considered in a general manner. The summary of the experimental data is given in Chapter 9. Then, in Chapter 10, facts in respect to biotransmutation are given, and Kozima implies that it is a general phenomenon in nature. The ticklish subject of reproducibility is presented and discussed. The facts here are clear. If an investigator tries to start a cold fusion experiment on any given day, there is about one chance in five that he or she would see a positive result. If one is willing to wait a few days and try again, a successful experiment may be seen. At the same time, each of the results discussed in this book has been repeated many times in many laboratories in many countries. Therefore, the results are repeatable, but not reproducible in the normal sense. Scientists will have to get used to looking at phenomena like this. Chapter 11 contains the main testing out of the trapped neutron model. The author produces a noncontroversial equation for the rate at which the reaction of neutrons occurs but has an adjustable parameter, that is, the concentration of neutrons in the solid. As stated above, most of the phenomena can be accounted for numerically if the concentration of neutrons in the solid is between 1 0 ~ - 1 0 'cc-l. It is diffi~ Book Reviews 131 cult to understand how the concentration of neutrons in a given solid would vary so much. In fact, if one studies the table of all the results that have been matched, the range to get a fit is much greater, more like lo2-1013. The author is not forthcoming in commenting on this range, he seems to have an easygoing attitude toward acceptability. By Chapter 12, the reader is immersed in the trapped-neutron theory, and it is compared with various other theories using discussions of the corresponding Mossbauer effect and the role of the electrolyte. Chapter 13 further develops seven other theories and how they compete with the trapped-neutron theory; Chapters 14 through 18 are "postscript chapters." The book really ends after Chapter 13. Chapter 14 is about the energy crisis and how cold fusion might solve it. Chapter 15 is a general chapter about revolutions in paradigms. Chapter 16 presents the views on the field of a number of Japanese scientists. Chapter 17 is about symbols and units, and Chapter 18 is the reference list. To call this a book about cold fusion is perhaps too much. It is particularly oriented toward a presentation of the author's theory. Thus, the typical examples of the reactions of trapped neutrons with the constituents of the lattice are: n +$ ~= i ~ e ( 2 . 1 i MeV) E + t(2.7 MeV), E. + d = n(E1)+ d ( "), d ( ) + d = ;He(0.82 MeV) + n (2.45 MeV) + n ( E) E t ( ~ + d = lHe(3.5MeV) +n(14.1 MeV) + E , ) The product particles of these trigger reactions create higher energies than does the thermal reaction and can induce successive nuclear reactions (i.e., breeding reactions). In a qualitative way, the trapped-neutron theory explains a great deal. Once one is convinced that free neutrons are inside the solid, one can see that several transmutation reactions might well take place and produce energy. The major problem is providing convincing evidence for the large number of trapped neutrons that the theory demands. It is difficult to measure in an independent way, and the author uses it as an adjustable parameter. Were Kozima able to establish agreement with the experiment at, for example, lo7-lo9 neutrons per cc, the reader might be able to believe in the model and swallow the discrepancies in the concentration of cold fusion present in palladium. Much greater ranges are needed to obtain a fit, however, and one has to ask why. Is this the origin of the famous irreproducibility? Could it be that various pieces of palladium have had various trapping times for neutrons? It is difficult to see that neutrons from the atmosphere, over the several years in which most of the pieces of palladium have existed, could build up to the necessary values. Thus, lo2 neutrons per cmv2and sec-' implies the need for lo4 years to build up 1013 neutrons cc-' (even if all were trapped). Another issue that raises doubts about the model is that Kozima always Book Reviews stresses the importance of LiOD (or LiOH) electrolyte and the diffusion of lithium into the palladium. It is true that lithium does this, as was shown by Oliver Murphy at Texas A&M in 1990, but other works that have used sodium or hydrogen as cations in the electrolyte, and these have lead to cold fusion too. Other phenomena that the trapped-neutron theory would seem difficult to accommodate are those observed by Chien at Texas A&M in 1992. He found that whenever he added fresh D 2 0 to, LiOH, the production of tritium stopped and then started again spontaneously after some hours. Correspondingly, it is not easy to see why the potential of the electrode alters the rate at which tritium is produced on the neutron theory. Finally, the impact method of provoking nuclear change-little known but now verified-does not have an obvious interpretation in terms of neutrons. Nevertheless, one feels that Dr. Kozima has provided a useful text by placing his attention on neutrons. The details need further working out-particularly the origin of the neutrons and the concentration that he has to assumebut his work clearly strengthens the neutron case. It is increasingly necessary to consider the sociology of this new phenomenon, which has been lurking in the literature for more than 50 years but came to prominence in the 1990s. There are now more than 2,000 positive papers in the literature written after 1990. It is scandalous that one still has to go to specialist journals to obtain acceptance for publication (i.e., the "establishment journals" of physics and chemistry still refuse to publish cold fusion papers, and the U.S. patent office will not accept patents for devices based on cold fusion). This is a historically important fact because it indicates a frozen physics. New phenomena may not be accepted if they disagree with the present paradigm. Thus, the importance of Dr. Kozima's book goes beyond his providing a compendium of cold fusion facts with an interesting attempt at interpretation. It should act as a clanging bell to scientists in general that something is wrong with textbook perceptions in nuclear physics-and perhaps with the lack of perception by physicists that there is always a next step. John O'M. Bockris 4973 Afton Oaks Dr. College Station, TX 77845 The Truth in the Light by Peter Fenwick (with Elizabeth Fenwick). New York: Berkley Books, 1997.278 pp. $12.00, paper. ISBN 0-425-15608-7. Light and Death by Michael Sabom. Grand Rapids, MI: Zondervan Publishing House, 1998.240 pp. $12.99, paper. ISBN 0-310-21992-2. No one can complain about a dearth of books on near-death experiences (NDEs), but many of the authors of the books published on this topic have either written autobiographical accounts or declared themselves researchers and Book Reviews 133 cobbled a book from the results of an informal survey. Some qualified physicians have even engaged ghostwriters to help them appeal to a wider audience, suggesting that writing a nonfiction bestseller seemed more important than searching for the truth in these unusual experiences. My disappointment with nearly all the books in this field made me both surprised and pleased to find the two I recommend to critical readers here. Both the authors are physicians, and both have a record of publishing serious scientific research. Peter Fenwick (of London, England) is a neuropsychiatrist, a psychiatrist who is also well informed in neuroscience. He based his book on the results of a questionnaire sent to persons who learned about his research on NDEs primarily through television programs. More than 350 persons completed the questionnaire. Although Fenwick had a large sample, its participants selected themselves, and he did not claim otherwise. He interviewed only a few of his respondents. He takes his readers, chapter by chapter, through the principal features of the near-death experience, such as the sense of serenity and enhanced vitality, engulfment by a bright but not blinding light, looking down on one's body from above, reviewing one's life, and losing one's fear of death. Earlier books on near-death experiences have made these features familiar to many readers. Yet such earlier books-I discount autobiographical ones, which too often betray missionary zeal or more venal motives-rarely contain lengthy accounts of individual cases. Fenwick includes many long extracts from his respondents' own accounts of their experiences. In the last four chapters, Fenwick exhibits his mastery of neuroscience in an excellent discussion of alternative interpretations of the near-death experience. He acknowledges our dependence on the brain for our perceptions and nearly everything else that we experience. "Nearly everything" is, however, not everything. How can some persons who are ostensibly unconscious, if not dead, nevertheless perceive their bodies from above and sometimes become aware of events occurring outside the reach of their ordinary senses? Above all, how can our memories depend solely on our brains when persons whose brains seem not to be functioning during these experiences afterward remember them much better than they remember most events that occurred when they were in their normal states? If this book sells as widely as it deserves, a second edition will provide an opportunity to correct some errors in the spelling of names. Those of Melvin Morse (an investigator of children's near-death experiences), Andrew Greeley (an American sociologist), and Richard Bucke (a Canadian psychiatrist and writer on mystical experiences) are misspelled. A philosopher, easily identifiable from the context as Thomas Nagel, is cited, both in the text and in the index, as Tom Nagal. In 1982, Dr. Michael Sabom, a cardiologist, published one of the few books on near-death experiences that I can recommend to other physicians. It is called Recollections of Death: A Medical Investigation (New York: Harper and Row). Now he has written a second admirable book. 134 Book Reviews In his latest work, Sabom describes results from a smaller sample of persons who have had an NDE (47 in all) than the sample on which Fenwick based his work. Sabom's investigation, however, does offer three features that are lacking in Fenwick's research. First, most of the patients he studied were from his own practice, and he could appraise their physical conditions himself. Second, he compared the patients who had unusual experiences when critically ill with more than 100 cardiac patients who had comparable physical conditions, but had had no unusual experience. Third, he interviewed all 160 of the participants in his study. Michael Sabom particularly inquired about the after effects of the NDE on attitudes toward life and death. Using a Life Changes Questionnaire and the previously developed Religious Motivation Scale, he found significant differences between the patients who had had an NDE and those who had not. The life changes noted included increased faith in God, increased belief that life has meaning, and increased positive involvement with other persons. The scale of religious motivation showed an increase in the influence of religion on persons who had had an NDE as well, something quite different from mere attendance at church. Light and Death also describes some remarkable cases, including one that is quite extraordinary. The patient underwent a delicate neurosurgical operation for the ablation of an aneurism of the brain. The operation required her entire body to be chilled (hypothermia) and temporarily exsanguinated. She was as dead as anyone could possibly be unless they were "really dead." And yet she perceived events of the operation including details it seemed most unlikelybut not quite impossible-that she could have learned about normally. Michael Sabom would not mind if I describe him as an ardent Christian. He frankly states that his investigation derived from his religious beliefs as much as from his interest in science. Some readers may think that his religion may have interfered with his science. I am sure it did not. Everyone has religious beliefs, even atheists, and I find it refreshing that Michael Sabom told his readers explicitly about his. Near-death experiences remain anomalous in science because few scientists, including physicians, think that their investigation would result in any useful information warranting the efforts required. Why is this? Could the plethora of popular and "best-selling" books-many asserting extravagant claims-have repelled scientists by making the study of such experiences appear ridiculous? Why waste one's time, some scientists must have thought, in examining further the scanty evidence based on genuine scientific inquiries? Sherwin Nuland, in How We Die: Reflections on Life's Final Chapter (New York: Knopf, 1994), took less than two pages to reassure himself, and presumably his readers, that metabolic changes in a failing brain can satisfactorily account for the features of near-death experiences. To his credit, Nuland added a paragraph in which he mused about the inscrutability of God and concluded that perhaps, after all.. .. God, however, sometimes shows Himself in data ob- Book Reviews 135 tained by scientists, and that is why I urge them to read these two books. Those who do so will minimally see the need for more investigations of near-death experiences. Ian Stevenson Division of Personality Studies Box 152, Health Sciences Center University of Virginia Charlottesville, VA 22908 Psychedelic Drugs Reconsidered by Lester Grinspoon and James B. Bakalar. New York: The Lindesmith Center, 1997 (originally published 1979). 385 pp. ISBN 0-9641568-5-7. When this book was written in 1979, the intense debate and sensational rhetoric about psychedelic drugs was still recent memory. The time was surely ripe for a book such as Psychedelic Drugs Reconsidered, which struck a welcome note of sobriety after a decade or more of polemics. Psychedelic Drugs Reconsidered is a comprehensive survey of all aspects of the psychedelic question. All its subject matter is accessible to the lay reader, encompassing history, sociology, pharmacology, subjective and medical effects, and therapeutic use of psychedelic plants and synthetic drugs. Although the scope of this book is vast, its concise writing style, careful organization, and judicious editing make it an eminently accessible survey of the field-circa 1977. As a reprint and not a genuine new edition, some of the book is significantly dated. The pharmacological information it provides is easily updated for those so inclined and generally is consistent with contemporary knowledge. More serious is the sociological datedness. This book was written before the Reagan era, before the War on (some) Drugs and the concomitant mushrooming of the prisons and erosion of civil liberties, before the explosion of the rave scene. Psychedelic drugs occupy a different niche in society than they did in the 1970s. To understand the cultural significance of these substances in the 1990s, one will have to look elsewhere. Psychedelic Drugs Reconsidered scrupulously maintains at least a superficial objectivity throughout. The authors themselves recognize that impartiality is a problematic goal, as virtually all the language used in the psychedelic debate has become freighted with associations that embody the views of one side or another. Even the label "psychedelic '-meaning "soul manifestingwis controversial. Advocates of psychedelic drugs obviously prefer it to "psychotomimetic" (simulating psychosis), an earlier term with less-than-benign connotations. The authors mitigate this difficulty through consistently high standards of scientific integrity. They are not afraid to cite studies that go against the general direction of their conclusions, and they ruthlessly criticize flaws in stud7 136 Book Reviews ies in their favor. If anything, they tend to be overcautious, offering only qualified conclusions when their evidence seems overwhelming. Such intellectual probity lends credibility to their reportage and conclusions, as do their careful citations and impeccable documentation. The bibliography can only be described as "compendious"; especially useful is the separate annotated bibliography of several hundred titles, which, although dated, would be an invaluable resource for those pursuing a deeper knowledge of the field. Despite the scrupulous impartiality of the authors in attempting to present only the facts, one sometimes suspects that they are concealing their true opinions. They go to great lengths to make the book palatable to a skeptical audience (according to the jacket cover, even Carl Sagan approved). The rigor of its arguments and its careful documentation make the book ideal for the openminded scientist with little understanding of the field. True believers will be disappointed, however, for Psychedelic Drugs Reconsidered assiduously avoids committing to any paradigm-shattering conclusions. For example, the work of Stanislav Grof is cited extensively, but even though Grof (who started his career as an orthodox psychiatrist) argues persuasively that the evidence of psychedelic psychotherapy demands a radical rethinking of the present scientific worldview, Grinspoon and Bakalar shy away from such conclusions. They prefer to sit on the fence with assertions such as, "When someone says that under LSD he has relived a past life as an ancient Egyptian embalmer and produces an accurate formula for constructing a mummy (Grof, 1975, p. 170), he has probably read the formula somewhere, even if he cannot remember doing so." Yet one detects an underlying sympathy for the views of Grof and countless others who believe that the psychedelic experience cannot be subsumed under the rubric of conventional science. They are less equivocal about advocating a resumption in psychedelic research and therapy. Thousands of studies were published in the 1950s and 1960s, many in mainstream scientific journals, only to disappear entirely by the 1980s. Today, when a handful of fledgling ibogaine (a powerful psychedelic alkaloid derived from the African shrub Tabernathe iboga) and LSD studies are fighting for survival, their arguments could well prove useful to advocates of these studies. They remain current, if only because virtually no new research has been conducted since the book was published. If the authors are right, then the politically impelled cessation of psychedelic research was profoundly unscientific. Although the overall tone of the book is scholarly and reserved, the authors introduce anecdotes and extended quotations that convey something of the power and wonder of the psychedelic experience. These users' stories animate an extremely sophisticated discussion of the subjective effects of psychedelics. Withholding judgment about the more mystical psychedelic revelations, they persuasively argue that "These phenomena deserve to be placed in a theoretical framework rather than dismissed with casual psychiatric epithets and ad-hoc explainings-away on the one hand or affirmed as self-evident reve- Book Reviews 137 lations of the highest truth on the other" (p. 153). Any reader who conceives of LSD as a "hallucinogen" will find their old preconceptions untenable. For 25 years now, any hint of tolerance or advocacy for psychedelic drugs has been anathema to established political and public opinion. Therefore, in spite of the authors' carefully cultivated objectivity and balance, it is unlikely their work will have any greater impact on policy than it did in 1979. For most of the ideological spectrum, psychedelics are a closed issue. This is unfair. Today, even more than in 1979, aspects of the psychedelic experience are cropping up in fields as diverse as ufology, holistic medicine, death and reincarnation studies, transpersonal psychology, and anomalies research, demanding that we do as the book's title suggests-reconsider psychedelics. Charles Eisenstein 104 Matthew Circle State College, PA 16801 [email protected] The Meaning of Consciousness by Andrew Lohrey. Ann Arbor, MI: The University of Michigan Press, 1997. 301 pp. ISBN 0-472-10821-2. To be fair to the author, I may not be the right critic to provide an unbiased review of his book. I am, to put my cards on the table, an unreconstructed Cartesian dualist when it comes to the mind-brain problem and a follower of Popper and Eccles, who come in for criticism in this book. The author, on the other hand, as Katherine Hayles makes clear in her foreword, "develops a perspective radically different from a presumed separation between subject and object. Instead of seeing a world 'out there' separate from an observer, Lohry emphasizes the mutual construction of world, subject and discourse" (p. xii). And our author, in his own introduction, confirms what she says when he writes: "a science of consciousness should rely, not on reason or empiricism and their attendant conventions, which asks how consciousness arises from matter, but on a perspective paradigm which reverses the question to ask how matter arises from consciousness" (p. 2). Later, he adds that his book both develops and works from a new, non-rationalist and non-materialist framework or paradigm which challenges much in the scientific and Cartesian tradition of analysis. This paradigm, he points out, owes more to the new discipline of discourse analysis. Having thus layed his cards on the table, the author presents a glossary so that we should be in no doubt as to exactly how he is using the 22 key words or phrases that will crop up in the text. The book is scholarly, and the author knows the literature and is familiar with the controversies it has generated. He has read his Sheldrake, although Sheldrake is taken to task for "his lack of a definition of and a lack of structural content for his notions of habit and morphic resonance." 138 Book Reviews It is unfortunate that there is no reference to the work of David Chalmers, author of The Conscious Mind, but, as that was published only in 1996, it presumably came too late for the present book, which appeared in 1997. As far as this reviewer is concerned, the parting of the ways comes with Chapter 9, "The Unity of Mind and Body." It is here that Lohrey challenges the pessimism of Karl Popper and John Eccles, who expressed doubts as to whether we shall ever be able to understand "the relation between our bodies and our minds." The author has no such doubts: I maintain that an isomorphic paradigm, which predisposes to recursive symmetry and therefore unity, can provide a positive answer to this relationship. The nature of this positive answer is that neurological activity in the form of excitation and inhibition are the actual semantic relationships of symmetry, nonsymmetry, and asymmetry and that through these the recursive systems in general of subjectivity are established. The unity of "mind" and body is therefore complete and absolute. (p. 180) The quotes around the word "mind" say a great deal about the way the author approaches the problem, which is to downplay the distinction, which to many of us seems so unbridgeable, between mind and consciousness on the one hand, and body and matter on the other. But the author is even willing to summon to his aid, U. T. Place, who famously argued that consciousness is just a brain process. It is noteworthy that there is nowhere a discussion of the "paranormal," so we do not know whether the author acknowledges or denies the reality of parapsychological abilities. Yet ever since J. B. Rhine, the existence of such abilities has been cited as demonstrating the independence of mind from material constraints. It may be that given my own philosophic allegiances, I could not be expected to see matters from the author's point of view. But, as I said at the beginning, I may not be the reviewer that the author deserves. John Beloff 6 Blacket Place Edinburgh EH9 1 RL Scotland UK Journal of Scientific Exploration, Vol. 14, No. 1 , pp. 139,2000 0892-3310100 0 2000 Society for Scientific Exploration ERRATUM rality, N o w , edited by Harald Atmanspacher and Evan entific E x p l o r a t i o n , 1 3 , 4 , 1999, pp. 695-703. The The references accidentally were omitted for the book review Time, TempoRuhnau, J o u r n a l of Scireferences, cited by reviewer Karl Gustafson, are listed below. References Atmanspacher, H., Primas, H., & Wertenschlag-Birkhauser, E. (1995). Der Pauli-Jung-Dialog und seine Bedeutung fiir die modern Wissenschaft. Berlin: Springer. Bernasconi, J., & Gustafson, K. (1998). Contextual quick-learning and generalization by humans and machines. Network: Comput. Neural Syst., 9, 85. Brent, J. (1993). Charles Sanders Peirce: A life. Bloomington, IN: Indiana University Press. Calude, C., Casti, J., & Dinneen, M. (Eds.). (1998). Unconventional models of computation. Berlin: Springer. Davies, E. (1976). Quantum theory of open systems. London: Academic Press. Gustafson, K. (1990). Reversibility in neural processing systems. Statistical mechanics of neural networks (L. Garrido, Ed.). Lecture Notes in Physics 368, 269. Gustafson, K. (1997a). Lectures on computationalfluid dynamics, mathematicalphysics, and linear algebra. Singapore: World Scientific. Gustafson, K. (1997b). Operator spectral states. Computers Math. Applic., 34, 467. Gustafson, K. (1998). Internal sigmoid dynamics in feedforward neural networks. Connection Science, 10, 43. Lahti, P., & Mittelstaedt, P. (1985, 1987, 1990). Symposium on the foundations of modernphysics. Singapore: World Scientific. Pauli, W. (1953). Die Klavierstunde. ETH, Zurich. Unpublished manuscript. Richmond, B., Optican, L., Podell, M., & Spitzer, H. (1987). Temporal encoding of two-dimensional patterns by single units in primate inferior temporal cortex. I. Response characteristics. 11. Quantification of response wave form. 111. Information theoretic analysis. Journal of Neurophysiology, 57, 132, 147, 162. Schwartz, D. (1981). Isomorphism of Spencer-Brown's laws of form and Varela's calculus for self-reference. International Journal of General Systems, 6, 239. Spencer-Brown, G. (1969). Laws offomz. London: Allen and Unwin. Webster's third new international dictionary. (1971). Springfield, M A : G. C. Merriam, Publishers. Wiener, N. (1948). Cybernetics. New York: Wiley. In addition, the correct keywords for "Experimental Systems in Mind-Matter Research" by Robert Morris, Journal of Scientific E x p l o r a t i o n , 1 3 , 4 , 1999, pp. 561-577 should be: statistical evidence, p-values, meta-analysis, and re- Journal of Scientijic Exploration, Vol. 14, No. 1, pp. 141-148,2000 0892-3310100 O 2000 Society for Scientific Exploration SSE 19th Annual Meeting at King's College in London, Ontario, Canada June 8-10,2000 Announcement and Call for Papers Program The Society for Scientific Exploration will hold its 19th Annual Meeting in London, Ontario, Canada, on the King's College campus, June 8-10, 2000. The SSE Local Host is Imants Baruss. A number of distinguished scholars have accepted invitations to speak on themes that embody SSE's focus on the frontiers of science. This year's program, at the beginning of a new millennium, will give special attention to the future of science and the contributions made by researchers investigating unusual, elusive, and anomalous topics. Contributed Papers Contributed presentations from SSE members and from researchers sponsored by members will be welcomed by the Program Committee, which consists of Roger Nelson (Chair), Brenda Dunne, Patrick Huy ghe, Wayne Jonas, and Marcel Kuijsten. We especially encourage contributions from young investigators. Abstracts no longer than 400 words should be submitted by April 15, preferably by e-mail and in plain ASCII text, to: Roger Nelson, C- 131, School of Engineering, Princeton University, Princeton, New Jersey 08544, USE. Email: [email protected]; Tel: 609-258-53709; Fax: 609-2581993 Invited Speakers A number of distinguished scholars have accepted invitations to address themes embodying SSE's purpose as a forum for frontier science. This year's program, at the beginning of a new millennium, will focus on the future of science and the special contributions that can be made by researchers investigating unusual, elusive, and anomalous topics. Talks by our invited speakers fall into the following overlapping categories: I. The Evolution of Science: Philosophy, Technology, Purpose R. G. Jahn, Princeton University: Plenary, A Future Science John Peterson, Arlington Institute: A Futurist Assesses Engineering Seth Shostak, SET1 Institute: Search for signals from the Universe 11. Biosciences and Medicine: Anomalies in a Pragmatic Framework Wayne Jonas, Univ. Uniformed Services: Homeopathy, Research Strategies for Hard Problems Elisabeth Targ, U. C. San Francisco: Distant Healing in AIDS Patient Population 142 SSE News Alexander Berezin, McMaster University: High Dilutions: Isotopicity, Quantum Coherence 111. Language and Consciousness: Window and Mirror Imants Baruss, King's College: Alterations of Consciousness Evan Pritchard, Marist College: Comprehensivism: The Way of the Future (and Past) IV. Forum: Creative, Constructive, Critical Anomalies Research Logistical Information: See registration form. The following information gives a brief orientation. Further information is available on the Society website: www.scientificexploration.org. For questions and detailed information, please contact the local host, Imants Baruss. Tel: 519-433-3491; Fax: 519-433-0353; Email: [email protected]. Airport International airport with flights from Toronto, Detroit, and Pittsburgh. Ground transportation from Detroit and Toronto. Hotel Station Park All Suite Hotel; Group reservation number 30214; 242 Pall Mall Street, London, Ontario; Telephone: 1-800-561-4574; Fax: 519-6422551. Canadian $1 10 (about $78 U.S.), parking included. Station Park is located downtown at Richmond Row, London's premier business, shopping, dining and entertainment district. It is 10 minutes away by car from the University campus. Shuttle to campus. Make reservations by April 28. Accommodations at King's College Campus rooms, spacious, shared bathrooms, hot breakfast included. $35 (U.S.) per day single occupancy; $30 per day per person for double occupancy. To reserve rooms in the student residences, use the MEETING REGISTRATION FORM. For questions about the campus and campus rooms, contact baruss@julian. uwo.ca. Field Trip Friday afternoon: Pinery Provincial Park on Lake Huron, one hour away, near Grand Bend, box lunch, kilometres of hiking and beach. Bring swimwear AND rainwear, sweater or windbreaker, if you stay on to see the sunset, considered among the most beautiful in the world. Alternate trip: One hour to Stratford Shakespeare Festival. Society Banquet Saturday evening at Sunningdale Gold Club (established in 1901 and ranked as one of the top 20 golf courses in the world), outdoor patio, lounge, formal dining room is all ours, grand piano entertainment and singing after dinner, beautiful grounds, woods, view of creek. Dress code: no jeans or halters. SSE News 143 REGISTRATION & PAYMENT INFORMATION FOR SSE 2000 MEETING Please Print Name: Address: All Amounts are given in US Dollars I Telephone: E-mail: Work: Home: FAX: I I Name for Badge: Badge Affiliation (or City): Number in Party: (Please copy this form as needed and fill out for each additional person.) I Early Meeting Registration Fee: $100 Registration Fee After May 5,2000: $125 Daily Registration Fee: $35 early; $45 after May 5,2000 $ $ $ I I Accommodation in King's College Residence (with shared bathrooms and hot breakfast) One person in room: $35 per night Two people in room: $30 per person per night Indicate which nights: Wednesday Thursday Friday $-$ Saturday Lunch, with some choice, in King's College dining room: $5 per day $ (otherwise you hike 2.5 km to downtown restaurants) n vegetarian Indicate if vegetarian: Thursday U Friday Indicate which days: Friday evening picnic at Pinery Provincial Park on Lake Huron: $25 (includes one hour bus trip, park entrance fee and box lunch) egg vegetarian Choose: fl meat rl fish $ Saturday Saturday evening banquet at Sunningdale Gold Club: $30 $ (includes 15 minute bus trip, pre-dinner bagpipes on terrace and after dinner lounge piano) Choose: U prime rib 1-1 puff pastry salmon en crouti O vegetarian Total Paid $ Please return with check payable to SSE or with VisaJMatercard information to: Charles Tolbert, PO Box 3818, Charlottesville, VA 22903. FAX: 804-924-3104 Card Number: - - - Expiration Date I SSE News 145 Fifth Biennial SSE European Meeting Announcement and Call for Papers The Fifth Biennial SSE European Meeting will be held from Oct. 20-22,2000, at the University of Amsterdam in the Netherlands. Dick Bierman and Chris Duif are hosting the meeting and coordinating registration. Ezio Insinna is the chairman of the program committee. Program A number of distinguished scholars have been invited to speak on the theme: Unorthodox Science: Past, Present, and Future. The following speakers have accepted at the time of publication: H. B. G. Casimir (Netherlands): On the Nature of Vacuum (video interview). S. J. Doorman (Dutch Skeptical Society): On the Future of Unorthodox Science. F. H. van Lunteren (University Utrecht, Netherlands): On the History of Unorthodox Science. W. Peschka (Germany): Kinetobaric Effects and Bioinformational Transfer. Pyatnitsky (Moscow Institute for High Temp., Russia): Effects of Consciousness on the Structure of Water. S. E. Shnoll (Moscow State University, Russia): Realization of Discrete States during Fluctuations in Macroscopic Processes. R. Sheldrake (UK): Animal Communication. R. van Wijk (University Utrecht, Netherlands): High Dilutions Research. K. Zioutas (University Thessalonike, Greece): Evidence of Dark Matter from Biological Observations. Further information on the program will be announced on the SSE website (www.scientificexploration.org/meetings/euro5 .html) Call for Papers All interested European SSE members are encouraged to submit abstracts for short contributed presentations to the program committee consisting of Ezio Insinna, Brenda Dunne, and Dick Bierman. Abstracts no longer than 500 words must be sent before August 15, preferably by e-mail to emi2Qworldnet.fr. Address: Ezio Insinna, 18 allee des Freres Lumiere, 77600 Bussy Saint 146 SSE News LOGISTICAL INFORMATION Traveling The International Schiphol Airport can be reached easily from almost everywhere. Train connections from the airport to the city (Central Station) are running every 15 minutes. Travel time from the airport to Central Station is about 15 minutes. If you are traveling by automobile, please note that there is no free parking around the conference location. Check with your hotel or the local organizer Dick Bierman about how to arrange for parking. Hotels Main Hotel (-$90): Lancaster Hotel, Plantage Middenlaan 22, Amsterdam, the Netherlands. Phone: +31 (0) 20 5356888. Email: res.lancasterhote1 @edenhotelgroup.nl. Mention group code UVA-SSE when making reservations. Budget hotel, backpackerstyle (-$45): Arena Budget Hotels, Gravezandestraat 51, 1092 AA Amsterdam, the Netherlands. Phone: +31 (0) 20 6947444. Email: infor @ hotelarena.nl. Information on other smaller hotels within walking distance will be provided on the website. The hotels are situated in a relatively green area near the Zoo and can be reached by taking either a cab from the Central Station or tram nr. 9 (get off at the Zoo for Lancaster or Tropenmuseum for Arena). Social Activities The University of Amsterdam is situated near the center of the city and the program will be set up in such a way that attendees have opportunities to enjoy the cheerful atmosphere and picturesque canals in Amsterdam. A reception and registration will take place on Thursday evening (October 19) at a nearby Cafe (to be announced), all within a two-minute walk from the Lancaster Hotel. The Society banquet is planned for Saturday evening at the Zoo restaurant with a splendid view over the animals and gardens. The banquet will be preceded by a trip through the Amsterdam canals. The dress code is informal. Sightseeing Friday evening is free for exploring the city. Sunday afternoon might be used to visit the museums, which have large collections of Rembrandt and van Gogh paintings. The Anne Frank house is open for visitors on Sunday afternoon. For more information on local arrangements contact: Dick Bierman, email: [email protected], fax: +31-20-639 1656. Registration Copy this registration form and send it to: Dick Bierman University of Amsterdam Roetersstraat 15 1018 WB Amsterdam The Netherlands Or: Register using the web registration form at our website. SSE News Registration Form European SSE 2000 Conference Contact Information Affiliation for Badge: Street Address: City/State/ZIP/Country: Phone: E-mail: Registration Fee Select Registration Type: Early Fee (before Sept. 1): $85 / Dfl. 190,Late Fee (after Sept. 1): $100 / Dfl. 220,n Daily Registration (1 day): $40 / Dfl. 100,n Daily Registration (2 days): $75 / Dfl. 170,U 50% DISCOUNT. Check this if you are an enrolled student. U Lunches and Banquet n On Site Lunches ($20 for 3 days) Friday Night Banquet ($35) (includes trip through canals) fl Check this if you are a vegetarian Payment Method Total amount (the sum of the registration fee, lunches, and banquet): Select Payment Method: Credit Card CardNumber: Signature: fl $ - - - Expiration Date 1 Check (make check payable to Euro SSE 2000) Bank transfer to PostBank Account 8417584 of Euro SSE2000, the Netherlands Please mail registration form to Dr. Dick J. Bierman, SSE 2000, University of Amsterdam, Roetersstraat 15, 1018 WB Amsterdam, The Netherlands Journal of Scientijic Exploration, Vol. 14, No. 1, pp. 149-161,2000 0892-3310100 02000 Society for Scientific Exploration Corrected Index for Volume 13 The following is the revised set of comprehensive subject and name indices for Vol. 13, No. 1 through Vol. 13, No. 4 of the Journal. The indexers apologize for the accidental omission of JSE Vol. 13, No. 3 in the version which appeared at the back of JSE Vol. 13, No. 4. The Name Index includes, by author, all articles and book reviews that have appeared during 1999, as well as books by authors that have been reviewed by someone else. Substantive citations to, and quotes from, an author's work have also been indexed; routine references have not. The Subject Index includes a list, sorted alphabetically by author, of all books reviewed in the Journal, under the header "Book reviews." The name of the reviewer is included in parentheses at the end of each entry. The Editors wish to thank Dawn Hunt and James Matlock for compiling the indices. Name Index Adams, John Couch, 259,260 Ajello, Elena, 463 Akers, Charles, 627 Alcock, James E., 317,318 Altschuler, M. D., 201 Amann, Anton, 559,696, Introductory Remarks on Large Deviation Statistics, 639-64 (co-author) Anderson, Jane, The Shape of Things to Come, 552-53 (book review) Antoniou, Ionnis, 696 Arechi, F. Tito, 696 Arp, Halton, book review by, 341-43 Arvidson, P. Sven, Intuition: The Inside Story, 337-4 1 (book review) Atmanspacher, Harald, 559,696-97 Guest Editorial: Data Analysis in Mind-Matter Research, 557-60 Introductory Remarks on Large Deviation Statistics, 639-64 (co-author) Time, Temporality, Now, 695-703 (co-author) (book review) Babinski, J., 465 Bacon, Francis, 266 Banting, Sir Frederick, 500 Barbour, Ian G., Religion and Science: Historical and Contemporary Issues, 333-35 (book review) Barbour, Julian, 696 Batcheldor, Kenneth, 247 Bayer, Resat, 190, 193 Beloussov, L. V., book review by, 346-48 Bendernagel, John A., North America's Great Ape: The Sasquatch, 549-51 (book review) Benveniste, J., 146 Bernard, Claude, 266 Birkhoff, George, 594,597 Black Elk, 243,255 Blasband, Richard A., book review by, 706-9 Blofeld, J., 10-11 Bockris, John O'M., book review by, 343-46 Bodner, A. G., 495 Bohm, David, 106 Bohr, Niels, 104-5,500 Boller, Emil, A Rescaled Range Analysis of Random Events, 25-40 (co-author) Bondi, Hermann, 282 Boole, George, 586 Born, Max, 582 150 Name Index: Volume 13 Family History: Pedigree and Segregation Analyses, 351-72 Colby, William, 72,73 Comfort, A., 489 Condon, Edward U., 422,454 Conte, 344,345 Coon, C., 551 Costa de Beauregard, O., Comment on "The Timing of Conscious Experience" by F. A. Wolf, 323-25 Cournot, Antoine Augustine, 581,585 Cox, Edward, 232,234,243,247,248, 250,252 Cram&-,H., 642,643 Crater, Horace W. Mound Configurations on the Martian Cydonia Plain, 373-96 (co-author) Reply to P. A. Sturrock's Criticisms, 398-400 (co-author) Response to "Geomorphology of Selected Massifs on the Plains of Cydonia" by David Pieri, 413-19 (co-author) Creegan, Robert F., book review by, 333-35 Curran, Pearl, 250,255 Dalenoort, Gerhard, 696 Darwin, Charles, 262,346-48,349-50, 551 David-Neel, Alexandra, 247 Davis, Elizabeth, 340 Davis-Floyd, Robbie, Intuition: The Inside Story, 337-41 (book review) Davy, John, The Marriage of Sense and Thought: Imaginative Participation in Science, 335-37 (book review) de Broglie, L., 45 de Finetti, Bruno, 584 Dean, D., 147 DeGracia, Donald J., 152-53 Report of Referee on "The Effect of 'Healing with Intent' on Pepsin Enzyme Activity," 149-52 DeMeo, James, The Orgone Boschinger, Christoph, 116 Boucouvalas, Marcie, 338 Bourron, Michel, 117 Bradish, G. J., Response to Stanley Jeffers, 329-31 (co-author) Brame, E. G., 147 Brandenberg, J. E., 418 Braud, William G., 516-17,518 Braude, Stephen, 3 18 Bridgman, P. W., 557 Broughton, Richard S., 232 Bunnell, Toni, 149-52 Bunnell's Response to deGracia, 152-53 The Effect of "Healing with Intent" on Pepsin Enzyme Activity, 139-48 Burneko, Guy, 339 Bush, George, 80 Butler, Alisa, 236,237,244 Calder, N., 282-83 Calvin, Steve, 232,234,235 Cann, R. L., 496 Carlotto, Mark J., 378 Response to "Geomorphology of Selected Massifs on the Plains of Cydonia" by David Pieri, 413-19 (co-author) Carnap, Rudolf, 584-85 Carpenter, James, 575 Cartier, Th&reseJosephine, 480 Castaneda, Carlos, 10 Chai, C. K., 494 Charcot, Jean-Jaques, 464 Chauvin, Remy Le darwinisme ou la$n d'un mythe [Darwinism, or the End of a Myth], 346-48,349-50 (book reviews) Psychological Research and its Alleged Stagnation, 317-22 Childerhose, Robert J., 207-8 Church, Alonzo, 597-98 Ciochon, R., 551 Clarke, J. M., 494 Clayton, Craig, book review by, 122-24 Cohn, Shari A., Second Sight and Name Index: Volume 13 Accumulator Handbook, 706-9 (book review) Diaconis, Percy, 557,563 Dibble, Walter E., Jr., Electronic Device-Mediated pH Changes in Water, 155-76 (co-author) Dirac, P. A. M., 47, 185-87,287,324 Dobbins, Yvonne H., Response to Stanley Jeffers, 329-31 (co-author) Dorit, R. L., 496 Dumas, G., 461 Dunne, Brenda J., 105,108,339-40,569 Response to Stanley Jeffers, 329-31 (co-author) Dyson, Frank Watson, 273,275,276, 278,282 151 State-of-the-Universe(s) Report, 341-43 (book review) Feynman, 397 Fithian, Elly, 235,236,237,244,249 Flammarion, Camille, 303-4 Fleck, Ludwik, 266 Florshiitz, Gottlieb, Swedenborg's Hidden Influence on Kant: Swedenborg and Occult Phenomena in the View of Kant and Schopenhauer, 545-49 (book review) Flournoy, Thomas, 267 Fogarty, Quentin, 433 Fonkin, V. A., 146 Fox, Roy, 543 Freud, Sigmund, 93-94,3 19 Earman, John, 582 Eddington, A. S., 260, 271,273, 275-78,280-82,286,287,288 Edelglass, Stephen, The Marriage of Sense and Thought: Imaginative Participation in Science, 335-37 (book review) Edelstein, Stuart J., book review by, 349-50 Edmonds, James D., Jr., Variations on the Foundations of Dirac's Quantum Physics, 177-88 Ehlers, Jiirgen, 696 Eibesfeld, Eibl, 3 19-20 Einstein, Albert, 182, 183,260,264, 271,272,278,279-89,328,342 Eldridge, Don, book review by, 552-53 Endler, 541 Erjavec, James L., Response to "Geomorphology of Selected Massifs on the Plains of Cydonia" by David Pieri, 413-19 (co-author) Eshleman, Von R., 423-24,448 Estebany, Oskar, 145 Euler, Manfred, 696 Everret, Hugh, 553 Falconer, D. S., 490 Feder, J., 26,28,31 Fernandes, Marco, 697 Ferris, Timothy, The Whole Shebang: A Galle, Johann Gottfried, 259 Gebert, Hans, The Marriage of Sense and Thought: Imaginative Participation in Science, 335-37 (book review) Gebser, Jean, 297,298 Gehrcke, Ernst, 283-84 Gemmeli, A., 466 Gerber, P., 283-85,289 Gibbs, Josiah Edward, 580 Gillmor, D. S., 454 Goerres, J. J., 466 Goethe, J. W. von, 335-36 Gonzales, B. M., 494 Gould, Stephen J., 342 Grad, Bernard, 147,321 Grattan-Guiness, I., Real Communication? Report on a SORRAT LetterWriting Experiment, 231-56 Graudenz, Dirk, 696 Griffin, David Ray, Unsnarling the World Knot - Consciousness, Freedom, and the Mind-Body Problem, 121-22 (book review) Gross, Paul R., Higher Superstition: The Academic Left and Its Quarrels with Science, 554-56 (book review) Guest, J. E., 407 Gustafson, Karl, book review by, 695-703 152 Name Index: Volume 13 tor Intention: A Review of a 12Year Program" by R. G. Jahn, B. J. Dunne, R. D. Nelson, Y. H. Dobbins and G. J. Bradish, 327-29 Jeffrys, Harold, 583 Johnson, Gregory R., book review by, 545-49 Jonas, Wayne B., 541-42 Non-locality and Homeopathy, 539-41 Josephus, Flavius, 488 Jung, Carl Gustav, 94,292,298,301-5, 308,572,603 Kamber, Franz, 601 Kant, 667 Kant, Immanuel, 545-49 Kaplan, Joseph, 687 Katra, Jane, Miracles of the Mind: Exploring Non-local Consciousness and Spiritual Healing, 129-31 (co-author) (book review) Kauffman, Stuart, 347 Keil, Jiirgen, Do Cases of the Reincarnation Type Show Similar Features Over Many Years? A Study of Turkish Cases a Generation Apart, 189-98 (co-author) Kelleher, Colm A., Retrotransposons as Engines of Human Bodily Transformation, 9-24 Kelvin, Lord, 261 Kenyon, K., 485 Kepler, J., 262-63 Keyhoe, Donald, 686,690 Keynes, John Maynard, 584 Khrennikov, Andrei, 559 p-adic Information Spaces, Infinitely Small Probabilities and Anomalous Phenomena, 665-79 Kiefer, Claus, 696,699-700 Klass, Philip, 443-45,447,448 Klass, Phillip, 201,209 Klose, Joachim, 695 Koestler, Arthur, 89 Kolmogorov, Andrei Nikolaevich, 559, 588-89,590-92,594,595,597, 598,599,600,601,602 Guth, Alan, The InJEationaly Universe, 341-43 (book review) Hahnemann, 540 Halley, Edmund, 258-59,260 Hamilton, Jennifer L., book review by, 703-4 Hamilton, R., 178 Hansen, George P., 232 Harbort, Bob, 340 Harman, W., 500 Harman, Willis W., Biology Revisioned, 703-4 (co-author) (book review) Head, J. W., 411 Hemenway, Donald, 455 Hetherington, N. S., 264 Hiley, Basil, 697 Home, D. D., 318 Honorton, Charles, 626-27,628,629 Horrobin, D. F., 267 Houran, James, "The Jim Ragsdale Story" Revisited: An Appraisal of Early Unpublished Testimony, 113-16 (co-author) Howe, Elias, 500 Hoyle, Fred, 341 Towards and Understanding of the Nature of Racial Prejudice, 681-84 (co-author) Hunt, Dawn Elizabeth, book review by, 124-29 Hurst, H. E., 25-26,31,32 Hyman, Ray, 617,626-27,628,629, 632,634-35 Hynek, J. A., 116,117 Imbert-Gourbeyre, A., 462,466 Ireland, W., 447,448 Jahn, Robert G., 105, 108,569 Response to Stanley Jeffers, 329-31 (co-author) Jawer, Michael A., Human Factors and Anomalous Experience: A Survey Investigation, 542-43 Jeffers, Stanley, Comment on "Correlations of Random Binary Sequences with Pre-Stated Opera- Name Index: Volume 13 Konig, H., 593 Koopman, Bernard Osgood, 584-85, 596 Krengel, U., 557 Kress, Kenneth A., 89 Parapsychology in Intelligence: A Personal Review and Conclusions, 69-87 Targ response, 87-90 Krippner, Stanely, book review by, 554-56 Krishna, Gopi, 20 Kronz, Frederick, 695 Kuhn, Thomas, 265 Kull, Andreas, 695,698 Kura, Ertan, 190 Lakotas, I., 266 Laplace, Pierre Simon, 266-67,579-80, 584 Laszlo, E., 142-43 Lateau, Louise, 463 Laughlin, Charles D., 338 Leibnitz, G. W., 297,298-99 LeLorier, J., 631 Leverrier, Urbain, 259,260,272 Levin, Michael, Commercial Funding for Frontier Science, 537 Levitt, Norman, Higher Superstition: The Academic Left and Its Quarrels with Science, 554-56 (book review) Linde, Klaus, 540,541 Links, Jonathan M., book review by, 335-37 Lorentz, Konrad, 182, 183,285,319-20 Lyre, Holger, 695 Maccabee, Bruce Atmosphere or UFO? A Response to the 1997 SSE Review Panel Report, 421-59 Optical Power Output of an Unidentified High Altitude Light Source, 199-211 Maddox, J., 271-72,273,287,288 Mahler, Giinter, 697 Maier, Georg, The Marriage of Sense 153 and Thought: Imaginative Participation in Science, 335-37 (book review) Malin, Shimon, 695 Mandelbrot, B. B., 27 Manning, Matthew, 142, 145,321 Margnelli, Marco, An Unusual Case of Stigmatization, 461-82 Marnard Smith, J., 494 Martin-Lof, Per, 599,659 Mathews, Olga, 237,244 Matthews, Robert A. J., 630 Significance Level for the Assessment of Anomalous Phenomena, 1-7 Mauge, Claude, Response to the Pocantico Report, 116-20 Maxwell, J. C., 179, 181, 182, 183, 184 McCausland, Ian, Anomalies in the History of Relativity, 271-90 McClenon, James, 232,234,235,243, 254 McClintock, Barbara, 9-10, 11, 19 McCraty, R., 146 McDaniel, Stanley V. Mound Configurations on the Martian Cydonia Plain, 373-96 (co-author) Reply to P. A. Sturrock's Criticisms, 398-400 (co-author) Response to "Geomorphology of Selected Massifs on the Plains of Cydonia" by David Pieri, 413-19 (co-author) McDermott, Zachary, Exploring the Limits of Direct Mental Influence: Two Studies Comparing "Blocking" and "Co-operating" Strategies, 515-35 (co-author) McDonald, James, 209,689 McKague, Lee, Methuselah: Oldest Myth, or Oldest Man?, 483-97 McLaughlin, Robert, 687-88,691-92 McMahon, John, 88 McNally, 205 Mehra, Jagdish, 697 Mendeleev, 259-60 Menzel, Donald, 689,691 154 Name Index: Volume 13 Pauli, Wolfgang, 178,298,301-2,304, 318,586,603,698,700 Pearl, R., 493-94 Peierls, Rudolf, 697 Peirce, Charles Sanders, 305-7 Perrin, Jean Baptiste, 595 Pert, Candace, 553 Pieri, David C., 413-18 Geomorphology of Selected Massifs on the Plains of Cydonia, Mars, 401-12 Pineault, Ann, 338 Poher, C., 118 Polanyi, Karl, 265 Polat, Can, 190 Pollard, Frank G., Comments on Reincarnation, 116 Poppel, Ernst, 695-96 Porter, Stephen, "The Jim Ragsdale Story" Revisited: An Appraisal of Early Unpublished Testimony, 113-16 (co-author) Post, Emil Leon, 598 Price, Pat, 72-73,74-79,82,83-84,87, 88-89,130 Primas, Hans, 696,697,698,699,700 Basic Elements and Problems of Probability Theory, 579-614 Prince, Walter Franklin, 250 Puthoff, Harold E., 70,71,75, 80, 87-88 Putigny, Thkrese Putigny, 481 Pyatnitsky, L. N., 146 Ragsdale, Jim, 113-15 Randall-May, Cay, Analysis of Technically Inventive Dream-Like Mental Imagery, 499-513 (co-author) Randle, Kevin D., 114, 115 Ratnoff, 0 . D., 466 Ravenscroft, John, Exploring the Limits of Direct Mental Influence: Two Studies Comparing "Blocking" and "Co-operating" Strategies, 515-35 (co-author) Rees, Martin, Before the Beginning, 341-43 (book review) Methuselah, 483-96 Miley, George, 344,345 Mill, John Stuart, 260 Miller, A. I., 288 Miller, R., 146, 147 Milton, Julie, 629-30 Miollis, Madame, 480 Mizuno, Tadahiko, Nuclear Transmutation: The Reality of Cold Fusion, 343-46 (book review) Monsay, Evelyn H., 339 Moore, Charles, 687-88 Morgan, T. H., 265 Morris, Robert L., 559 Experimental Systems in Mind-Matter Research, 561-77 Murphy, Michael, 10 Napier, J., 551 Narlikar, Jayant, 342 Neihardt, John G., 231-32,240,243, 250,252,254-55 Nelson, Roger D., Response to Stanley Jeffers, 329-31 (co-author) Neppe, Vernon M, 353 Neumann, Therese, 463 Newton, Sir Isaac, 183,259,298 Nordberg, Robert B., book review by, 121-22 OIKeefe,Eleanor, 236 Ockham, William, 296 Oster, Eileen F., The Healing Mind: Your Guide to the Power of Meditation, Prayer, and Rejlection, 122-24 (book review) Owen, A. R. G., 247 Pais, Abraham, 279-80,283,288 Palladino, Eusapia, 318 Pallikari, Fotini, A Rescaled Range Analysis of Random Events, 25-40 (co-author) Palmer, John, 369 Parker, S. L., 493 Parker, T. J., 411 Partridge, Michael D., book review by, Name Index: Volume 13 Reich, Wilhelm, 706-9 Reichenbach, Baron von, 59-60 Rein, G., 142, 145, 146,321 RCnyi, AlfrCd, 589 Reverdit, 480 Rhine, J. B., 232,318,321 Rhunke, Lothar, 455 Richards, John Thomas, 232-33,234, 235,236,243,244,248,252,254, 255 Richet, Charles, 557 Ring, Kenneth, 19 Robertson, Cloretta, 463 Rodeghier, Mark, 114, 115 Roll, W. G., 542-43 Roncalli, Lucia, 340 Rose, Charles, 89 Ruhnau, Eva, 695 Time, Temporality, Now, 695-703 (co-author) (book review) Ruppelt, Edward, 686 Russek, Linda G. S., Registration of Actual and Intended Eye Gaze: Correlation with Spiritual Beliefs and Experiences, 213-29 (co-author) Russell, Bernard, 583 Ryle, Martin, 342 Sagan, Carl, 413,693 Sahtouris, Elisabet, Biology Revisioned, 703-4 (co-author) (book review) Saint-Laudy, 541 Savage, Leonard J., 583 Schlaug, G., 352 Schlitz, Marilyn J., 516-17,518 Schmitt, Donald R., 114 Schnorr, Claus-Peter, 599 Scholem, Gershom, 292,304 Schopenhauer, A., 298 Schopenhauer, Arthur, 548 Schuessler, J. F., 118 Schwartz, Berthold E., 232,237 Schwartz, C. J., 150,153 Schwartz, Gary E. R., Registration of Actual and Intended Eye Gaze: Correlation with Spiritual Beliefs and Experiences, 213-29 (co-author) 155 Schwartz, Stephen A., book review by, 129-31 Sciama, D. W., 272,273,282,287 Shattuck, Roger, Forbidden Knowledge: From Prometheus to Pornography, 124-29 (book review) Sheldrake, Rupert, 538,555 Sheridan, Joe, 338 Shklovskii, I. S., 692-93 Silberstein, Ludwig, 279-81 Simpson, C. J., 467 Smith, Cyril, 543 Smith, J., 145, 146 Smith, W., 117 Spencer-Brown, G., 557,695,698 Sprague, Roderick, 551 Stanford, Rex, 572 Steiner, Rudolf, 335-36 Stevenson, Ian, 116 book review by, 705 Do Cases of the Reincarnation Type Show Similar Features Over Many Years? A Study of Turkish Cases a Generation Apart, 189-98 (co-author) What are the Irreducible Components of the Scientific Enterprise?, 257-70 Stokes, Douglas M., The Nature of Mind: Parapsychology and the Role of Consciousness in the Physical World, 705 (book review) Stone, Marshall Harvey, 588 Strange, J. F., 400 Strasenburgh, Gordon, book review by, 549-51 Stringfield, Len, 691 Sturrock, Peter A., 5, 116-20 Referee Report on "Mound Configurations on the Martian Cydonia Plain," 397-98 Suchanecki, Zdaslaw, 696 Swedenborg, Emanuel, 545-49 Swords, Michael D., Clyde Tombaugh, Mars and UFOs, 685-94 Szelag, Elzbieta, 696 T., Anna Maria, 467-81 156 Name Index: Volume 13 Vandenberg, Hoyt, 686 Velasco, Jean-Jacques, 421-22 Venn, John, 581,585 Verbrugghe, G. P., 488 Virchow, R., 463 von Ludwiger, Illobrand, 421,422 von Mises, Richard, 582-83,586,597 von Neumann, J., 596 von Weizsacker, Carl Friedrich, 585, 695,697,698 Waismann, Friedrich, 582,591 Walach, Harald, 5 39-40 Magic of Signs: A Nonlocal Interpretation of Homeopathy, 291-315 Reply to Jonas, 541-42 Watt, Caroline Exploring the Limits of Direct Mental Influence: Two Studies Comparing "Blocking" and "Co-operating" Strategies, 5 15-35 (co-author) Rupert Sheldrake and the Objectivity of Science, 538-39 (co-author) Whitehead, Alfred North, 279,299-300 Wickersham, J. M., 488 Wickramasinghe, Chandra, Towards an Understanding of the Nature of Racial Prejudice, 681-84 (co-author) Wiener, Norbert, 594,595,596-97 Wiseman, Richard, 232,563,629-30 Rupert Sheldrake and the Objectivity of Science, 538-39 (co-author) Wolf, F. A. Reply to Costa de Beauregard, 325-27 Wright, S., 493 Yost, Casper, 250,255 Ziman, J. M., 263-64 Targ, Russell, 70-7 1,75,80 Comments on "Parapsychology in Intelligence: A Personal Review and Conclusions," 87-90 Miracles of the Mind: Exploring NonLocal Consciousness and Spiritual Healing, 129-31 (co-author) (book review) Tart, Charles T., 297,515 Tattersal, I., 551 Teller, Edward, 687 Tesla, Nikola, 499 Thomson, J. A., 495 Thurston, H., 11 Thurston, Herbert, 466 Tiller, William A. Electronic Device-Mediated pH Changes in Water, 155-76 (co-author) Towards a Predictive Model of Subtle Domain Connections to the Physical Domain Aspect of Reality, 41-67 Tinel, J., 463-64 Tombaugh, Clyde, 686-93 Tornier, E., 557 Torun, Erol, 378 Towe, Bruce C., Analysis of Technically Inventive Dream-Like Mental Imagery, 499-513 (co-author) Trizna, Dennis, 455 Truzzi, Marcello, 571 Turing, Alan Mathison, 598 Turner, Stanfield, 89 Uexkull, Thure von, 306-7 Ullman, Montague, Dreaming Consciousness: More than a Bit Player in the Search for Answers to the MindIBody Problem, 91-112 Utts, Jessica, 557,558 The Significance of Statistics in Mind-Matter Research, 615-38 Valentich, Frederick, 432-33 Vallee, Jacques, 117 Van Allen, James, 687,691-92 Index: Volume 13 Subject Index Academy of Consciousness Studies, 337 Altered states of consciousness crcativity and, 499-513 Alternative healing, 122-24, 129-31, 215 in clinical setting, 123 pepsin enzyme activity and, 139-48, 149-52,152-53 research on, 130-31 Anomalous phenomena. See also Orgonomy causality and, 292-97 significance testing of, 1-7 Astronomy, 258-59,264 alternative cosmological models, 271-90 UFO sightings and, 685-94 Atmospheric phenomena. See under UFOs and ufology Bigfoot, 549-51 Bilocation transposons and, 20 Biology, 703-4. See also Genetics Darwinism, 34648,349-50 DNA, spiritual healing and rewinding of, 146 Book reviews Anderson, J., The Shape of Things to Come, 552-53 (D. Eldridge) Atmanspacher, H. and Ruhnau, E., Time, Temporality, Now, 695-703 (K. Gustafson) Barbour, I. G., Religion and Science: Historical and Contemporary Issues, 333-35 (R. F. Creegan) Bendernagel, J. A., North America's Great Ape: The Susquatch, 549-51 (G. S trasenburgh) Chauvin, R., Le darwinisme ou l a j n d'un mythe [Darwinism, or the End of a Myth], 346-48 (L. V. Beloussov), 349-50 (S. J. Edelstein) Davis-Floyd, R., and P. Sven Arvidson, Intuition: The Inside Story, 337-41 ( M . D. Partridge) DeMeo, J., The Orgone Accumulator Handbook, 706-9 (R. A. Blasband) Edelglass, S., The Marriage of Sense and Thought: Imaginative Participation in Science, 335-37 ( J . M . Links) Ferris, T., The Whole Shebang: A State of the Universe(s)Report, 341-43 ( H . Arp) Florshiitz, G., Swedenborg's Hidden InJEuenceon Kant, 545-49 (Gregory Johnson) Griffin, D. R., Unsnarling the World Knot - Consciousness, Freedom and the Mind-Body Problem (R. B. Nordberg), 121-24 Gross, P. R. and N. Levitt, Higher Superstition: The Academic Left and Its Quarrels with Science, 554-56 (S. Krippner) Guth, A., The Injationary Universe, 341-43 ( H . Arp) Harman, W. W. and Sahtouris, E., Biology Revisioned, 703-4 ( J . Hamilton) Mizuno, T., Nuclear Transmutation: The Reality of Cold Fusion, 343-46 ( J . O'M. Bockris) Oster, E. F., The Healing Mind: Your Guide to the Power of Meditation, Prayer, and Rejection (C. Clayton), 122-24 Rees, M., Before the Beginning, 341-43 (H. Arp) Shattuck, R., Forbidden Knowledge: From Prometheus to Pornography (D. E. Hunt), 124-29 Stokes, D. M., The Nature ofMind: Parapsychology and the Role of Consciousness in the Physical World, 705 (I. Stevenson) 158 Subject Index: Volume 13 random event generators, 327-239, 329-31 Entropy, 639-64 Evolution, 258,265,346-48,349-50 creationism and, 495-96 racial prejudice and, 681-84 role of retrotransposons in, 13,21 Experimenter effect, 558,573 in mainstream scientific research, 538-39 Extra-sensory perception (ESP), 262, 562,565,566,570 belief in, and eye gaze, 222-24,227, 228 Einstein on, 328 experimental designs, 568-70 Eye gaze spiritual beliefs and experiences and, 213-29 Faith healing. See Alternative healing Fund for UFO Research, 114 Genetics. See also DNA aging studies and, 494-95 inbreeding depression and life span, 489-94 inheritance of psi ability, 351-72 racial prejudice and, 681-84 retrotransposons, and enlightenment, 9-24 Homeopathy, 263,291-315 non-locality and, 539-41,541-42 Hypnotism, 553 Intention. See also Engineering anomalies research; Parapsychology eye gaze and, 213-29 pepsin enzyme activity and, 139-48, 149-52,152-53 pH changes in water and, 155-76 Intuition, 337-41 J. Allen Hynek Center for UFO Studies, 114 Journal of ScientiJic Exploration, 268 Targ, R and Katra, J., Miracles of the Mind: Exploring Non-Local Consciousness and Spiritual Healing (S. Schwartz), 129-31 Central Intelligence Agency parapsychology and, 69-85,87-90 Chaos theory, 595-600 Clairvoyance, 552-53 of Swedenborg, 545-47 Cold fusion, 343-46 College of Psychic Studies, 232 Consciousness, 17-18. See also Alternative healing dreaming, 91-112 intuition and, 337-41 mind-brain relationship question and, 121-22,557,705 parapsychology and, 705 pervasiveness of, 703-4 quantum mechanics and, 104-6, 109 self-healing and, 122-24 spatial-temporal boundaries and, 323-25,325-27 spirituality and, 553 Consciousness-related phenomena. See also Psychophysiology effects of thoughts, prayers and wishes on, 139-48 eye-gaze and, 213-29 human-machine interactions, 155-76 quantum physical correlations of, 92, 104-7 Cosmology, 341-43 Cryptozoology. See Bigfoot DNA mitochondrial, and human ancestry, 495-86 Dreams and dreaming consciousness and, 91-112 precognitive, 552-53 Engineering anomalies research Princeton Engineering Anomalies Research (PEAR) lab, 108, 327-29,329-3 1 Subject Index: Volume 13 Kant, Immanuel Swedenborg and, 545-49 Koestler Parapsychology Unit, 562 Levitation transposons and, 20 Life spans Old Testament. 483-97 Mars "face" on, 401-2,404,407-8,413-18 Cydonian plain formations, 373-96, 397-98,398-400,401-12,413-19 UFOs and, 691-93 Meditation spiritual healing and, 122-24 transposons and, 20,22 Mediumship physical, 231-56 Mental imagery scientific invention and, 499-513 Mind-brain relationship question, 9 1, 121-22,557,705 non-Boolean logic and, 557 Mind-matter interactions alternative interpretations of, 563-65 defined, 562 Mind-matter research, 557-60,56 1-77. See also Parapsychology information spaces and, 671-72 statistics and methodology in, 558-59,572-75,615-38 study design in, 565-71 Near-death experiences (NDEs) retrotransposons and, 19,20,22 Neurolinguistic programming (NLP), 552-53 Orgonomy, 706-9 Parapsychology, 1, 129-30,228,557, 562. See also Mind-matter research consciousness and, 705 criticism of, 634-36 fraud in, 563,568 ganzfeld, 626-30,63 1-33, 636 159 intelligence work and, 69-85,87-90 limits of direct mental influence, 515-35 replicability in, 539-40,630-3 1 state of research in, 317-22 Philosophy psi phenomena and, 545-49 Physics de Broglie pilot waves, 45-48,49,54 dual four-spaces, 48-65 electromagnetism, 41,43-45,54-62, 65-67 magnetic anomalies, 43-45 multiple universes, 553 non-locality, 300-301 quantum mechanics, 42,43, 104-6, 109 quantum theory, 177-88,300-301, 337,552,602-3,672-75 theory of relativity, 271-90 Precognition dreams and, 552-53 Princeton Engineering Anomalies Research (PEAR) laboratory, 108, 327-29,329-31 conference sponsored by, 337,339, 341 Psi ability inheritance of, 351-72 Psi phenomena. See also Clairvoyance; Precognition dreaming consciousness and, 103-4, 107-8,109 environmental illness and, 542-43 philosophy and, 545-49 Psi research. See Engineering anomalies research; Parapsychology Psychokinesis (PK), 28-39,562,565-66 experimental designs, 570-7 1 large-scale, 567 macro, 231-32, 233 micro, 537 Psychological tests Minnesota Multiphasic Personality Inventory, stigmatization and, 479 Psychology state of research in, 317-22 160 Subject Index: Volume 13 Service dlExpertisedes Phenomenes des Phenomenes de Rentrees Atmospheriques (SEPRA), 422 Sitter groups, 231-56 Society for Planetary SETI Research (SPSR), 399,413,414,417-18 Society for Psychical Research (SPR), 232,234,236 Society for Research in Rapport and Telekinesis (SORRAT), 231-56 Society for Scientific Exploration (SSE), 265,267-68 Pocantico workshop, 421-22,455 Solar energy, 502-9,5 11 SORRAT. See Society for Research in Rapport and Telekenesis SRI International, 633 Stanford Research Institute (SRI) CIA-sponsored psi research at, 7 1-74,79,80-8 1,129,130 Statistics and methodology, 149-52, 152-53,557-60 Bayesian analysis, 2-6,559 Boolian logic, 586-89,600-603 fractional Brownian motion, 27-28, 39 Hurst effect, 25-40 information spaces, 665-79 large deviation statistics, 639-64 meta-analysis, 531,540,541,559, 624-25 Percentage Score Index (PSI), 518-19 probability theory, 557-58,559, 579-613,665-79 range analysis, 25-40 replication, 539-40,541-42 segregation analysis, 363-66 simple radar equation, 456 Stigmata and stigmatization, 461-82 Subtle energies, 41-67 Survival after death, 10, 11. See also Reincarnation belief in, and eye gaze, 223,228 Swedenborg, Emanuel Kant and, 545-49 Synchronicity, 572,603 homeopathy and, 301-5 Psychophysical Research Laboratories (PRL), 627,629,633 Psychophysiology limits of direct mental influence on, 515-35 stigmata and, 461-82 Racial prejudice color perception and, 681-84 Random event generators, 228 Reincarnation, 116,262 cases in Turkey a generation apart, 189-98 Religion Old Testament life spans, 483-97 stigmata and stigmatization, 461-82 Remote perceptionlviewing, 129-30, 228,569,624,632,633,636 Royal Society, 267 1919 joint meeting with Royal Astronomical Society, 278-82 Sasquatch. See Bigfoot Science censorship of discordant results, 267, 27 1-90 commercial funding for frontier, 537 definitional problem, 257-70 epistemology, 125-29,335-37 experimenter effects in research, 522, 525,530,538-39 history of, 557 intuition in, 339-40 limits on research, 125-29 mental imagery and invention in, 499-513 philosophy of, 257-70,335-37,557, 703-4 religion and, 333-35 sociology of, 554-56 Search for Extraterrestrial Intelligence (SETI), 402,414. See also Society for Planetary SETI Research (SPSR) Second sight inheritance of, 351-72 Semiotics homeopathy and, 305-10 Subject Index: Volumes 13 precognitive dreams and, 552-53 Time conscious experience and, 323-25, 325-27 philosophy of, 695-703 quantum mechanics and, 695-703 UFO sightings, 685-94 New Zealand, 422,431-58 Roswell (New Mexico), 113-16 Switzerland, 422,424-31 Ubatuba (Brazil), 1 17 UFOs and ufology abductions, 551 atmospheric phenomena, 421-59 Colorado UFO report, 422 experiencers, 22 GENAP, 117 GEPANISERPA project, 1 17 Hessdalen Project, 1 19 photographs, image analysis of, 199-211 Pocantico workshop, 116-20, 421-22,455 retrotransposons and, 20,22 Unicat Project, 117 Society for Scientific Exploration Officers Prof. Peter A. Sturrock, President Varian 302 Stanford University Stanford, CA 94305-4060 Ms. Brenda Dunne Executive Vice President for Education C 131, School of Engineering &Applied Science, Princeton University, Princeton, NJ 08544-5263 Prof. Robert Jahn, Vice President D334, School of Engineering & Applied Science Princeton University Princeton, NJ 08544-5263 Prof. L. W. Fredrick, Secretary Department of Astronomy P. 0.Box 3818 University of Virginia Charlottesville, VA 22903-08 18 Prof. Charles R. Tolbert, Treasurer Department of Astronomy P. 0.Box 3818 University of Virginia Charlottesville, VA 22903-0818 Council Dr. Marsha Adams 1100 Bear Gulch Rd. Woodside, CA 94062 Prof. Henry H. Bauer Chemistry, 234 Lane Hall VPI & su Blacksburg, VA 2406 1-0247 Dr. Stephan Baumann fMRI Project Director Psychology Software Tools, Inc. 2050 Ardmore Blvd. Pittsburgh, PA 15221 Prof. John Bockris Department of Chemistry Texas A&M University College Station, TX 77843 Dr. John S. Derr Albq. Seismology Center Albuquerque, NM 871 15 Dr. Roger D. Nelson C 131, Engineering Quad. Princeton University Princeton, NJ 08544 Dr. Harold E. Puthoff Institute for Advanced Studies-Austin 4030 W. Braker Ln., Suite 300 Austin. TX 78759-5329 Dr. Beverly Rubik Institute for Frontier Science 6114 LaSalle Avenue, #605 Oakland, CA 94611 Dr. Robert M. Wood 1727 Candlestick Ln. Newport Beach, CA 92660 Journal o Scientific Exploration, Vol. 14, No. 2, pp. 163-194,2000 f 0892-3310100 O 2000 Society for Scientific Exploration Overview of Several Theoretical Models on PEAR Data Princeton Engineering Anomalies Research Princeton University, Princeton, NJ 08544 Abstract-Existing PEAR datasets on microelectronic experiments are reviewed for their consistency with several theoretical models. The analysis includes a comparison of the observed data with predictions from bitwise-effect (BIT), Decision Augmentation Theory (DAT), time-normalized (TIM), and teleological (TEL) models for the phenomena. Methods for constructing a model comparison are discussed, together with the merits of relative versus absolute tests and the sensitivity of various test representations. Final results are presented in both frequentist (p-value) and Bayesian (hypothesis odds) formats. Keywords: humanlmachine interaction-Decision Augmentation Theorymodeling methods-analysis methods 1. Terminology The following terms will be encountered frequently: Z score. This is a normalized score used as a statistical figure-of-merit for the degree to which an observation differs from a theoretical prediction. In general, a Z score is defined by Z = (x, - xt)lat, where x, is the observed value, x, is the theoretically predicted value, and a, is the theoretical uncertainty (standard deviation) of the measurement. If the measurement error is normally distributed and the theory correctly describes the experimental situation, one expects Z to follow the standard normal distribution with mean 0, variance 1. A large value of Z is evidence against the theoretical model under which it is calculated. When different models make different predictions for the same observation, a Z score can be calculated for each model. EfSect size. If some number N of identical, independent observations is made, the mean of those observations has a theoretical standard deviation of/*, where at is the theoretical standard deviation of a single observation. proTherefore the Z score of the collected database will tend to grow as vided the same effect (i.e., the same quantitative degree of departure from the theoretical prediction) is present in each observation. This means the Z score is not directly useful for comparisons between experiments because differences in Z may simply reflect different amounts of data in situations where the same phenomenon is observed. To get around this problem one customarily defines n, I 164 York H. Dobyns an effect size c = Z I ~ It .is obvious that E will remain constant as N increases, provided the repeated observations remain mutually consistent. Dataset and database. In general database refers to the entire collection of experimental data under consideration, while a dataset is a subset of the total database extracted according to some analytical criterion. For example, a parameter relevant to a particular theory, which takes on several discrete values in the database as a whole, may be used to segregate the data into several datasets for analysis. 2. Introduction The current analysis focuses on a subset of data from PEAR published in the Journal of Scientific Exploration (Jahn et al., 1997). Of those data, this analysis will examine only those obtained from microelectronic random sources; in the interests of a consistent database, we shall not include the data from deterministic sources or those from the macroscopic, mechanical experiments. A partial reproduction of Table 2 in Jahn et al. (1997), presenting the relevant data, is given as Table 1 below. The primary consideration in an anomalies experiment is, of course, the question of whether an anomaly is present. Our main tool for examining this will be a Z score against the null-hypothesis prediction. However, different models of the nature of the phenomenon lead to different formulas for calculating the composite Z score across several datasets, for reasons that will be discussed below. Thus, when there are competing explanations, it may be necessary to evaluate them against each other before attempting to make a definitive, quantitative evaluation against the null hypothesis. We shall examine four hypotheses that have been proposed as models for the observed phenomena. Any database of this sort is necessarily a combination of efforts from many individual experimental sessions. There are variables that cannot be controlled effectively in experiments in which the data must be generated by human volunteers at their own convenience. A single experiment combines data from many individuals with different personal characteristics, at different times of day and year, and so forth. In attempting to compare overall models of the phenomena, one must assume that such uncontrollable variables are not conse- TABLE 1 Data by Experiment Experiment Original Remote Alternate source Co-operator REG2000 REG20 Series 522 212 46 45 44 20 Bits 3.35 x 1.83 x 4.94 x 3.62 x 3.25 x 1.64 x lo8 lo8 lo7 lo7 lo8 lo6 Z 3.809 2.214 2.765 1.635 2.7 18 -0.956 Models on PEAR Data 165 quential to the phenomenon, or at least that their effects are averaged out over the accumulation of a large database. When we say that a model predicts that some measure or other should be "constant across all experiments," we naturally mean this to include the assumption that there are no consequential factors that systematically change between one dataset and another. In fact, we will illustrate one instance in which the presence of such a confound can lead to an incorrect model evaluation if it is not properly taken into account. All of the models discussed here are phenomenological; that is, they discuss what takes place under certain experimental conditions, without addressing the question of how it takes place. For example, the bitwise-effect model explains the anomalous observation as the result of a change in binary probabilities; that change is assumed, and no explanation is offered for how a human consciousness could alter the probability of a physical outcome. It seems advisable to compare such proposed models, even though they offer no deep theoretical insight, because it is unlikely that the deep questions can be addressed adequately while we are still uncertain about the phenomenology. 3. The Models Bitwise effect. The observed phenomenon consists of a shift in the mean output value of a binary random number generator, in accordance with the human operator's prestated intention. This can be modeled as a change in the probability of the random process, which favors the outcome in accordance with the operator's intention. The bitwise effect model simply states that this is in fact what happens; the probability of a binary outcome is changed by some small amount in accordance with intention. The natural unit for the bitwise model is the bit. The effect size measure, under the bitwise model, is Z I ~ where Nb , is the number of bits. The bitwise model predicts that this measure should be constant across experiments. For brevity, we will use the mnemonic BIT to refer to this model. Decision Augmentation Theory. This theory, commonly abbreviated DAT even by its authors, suggests that no change to the performance of the experimental apparatus actually takes place (May, Utts, and Spottiswoode, 1995). Instead, they suggest that the human operator is able to foresee, unconsciously and by anomalous means, the outcome of an experimental run. The operator thus can choose a moment to start the run when random fluctuations will produce a favorable outcome. The natural unit for this model is the "DAT event" of a single start-of-data-collection by the operator. The effect size measure is Z I ~where N D is the number of DAT events in a single dataset; DAT pre, dicts that this should be constant across experiments. Time normalization. This model suggests that the critical variable in an anomalous experiment is the amount of time, or equivalently the amount of effort, contributed by the human operator (Nelson, 1994). The natural effect size measure is where N, is the number of seconds spent generating data. ~/a, 166 York H. Dobyns The time-normalization model predicts that this should be constant across experiments. The three-letter mnemonic we will use for this model is TIM. Teleological. Several models have been discussed internally at PEAR that predict a decline over the course of a single experiment, specifically that the effect size should decline as l / f l , where N, is the total amount of accumulated data in the experiment to date. These models are collectively called "teleological" since the first such proposal was based on teleology at the level of individual experiments: It proposed that any experiment rapidly attains a "characteristic" overall Z score, which then holds more or less steady no matter how much further accumulation of data takes place. Unlike the previous models, this one does not have a "natural" effect size associated with it; rather, it predicts that the locally measured effect should decline in a specific way as data accumulate. It also does not make predictions across experiments, but it can be evaluated only by examining the sequential evolution within a single experimental dataset. The mnemonic for this model is TEL. 4. Comparing Models It is easiest to compare models in pairwise fashion. In general, such a comparison requires identifying a variable such that (a) data with different values of this variable are present in the database and (b) the models being compared make different predictions for the functional dependence of the effect on the variable. We shall make all our pairwise comparisons between theoretical models by comparing the BIT model against another model. This is primarily a matter of convenience and historical preference: The original data collection systems used an implicit bitwise model in their assumptions, and this is still the most convenient representation for analyses. For the comparison of BIT against DAT, the key variable is the sequence length, or number of bits per DAT event. The data under consideration have eight different sequence lengths, ranging from 20 to 200,000. We shall refer to this sequence length variable as n; it is related to Nb, the number of bits in a given dataset, and ND,the number of DAT events, by the obvious Nb = nND. As noted above, BIT and DAT identify different increments of data as fundamental, and the specific functional dependence upon n will depend on which of these is used. BIT predicts that the bitwise effect size, tb= Z I ~ is constant across datasets, while DAT predicts the same of the DAT effect A comparison of these formulas will immediately show size tD= that cb = tD/l/i;. Thus, if we adopt ebas our measure, BIT predicts that the effect is constant, and DAT predicts that the effect varies as I/+. Conversely, if we measure effects by ED,then DAT predicts constancy and BIT predicts an effect that increases as &.We will have some further comments on this duality of representation later. A precisely similar situation applies between BIT and TIM except that the key variable is now B, the number of bits per second, forming the connection , ~/m. Models on PEAR Data 167 Nb = BN,. As above, each model predicts that its own measure is constant; TIM predicts that rb oc 1 / a , while BIT predicts c, oc For the comparison of BIT versus TIM, we have three available bit rates: 20, 200, and 2,000 bits per second. The data in the first and last of these categories are somewhat scanty, so we may expect that this issue will be resolved much less clearly than the previous comparison. TEL is somewhat different in that it does not present an alternative fundamental unit for the local interpretation of data; rather, it predicts a specific course of historical change over the accumulation of data in a particular experiment. This prediction applies to any of the effect size measures discussed in the previous paragraphs. For a pairwise test of BIT versus TEL, the obvious choice is to use rb; then BIT predicts that the value should be constant, while TEL predicts rb oc I/&. Finally, there is the question of a null-hypothesis comparison. The null hypothesis simply predicts that there is no effect: rb = ED = c, = 0 . Nonetheless, because the different models weight contributing datasets differently, the confidence with which one can assert E # 0 differs from model to model. The procedure we will develop for the pairwise comparison between models includes, as a byproduct, the generation of a Z score against the null hypothesis for each model involved. Our ultimate candidate for an interpretive test against the null hypothesis should be the preferred model emerging from the pairwise comparisons. a. 5. Comparison Procedure In each of the pairwise comparisons, we are confronted with a set of empirical data and a key independent variable: One model predicts that the data should be independent of that variable, and the other predicts that there should be a specific functional dependence. The simplest evaluative procedure at this point is a goodness-of-fit test. We know, for example, that BIT predicts rb = kb, while DAT predicts eb = k D / f i , where kb and kD are undetermined constants of proportionality. We can find empirical values for these constants by minimizing the total squared error between the model and the observed data, and then calculate the x2 of the residual differences between the observations and the fitted model. This is not, however, the most sensitive test available in the current situation. A standard and often-used procedure for such tests is to linearize the model and then perform a linear regression to identify the slope. In this particular case, "linearizing" means not discarding higher-order terms, but constructing a new variable such that the model predictions are linear functions of that variable. We will continue to use the BIT versus DAT comparison as an initial example. If we work in the eb representation, then the DAT model predicts a dependence on 1/fi. Therefore, let us define x = I/&; then the BIT prediction 168 York H. Dobyns described above remains unchanged, and the DAT prediction becomes Eb = k D x . These models then can be compared directly with a weighted linear regression performed on the observations of cb versus X . Indeed, because we must perform a least-squares fit to find the empirical constants of proportionality, each of the model fits is a special case of a constrained linear regression. The general linear regression will find parameters b l , a slope, and bo, an intercept, such that the line y = bo + b l x minimizes the overall squared error C ( c - y ) 2 . The constant model is a special case of a linear least-squares fit with the slope constrained to be 0, and the model proportional to x is a special case of a linear model with intercept 0. The parameter values found by the linear regression, as optimal fits to the data, can be compared with the values found for the constrained models to establish which model is the better fit for the data. To compare the sensitivity of the linear-regression test with the goodnessof-fit test, a Monte Carlo procedure was used comparing the efficacy of both tests on synthetic data generated from both constant and linear models. For the linear model, a Z score for the slope against a nominal value of 0 was used. This was found to be correctly distributed in constant models and to grow with the size of the effect in linear models. The x2goodness-of-fit test also correctly identified which model had been used to generate the data but generally achieved a lesser degree of statistical significance than the regression test on the same data. This loss of significance was equivalent to that produced by a 17% deflation of the Z score in the regression test. It therefore was decided that the linear regression tests would be used as the primary evaluator for these hypothesis comparisons. The x2goodness-of-fit tests, however, retain a secondary utility. The regression test must always report some values for its parameters, whether the data are actually linear or not. Comparing these with the model predictions can identify a preferred model but will not give a warning if neither model is a good fit to the data. The x2tests, on the other hand, can identify cases in which neither model fits the data well. If x2is large for both models, some further source of variation, not well described by either model, is present. 5.1 Formulas Each subdivision of the overall database produces a group of datasets, each with its own Zi and Ni, where the Ns may count bits, DAT events, or seconds as appropriate to the effect size measure in use. The ith dataset thus has an effect size observation ci = Z i / f l . In the description of Z scores above, it was noted that Z is always normalized to the standard deviation, or measurement uncertainty, of the observation in question. Thus, a Z score must by construction have a standard deviation of 1. The uncertainty in the effect size ci is therefore ai = 1 I f l . In addition, there is a set xi of values of the key model-discriminating Models on PEAR Data 169 parameter. After suitable definition of x, we are always comparing two models, one of which predicts ri constant, while the other predicts ri cx xi. The constant model is E" = kc where kc = - 1 (1) Here o(kc) denotes the overall uncertainty on the value of kc that propagates from the individual measurement uncertainties of the ri. The alternative model always predicts a linear dependence E" cx x; we will call this the slope model to distinguish it from the linear model produced by the empirical two-parameter regression. The formulas for this model are: i = ksx where k, = c xi yi x+; 1 , 0' and o(ks) = 1 The general two-parameter regression, on the other hand, involves formulas that are rather complicated and can be understood more readily in an incremental presentation. If we first define the intermediate quantities, we then can simply define the linear parameters: E" = bo where i + blx, bl=Ccl,i~i; i bo=C~o,i~i and o ( b o ) = J m ; o ( b l ) = J m . 5.2 Problem with the Simplest Approach It might seem that the linear regression provides a direct and immediate test between the two models; because one predicts a slope and the other does not, we need only examine the slope parameter (bl) for statistical significance. This approach has even been used in other work comparing linear models (May, Utts, and Spottiswoode, 1995). However, Figure 1 illustrates a disconcerting feature of such a direct comparison. In Figure la, the eight datasets at different values of the DAT sequence length n are plotted against I/&; the vertical axis is the bitwise effect size rb. Figure l b uses the dual representation, plotting rd against &. The two dotted lines in each graph show the two models; the solid slanting line shows the regression fit. In the first graph, the slope of the regression line seems rather am- 170 York H. Dobyns 0 100 200 300 400 Square root of sequence length DAT=Constant, BIT=Slope Fig. 1 . Model comparison in two representations. biguous between the two predictions; in the second, it is extremely close to one of the predictions. It would seem that the representation in the first graph allows no very strong preference between models, while the representation in the second graph definitely supports one. But it seems very odd to arrive at such different conclusions, when one considers that we are examining the same two models on exactly the same data; only the representations differ. Some thought will demonstrate that a simple comparison, whether visual or numeric, between the slopes of the regression line and the slope model is suspect. Both slopes are obtained from a least-squares fit to the same data; a com- Models on PEAR Data 171 parison between them, using the error figures in the formulas above, is liable to overestimate the variance of the difference and thus underestimate the significance of that difference. That this is indeed the case was established by a systematic Monte Carlo evaluation of the slope-comparison procedure. The normal statistical interpretation of a Z score requires that it be normally distributed with mean 0 and variance 1 when the model it is testing is true. So to determine the validity of a slope-comparison Z score, the Monte Carlo program constructed synthetic datasets according to the requirements of a slope model. Figure 2 shows the result of a 10,000-iteration Monte Carlo evaluation of the slope-comparison Z score (bl - k , ) / a (bl) for data spaced as in the BIT versus DAT comparison. The points with error bars show the empirical population density, the solid curve shows the corresponding normal fit, and the dotted curve shows the standard normal probability density. We can see that the Z score remains normally distributed, but its standard deviation is much too small, about .48 rather than 1. 5.3 Vector Formulation How, then, can we obtain reliable tests of the two models? A certain amount of analytical development is needed to demonstrate the proper tests. Let us, first, reprise the situation: For any model comparison, after identifying the key discriminator variable in the database, we have a set of observations ci of the effect size, each with uncertainty ai, obtained at a key variable value xi. We have, in essence, four models to consider. The null hypothesis contends that -2 -1 0 Z score 1 2 Fig. 2. Monte Carlo population density for slope-comparison Z score. 172 York H. Dobyns there is no effect and that all of the observations differ from 0 only by noise. The constant and slope hypotheses have been discussed in the previous section, and the linear regression formulae were given in Equation 3 amount to the assumption of a two-parameter model. By explicitly adding a noise term v i , assumed to be normally distributed with mean 0 and standard deviation ai, we may express the models in terms of predictions for individual observations: (Null) (Constant) (Slope) (Regression) : €i = Vi = kc = kSxi Vi : : : Ei Ei €i + vi = bo + blxi + vi. (4) Now let us change notation by dividing each of these equations by ai. Defining the new variables, allows us to re-express Equation 4: (Null) (Constant) (Slope) (Regression) : zi = ~i : zi = kc(c,i : Z i = ks(s,i : zi = bo(c,i + + Ei Ei (6) + bl(s,i+ E i . Note that the new noise terms, c i , all have variance 1, from their definition in Equation 5. Also, because we assumed that the initial noise terms, v i , were normally distributed, the normality of the error terms will continue to propagate through subsequent analyses. (This assumption of normality is quite accurate for the empirical data presently under consideration.) To alleviate an excess of subscripts, we may take advantage of the fact that the model equations can readily be written in terms of vectors rather than individual indexed equations. Thus, for example, 2 is the vector whose individual elements are the zi corresponding to individual observations: 2 = { z i ) = {zl, z2, . . . , z m ) ,where there are a total of m observations. With similar notations for the other indexed variables we may write (Null) : (Constant) : (Slope) : (Regression) : = i = kc& + 2 i = k,& + 2 2 = bogc + bl gS+ 2. (7) Before proceeding further, let us recall the basic operations of vector arithmetic. If we have two vectors x' = { x i } and y' = { y i } ,the vector sum, x' + y' = {xi + y i ), is the vector obtained by adding the components of the two vectors. Models on PEAR Data 173 The product of a scalar (simple number) with a vector, as in kx' = {kxi),is the vector obtained by multiplying each element of x' individually by k. The inner product, -? . 9 = Cixiyi, is the result of multiplying each element of 2 by the corresponding element of y' and adding up all the resulting products. The length of a vector (x'( can be found by taking the square root of its inner product with itself: -? - x' = 1-?12. more useful result is not a vector identity but One follows from applying these formulas. Because the zi defined in Equation 5 have been normalized to the observational uncertainty, they all have a variance 1. The vector 2 used in Equation 7 therefore has a variance 1 in each element: 02[i] . Such a vector has the property that the variance of its inner =i product with any constant (i.e.,nonrandom or zero-variance) vector x' is This can readily be verified by explicitly writing out the inner product and applying the standard rules for variances of sums and products. All of the formulas in Equations 1 through 3 can be reexpressed in this vector notation, with exactly the same values for the model parameters, kc, k,, bo, and bl. Let us begin by considering a least-squares fit to the constant model. We note from Equation 7 that both the constant model and the slope model have exactly the same functional form, 2 = kg + Z , differing only in a change of subscript; therefore, a single derivation will do for both. Because the noise tern, 2, is unknown, we can solve for it as the error between the model prediction and the empirical value: 2 = 2 - k$. The optimal value of k is the one that minimizes the length, or equivalently minimizes the squared length, of this error. This squared length is To find the minimizing value of k, we take the partial derivative a(l;12)/ak and set it to zero: By substituting the definitions of Equation 5 for zi and (c,i, we find that this gives us Equation 1 for kc; by instead substituting {sti, we recover Equation 2 for k,. By applying the variance identity (Equation 8) to the last line of equation 10 above, we find York H. Dobyns which because l$cl = and I$,1 = appropriately reproduces Equation 1 for (kc) or Equation 2 for o(ks), depending on whether we , insert f C or $ for $ in the last line of Equation 11. The vector formulas for bo and bl are likewise identical, whether one rewrites Equation 3 in terms of 2, gc, and or performs a least-squares-fit calculation directly on the model formula in Equation 7. Because the vector notation avoids the tedious enumeration of indices and summation signs, we present the vector forms directly rather than through the intermediate step of the co,1 coefficients: ,/= ,/6 = y' . x'), we may note from Because the inner product is commutative (x' Equation 12 that bo and a(bo) differ from bl and o(bl) solely through the systematic exchange of c and s subscripts. There is one more set of quantities of interest that can be computed directly from Equations 10 through 12. These equations give values and uncertainties for the four parameters kc, k,, bo, and b l . For each parameter, we can therefore calculate a Z score against the hypothesis that its value is 0. From Equations 10 and 11, we can calculate that which can be applied to either k simply by attaching the appropriate subscript. The values of Zo and Z1 corresponding to tests of nonzero value on bo and bl are from Equation 12: Models on PEAR Data 175 , As we might expect from the earlier observation on Equation 12, Zo t Z1 upon substituting & t* &. 5.4 Abstract Modeling Problem Let us temporarily take our leave of the complicated least-squares-fit formulas and consider a very abstract representation of modeling data in multiple dimensions. We start with the assumption that we have a set of observations, which have been normalized such that each observation has unit variance. We summarize the set of observations as a vector ? with as many dimensions as we have observations. Because ? has been constructed to have unit variance, we can express it as two components, Z = (2) + 2, with a deterministic part (?) and a random c_omponent 2 defined as in the previous section: (2) = 0 , a2[2] = a2[1] = 1. Let us presume further that we have two vectors, G1 and G2, which are selected on theoretical grounds as being plausible models for the phenomena. The observed results might be explained by the first model, by the second, or by some linear combination of the two. We can summarize all three possibilities, along with the null hypothesis, by these four formulas for (2): Ho : (Z) = 0; HI : (2) = alG1; H2 : (1) = a2G2; H3 : (2) = BiGi + 8232. Figure 3 illustrates a two-dimensional example of this formalism, with the observation vector ? and the theory vectors GI and 32 shown in solid black, while one speculative possibility for the model vector (2) and the noise vector 2 are drawn in gray. The equations (Equation 15) form a hierarchy, in the sense that lower-numbered models are contained within higher-numbered ones as special cases: Ho is a special case of either H I or H2, while either of the latter is a special case of H3. In each case the simpler model is produced by setting a parameter of the more complex model to 0. The questions we wish to answer are all of the form of asking whether this in fact holds true. Within the context of HI or H2, is a = 0 so that Ho is actually true? Within the context of the two-parameter model H3, is one of the parameters actually 0 so that the model reduces to HI or H2? It is relatively straightforward to find an appropriate test for the null hypothesis Ho within the context of either H I or H2. First, noting that H I and H2 are 176 York H. Dobyns Fig. 3. Example of theory and observation vectors. identical in functional form, differing only in subscript, we may for purposes of the derivation drop the subscript and simply work with (2) = a;. Since the observation vector 2 is constructed to have unit variance, a properly distributed Z score for Ho can be constructed by taking the parallel projection of ? along the hypothesis vector: that is, the component of the length 121 in the direction defined by the hypothesis vector v' . This projection is We can readily verify that this formula gives us a valid Z score for Ho. The expectation and variance of Equation 16 are given by: making use of the variance identity (Equation 8) and the model substitution (2) = a;. We see that the variance of Z is 1, as required, while the expectation of Z is proportional to a, so that it is 0 when Ho is true and non-zero when Ho is false. The question of evaluating whether the general two-parameter model H3 reduces to a one-parameter model is more involved. If one simply looks at the projection of along one of the hypothesis vectors, one is repeating Equation Models on PEAR Data 177 16 and computing a test statistic for Ho within H 1 or H2, rather than computing the desired test statistic for H 1 or H2 within If3. The fundamental problem is that we have no guarantee that the vectors S1 and G2 are orthogonal. (Indeed, in our intended application, they are not.) Given nonorthogonal hypothesis vectors, simple projections o f ? along and G2 will both have nonzero expectation if either has a nonzero coefficient in reality. If, let us say, the actual model is H I , so that (Z) = alGl by hypothesis, the expected value of Z i = 2 51 / 1 GI I is (Zl) = a 1 I GI I. The expected value of Z2, on the other hand, is (2) . ;2/1521 = a 1 Gl ;2/1 s 2 1 ; this vanishes only if GI - G2 = 0, i.e., if the two hypothesis vectors are orthogonal. For the general case with nonorthogonal hypotheses, both (Z1)# 0 and ( 2 2 ) # 0 regardless of which hypothesis is true. Clearly neither Z 1 nor Z2 can be a satisfactory test against the hypothesis that one of the p parameters in H3 is 0. For definiteness, let us consider the case in which we wish to test whether p2 = 0, so that H3 reduces to H I . (Since the notation is completely symmetric, we can do our derivations once and then get the formulas for the other case by interchanging 1 t 2.) To obtain a Z score against this possibility, we need to , know the extent to which the observation ? requires a component not simply proportional to G1 ; that is, we need the projection of ? not along G2, but along that component of s2orthogonal to 3 1 . Let us label this orthogonal component 6 2 ; its value is given by s1 One can readily verify that Gl . 6 2 = 0, so that G2 is indeed, as we require, orthogonal to G l . The Zscore against p2 = 0 is then just the projection of? in the G2 direction: Let us verify that this does in fact have variance 1 and expectation 0 if, and only if, p2 = 0. We see that the proposed Z score against p2 = 0 has a variance 1 always and has an expectation value proportional to p2; therefore, it is Z-distributed if and only if /32 = 0, as required. If we choose to expand 6 2 in terms of its definition (Equation 18), so as to express all quantities in terms of the original hypothesis vectors and s 2 , the form of ZB2from Equation 19 becomes s1 178 York H. Dobyns zp2= -- 2 I"' ij2 i [ G ~.( + ( [G2 - - ) - Hjl) il.i2 + -2 If we multiply the numerator and denominator of the last line of Equation 21 by (GI 12,we finally obtain 1 I with a corresponding Zp obtained by interchanging 1 t , 2 in Equation 22: Equations 22 and 23 should look very familiar in form. The equations in Equation 15 were deliberately laid out to suggest the form of the equations in Equation 7; Ho, HI, H2, and H3 are identical to the null, constant, slope, and regression models, respectively, if one identifies the symbols: These identifications make it very clear that Equation 16, giving the Z score for a one-parameter model against the null hypothesis, is the same as Equation 13 derived from least-squares parameter evaluation, and that likewise Equations 22 and 23 for the Z scores of the parameters of the two-component model are identical to Equation 14 for the Z scores of the regression parameters. These identities, particularly those equating Equations 22 and 23 with Equation 14, are critically important. We have already derived and demonstrated the fact that Zgl and Zp2 are correctly distributed, properly normalized, and independent Z scores for the evaluation of their respective hypotheses. We now see that the least-squares method of parameter evaluation automatically generates the extra, orthogonalizing terms required to make sure that Zgl tests only the parameter Dl regardless of the value of P2, and vice versa. Because we made no particular assumptions about the values of or G2 in the derivation from Equations 15 through 23, this derivation of the validity of the Z formulas is entirely general; in particular, it holds despite the fact that our regression formulas (Equation 3) use the origin of coordinates, rather than the mean of the xi values, as the origin of the regression. One small subtlety needs to be kept in mind when evaluating these models. Models on PEAR Data 179 The Z scores of Equations 22 and 23 are Z scores against the null hypothesis that their respective parameters are 0. To consider Zg2 (which is Z1 in the notation of Equation 14), this is a Z score against the hypothesis that p2 = 0 (bl = 0), in which case H3 (regression) reduces to H1 (constant). Thus, the Z score of the slope is the test that can refute the constant model, by demonstrating a nonzero value of the parameter that must be 0 if the constant model is true. Likewise, the Z score of the intercept (Zgl, Z o ) is the Z score against H2 (slope model). Despite the power of the vector formalism, clarity of visualization requires that the original notation of €i, xi, and a be retained for illustrative purposes. i Because graphs on paper are inherently two-dimensional, and most readers are three-dimensional, it grows very difficult to render graphically vectors in spaces of four or more dimensions. It is true that any two arbitrary vectors can define a plane, so that a representation such as Figure 3 can show the relations of the theory vectors to each other and to the projection of the data into their common plane. However, such a compressed representation makes it impossible to visualize the actual data for most analyses. Therefore, the analyses in subsequent sections will retain the familiar graphs in xi, Ci space as well as the vector representation developed here. 5.5 Resolution of the Quandary With this new approach to calculating the Z score between the linear fit and each model, we find a fully symmetric duality between the two representations of the data in any model comparison. If we switch from one model's definition of the fundamental effect size to the other, the role of constant model and slope model are exchanged. But all of the statistical parameters also change places in a symmetrical fashion. The Z scores of the model parameters against the null hypothesis, of the linear regression against each model, all change places so that the same statistical figure-of-merit is associated with the same model. Even a X 2 goodness-of-fit test against the residuals from the model displays the same behavior. All of the statistical measures are associated with a specific model, regardless of the data representation used (see Table 2). TABLE 2 Duality of Representations I i Parameter z/A effect l/fi BIT DAT 5.80 3.97 1.30 4.43 z effect I ~ X axis Constant model Slope model Constant Z vs. null Slope Z vs. null Regression Z vs. constant Regression Z vs. slope X$ vs. constant x;vs. slope DAT BIT 3.97 5.80 4.43 1.30 fi 180 York H. Dobyns This exact equivalence between representations is, of course, further evidence that we are justified in using the techniques discussed in the preceding section: The comparison between two models and a set of data should be independent of the choice of representation. Having established that both representations are exactly equivalent when properly analyzed, we are now free to choose either representation without concern that our choice biases the conclusion. The remaining figures therefore will display the BIT representation exclusively. 6. Model Test Outcomes Although we have already seen some of the DAT versus BIT comparison results, since it was the example used in the previous section on developing the comparison methodology, we will cover the matter in more detail here. First, there is the matter of the raw data. The experimental summary in Table 1 is inadequate, because several of the experiments enumerated collected data at more than one sequence length, while in other cases more than one experiment used the same sequence length. Table 3 shows the result of collecting the data according to sequence length, the key variable for the DAT versus BIT comparison. When one calculates the statistical parameters for the comparison between models-which, as noted in the previous section, are the same whether one TABLE 3 Data for DAT vs. BIT n 20 200 1,000 2,000 10,000 20,000 100,000 200,000 Nb ND lo5 lo7 lo6 lo7 lo8 lo7 lo8 10' 1,600 163,350 1,200 24,340 15,182 3,522 2,040 2,106 Z 3.2 x 3.267 x 1.2 x 4.868 x 1.5182 x 7.044 x 2.04 x 4.212 x Note: DAT = Decision Augmentation Theory; BIT = bitwise effect. TABLE 4 BIT versus DAT Comparison Comparison BIT vs. null (2) DAT vs. null (Z) BIT vs. regression (2) DAT vs. regression (2) 2 X7 on BIT 2 X7 on DAT Score P Note: BIT = bitwise effect; DAT = Decision Augmentation Theory. Models on PEAR Data 181 uses z/A Z or J as one's measure of effect size-one finds the values ~ presented in Table 2, represented here in Table 4 for more convenient model identification. These outcomes are summarized graphically in Figures 4 and 5. Figure 4a shows the familiar regression graph, and 4b shows the various model parameters directly so as to facilitate comparison. Figure 4b plots model slope against model intercept, in units of the empirical uncertainty associated with each measurement. In other words, it plots the Z vs Constant: 1.303 ; Z vs Slope: 4.427 0 2 4 6 Intercept Z score 8 Fig. 4. Model comparison: BIT = constant and DAT = slope. 182 York H. Dobyns (b): Normalized Residuals Fig. 5. BIT constant and DAT slope as vectors. Z scores of the two models. The null hypothesis is represented by the filled dot at slope 0, intercept 0. The horizontal dotted line represents the family of all constant (zero-slope) models. Recall that in the current data representation, these are BIT models. The open circle on this line shows the actual model in this family that is the best fit to the data. The vertical line shows the family of all DAT-like, zero-intercept models. The open diamond on this line shows the actual DAT model that best fits the data. The two-parameter linear regression fit also is shown, but not directly as a Z score. To provide a valid visual comparison, its slope and intercept are plotted to the same scale as the slope of the slope model and the intercept of the con- Models on PEAR Data 183 stant model. Because the two-parameter regression in general has larger uncertainties on each parameter than the one-parameter models, this means that the position of the regression fit cannot simply be read off the axes to show a Z score. To provide a visual cue to the degree of uncertainty in the regression, 1 o and 2ocontours have been drawn around the fit point. We can clearly see that the general linear regression is quite close to the constant, BIT model, and very different from the entire DAT family of models. Figure 5 illustrates the vectorial representation of the problem. As noted in section 5, any two (nonparallel) vectors define a plane, however high the dimensionality of the space in which the vectors themselves exist. Figure 5a presents a view of this "theory plane" defined by for BIT and & for DAT. The two theory vectors themselves are drawn at unit length (the absolute scaling of this space is irrelevant to the theoretical comparison), with dots drawn at every unit increment of the vector length to help the eye project the _vector's direction. Shown as a closely dotted line is the observation vector Z, or rather its two-dimensional projection into the theory plane. This representation, free of the visual biases induced by a regression plot, allows an immediate appreciation of how much more closely the observation falls along the direction than the direction. Finally, Figure 5b deals with the residual vector 2 = ? - (bogc bigs), that part of the observation vector ? that is not accomodated by eitker model at all. A few moments of vector algebra will allow one to verify that R is perpendicular to both and and is therefore perpendicular to the entire theory plane. The viewer is assumed to be sighting along the theory plane, so that it forms the dotted line at y = 0.The solid line embedded in that dotted line is the projection o f ? into the theory plane, while the vertical line and the vector .? itself are shown extending slightly into a perpendicular dimension. If one pretends, for a moment, that Figure 5a is a three-dimensional object rather than a flat plot on paper, the view in Figure 5b is what one would achieve by rotating the page slightly clockwise, so as to make the vector ? shown be level rather than inclined, and then tilting the page away from oneself so that one is sighting along the paper, rather than looking down on it. It must be noted, however, that this lower graph is labeled as showing the "normalized residuals," rather than simply the "residuals." The reason for this is to avoid a visual deception arising from our accustomed three-dimensional experience. In the three-dimensional space of our everyday experience, a given plane has exactly one perpendicular axis. This is not so in more dimensions. The eight-dimensional space that Figure 5 attempts to summarize permits six mutually perpendicular directions all of which are also perpendicular to the theory plane. Under the hypothesis that (?) lies in the theory plane, the actual observation Z is expected to depart by a Z-distributed amount in each of t_hos_esix orthogonal directions, so the squared length of the residual vector, R R, is clearly a with+a number of degrees of freedom equal to the number of dimensions in which R exists. Therefore, the visual presentation normalizes this length. Figure 5b reports the x2 value, and its degrees of freedom, explicitly: tc ts tc + zc ts x2 184 York H. Dobyns but, because the eye only expects one perpendicular to a plane, rather than six, the vertical distance shown is the Z score that has the same two-tailed p-value as the X 2 calculation. The vertical axis can be read off directly as a Z score; it represents the length that the residual vector would have if the entire vector space were confined to the three dimensions we are accustomed to visualizing. Turning now to the comparison of BIT with TIM, Table 5 presents the raw data when segregated according to the key variable of bits per second. These figures result in the model comparisons summarized in Table 6. Figures 6 and 7 illustrate the model comparison between BIT and TIM in the same way as Figures 4 and 5 do for BIT versus DAT. In Figure 6a, the x axis is, of course, 1 / a . As it happens, we have only three data points for this plot, and one of them has rather large error bars. This is the reason for the very wide error contours visible on the lower plot; despite the fact that both the BIT and TIM models are almost 6 0 away from the origin, the ability of the regression to discriminate is very much weaker, so that the 2 0 contour covers almost the whole plot. Nevertheless, the regression fit is visibly much closer to the BIT model than to the TIM model. Figure 7a illustrates the same point: The acute angle between the two theory vectors shows why ? has little power to distinguish between them, despite being much closer to one. Figure 7b once again shows no consequential residual vector. It is worth noting that because we actually have only three dimensions in this vector space, the normalization of Figure 7b is one-to-one and we see the residual vector at its true length. The TEL model refers to a historical evolution within a single experiment, rather than across experiments. The model parameter becomes the amount of TABLE 5 BIT vs. TIM Comparison Data 20 200 2,000 1.64 x lo6 6.0409 x lo8 3.246 x lo8 8.2 x lo4 3.02045 x lo6 1.623 x lo5 -0.9558 5.2477 2.7179 Note: BIT = bitwise effect; TIM = time normalization. TABLE 6 BIT vs. TIM Comparison Results Test BIT vs. null (Z) TIM vs. null (2) BIT vs. regression (Z) TIM vs. regression (Z) Score P Xivs. BIT 2.27 5.57 .32 .062 X; vs. TIM Note: BIT = bitwise effect; TIM = time normalization. Models on PEAR Data 185 Z vs Constant: 0.515 : 0.0015 0.0010 Z0.0005 0 Z vs Slope: 1.887 (a) ~0.0005 -0.0010 -0.0015 0.0 1 1 1 Q) 1) A 1 0.05 0.10 0.15 0.20 Inverse square root of (bits per second) - 4 - 2 0 2 4 6 Intercept Z score 8 1 0 Fig. 6 . Model comparison: BIT = constant and TIM = slope. data accumulated in the experiment to date. To test this model, we therefore must perform evaluations within single experiments. Furthermore, because the hypothesis predicts a progressive diminution in effect from series to series, a tabular presentation of the raw data would be excessively cumbersome, with as many lines as the total of series enumerated in Table 1. Therefore, Table 7 presents only analysis results: the Z score versus the null hypothesis, the Z score against the linear regression, and the x2 fit test, for both BIT and TEL models, for the six experiments summarized in Table 1. 186 York H. Dobyns (b): Normalized Residuals 0 2 4 6 Fig. 7. BIT constant and TIM slope as vectors. Here we have a rather disconcerting lack of unity. Some of the experimental datasets show a strong preference for TEL and some show a modest preference for BIT. The result cannot be attributed entirely to statistical power issues, as the two largest datasets show oppositely inclined preferences. The largest of these datasets, furthermore, contains a known confound. As was noted in section 2, an analysis of this sort proceeds on the implicit assumption that conditions affecting the experimental yield, which are not related to the hypothesis under consideration, are uniformly or at least randomly distributed across the subsets defined by the hypothesis parameter. This is Models on PEAR Data TABLE 7 BIT versus TEL Summaries - Experiment Znb znt zrb z, -0.626 1.534 0.072 0.628 -0.523 -1.762 X: (df 571.6 (521) 200.5 (21 1) 57.12 (45) 39.68 (44) 45.41 (43) 18.31 (19) X: (df ) 555.9 (521) 202.8 (21 1) 54.49 (45) 40.03 (44) 41.17 (43) 19.17 (19) Original 3.809 Remote 2.214 Alternative source 2.765 Co-operator 1.635 REG2000 2.7 18 REG20 -0.956 5.405 1.615 3.205 1.525 3.410 -0.229 3.886 -0.244 1.624 0.218 2.125 1.498 = Note: BIT = bitwise effect; TEL = teleological; Znb BIT Z score versus null; Zn,= TEL Z score versus null; Zrb= BIT Z score versus regression fit; Z, = TEL Z score versus regression fit; &df) = Goodness-of-fit x2 for BIT (degrees of freedom). &df) = Goodness-of-fit X 2 for TEL (degrees of freedom). known not to be the case for a specific subset of the "Original REG" data with respect to the TEL parameter. As has been discussed in fuller detail elsewhere (Dobyns and Nelson, 1998), the earliest REG experiments were run with a slightly different experimental protocol, which distinguishes itself by a much larger effect size than the subsequent continuation of the experiment. Because the change in protocols is a historical one and distinguishes part of the earliest data from the remainder of the dataset, it is directly confounded with the measure of historical progression that defines the TEL hypothesis. Merely identifying the presence of a confounding factor, however, cannot by itself determine the status of the two hypotheses. The fact that the difference between the early data (designated as the X protocol) and the subsequent data was well established long before the TEL hypothesis was considered prompts an inclination to consider the support for TEL a confounding result of the difference in protocols, but this reasoning is spurious. It is equally plausible that the unexplained difference in the protocols is driven by the hitherto-unrecognized effects of teleological decline. To resolve this ambiguity, we would like ideally to compare data generated under the two protocols at the same teleological status. This unfortunately is impossible, because "teleological status" is a historical parameter that is distinct for any two series. We can do the next best thing, however, by segregating the two conditions and examining each independently for teleological effects. There is a small subtlety involved in this segregation, however. When we remove the X protocol, we are taking away the earliest series in the Original REG dataset. Did the protocol change "restart" the teleological "clock," so that the new data represented a new beginning? Or should the subsequent data be evaluated under the assumption that they are continuing a teleological decline started in the early data? We do not know enough about the constraints of the TEL model to answer this question, so we have no alternative but to ana- York H. Dobyns TABLE 8 BIT vs. TEL Summaries, Reduced Datasets Experiment X only Continue Restart Znb znt Zrb zrt ~@f) 46.19 (15) 516.9 (505) 516.9 (505) X (df : 34.69 (15) 519.0 (505) 521.3 (505) 3.519 3.102 3.102 4.886 2.734 2.286 4.331 0.034 -0.124 -2.696 1.466 2.101 Note: BIT = bitwise effect; TEL = teleological. Z vs Constant: 4.332 ; Z vs Slope: -2.696 -10 -5 0 5 10 Intercept Z score 15 Fig. 8. BIT constant versus TEL slope for X protocol only. Models on PEAR Data 189 1 lyze the "remainder" data in both ways. The results of such analysis are presented in Table 8. It is plainly evident that while the X protocol shows a preference for TEL internally, neither model is a very good fit to the data in this unique dataset. Both x2 values are quite large, withp-values of 4.96 x and 2.72 x lo-', respectively. Neither Z score is within the range that would be considered plausibly consistent with model variation. This strongly suggests that this dataset contains some structure other than that explained by either model. Conversely, the data without the X protocol no longer show any support for TEL, regardless of the status of the "restart" assumption. Figures 8 and 9 display the X protocol data in the format used for Figures 4 and 5 and Figures 6 and 7. In Figure 8a, we note what appears to be almost a bifurcation in the experimental data, with the data points below x = 0.0004 seeming at least visually to cluster into two vertically separated populations. It is inevitable that even two-parameter linear regression fails to fit this dataset well, because no single-valued functional fit can possibly reproduce such a feature. In Figure 8b, the regression fit is conspicuous for its wide departure from both families of theoretical model. Figure 9, illustrating the vector space representation, drives home the same point. In the theory plane of Figure 9a, the observation vector is well outside the acute angle subtended by the two theory vectors, differing more from either theoretical prediction than the theories differ from each other. The residual graph in Figure 9b shows us a vector lying far outside the theory plane, with an equivalent Z score well over 2. Figures 10 and 11 give the same reports for the non-X protocol data, using the "restart" assumption. We can see that the results are now a very good fit to BIT, whether we look at the parameter plot in Figure 10 or the vectors of Figure 11. We also see a completely inconsequential residual in Figure 11b. The overall conclusion remains ambiguous. Computing a weighted Z across all of the datasets reported in Table 7, with the "Original REG" further broken down as summarized in Table 8, leads to the unenlightening results presented in Table 9. As in Table 6, two results are reported, depending on whether the transition from the X protocol to the standard protocol is considered a continuation of the Original REG experiment, or a transition to a new experiment. We are left with the conclusion that we have no good grounds for preferring either BIT or TEL on the basis of the current data. TABLE 9 Overall BIT vs. TEL X Treatment Continue Restart Note: BIT = bitwise effect; TEL = teleological. Z vs. BIT Z vs. TEL 190 York H. Dobyns (b): Normalized Residuals Chi-Squared: 27.424 (14 d.f.) 4- I z 2- 0 --.--t ---. I t 1- 0 2 4 6 Fig. 9. BIT constant and TEL slope, as vectors, X protocol data only. 7. Conclusions We noted above that the choice of model can affect even the the primary conclusion about the existence of an anomaly. One reason for this follows from the differences in the definition of the fundamental data unit. Let us suppose that we have several datasets, each of which contains Ni observations, and has a Z score Zi, where i = I,... ,m for m = the number of datasets. What is the overall Z score of the aggregate database? It is provably correct that this ~ifl)/dm, in the sense that this reproduces composite score is Z = (x Models on PEAR Data 191 0.0041 Z vs Constant: -0.1 ; 24 Z vs Slope: 2.101 1 0.002 0.0 .0.002 .0.004 0.0 0.0002 0.0004 0.0006 0.0008 0.0010 Inverse square root of accumulated data total -2 0 2 4 6 Intercept Z score Fig. 10. BIT versus TEL with X data removed, restart version. the Z score that would be calculated from pooling all of the raw data. We have noted, however, that different models of the phenomenon may disagree on the definition of the fundamental data unit and therefore may assign different values of Nito the various subsets. In general, this changes the value of the composite Z. This is the reason for the different Z scores against the null hypothesis displayed by the different models in section 6. Therefore, when competing models of the effect have been proposed, it seems advisable to select the model that is the best fit to the data before at- 192 York H. Dobyns (a): Projection of Z In Theory Plane (b): Normalized Residuals Fig. 1 1 . BIT constant and TEL slope as vectors, X protocol data removed. tempting an absolute test for the existence of the effect. We have done this, above, for the four models under consideration and find in each case that either the BIT model is preferred, or that no preference can be established. If the BIT model is thus taken as the best candidate model of a real effect, we note that its Z score against the null hypothesis is 5.8, with ap-value of 3.3 x We may be confident that the database examined here shows a real effect and that the hypothesis that this effect comprises a constant bitwise probability shift serves at least as well as does any alternative thus far considered. Models on PEAR Data 7.1 Bayesian Calculations The foregoing material has employed a "frequentist" approach to the interpretation of the statistical scores, by the presentation of p-values. It would be inappropriate to reprise the entire theoretical debate between frequentist and Bayesian statistics here, and rather than take sides in this dispute, PEAR has found it more productive to present results in both formalisms. The Z scores calculated above can be put to work immediately in a Bayesian hypothesistesting framework. Where possible, we prefer to present results as an odds ratio, or odds adjustment factor, which represents the relative support that a piece of empirical data provides to two well-defined hypotheses. Because this describes the proportional change that must be made in going from prior to posterior probabilities on the hypotheses concerned, it can be computed without reference to the prior probabilities. Table 10 summarizes the odds ratios for the pairwise comparisons made above. We see that we have strong grounds for supporting BIT against DAT, and much weaker but still positive support for preferring it to TIM. The two conflicting Bayes factors, both close to 1, for the "continuation" versus "restart" interpretation of the X protocol transition, reflect the lack of conclusive information discussed in the previous section. The Bayesian evaluation against the null hypothesis becomes slightly more complicated. In the pairwise comparisons above, the two-parameter linear regression is simply compared with two families of one-parameter models; no prior information about the model parameters is required, since they are handled in a perfectly symmetric fashion. However, once one of these models (in this case, BIT) is chosen as the favored alternative for comparison with the null, the actual odds adjustment one computes will depend on the prior assumptions one makes for the actual parameter values, rather than merely their constraint equations. Because we are using a BIT model for our comparison, the variable under consideration is the probability that a binary decision chooses the option targeted by human intention. The null hypothesis predicts that this probability TABLE 10 Pairwise Odds Adjustments Comparison BIT vs. DAT BIT vs. TIM BIT vs. TEL (continue) BIT vs. TEL (restart) Note: BIT = bitwise effect; DAT = Decision Augumentation Theory; TIM = time normalization; TEL = teleological. Odds Adjustment Favoring BIT 194 York H. Dobyns physical processes and that the probability might take on any value between 0 and 1. In other words, the effect of intention is a probability shift Ap in the range -?h < Ap < %. This is in many ways a maximally conservative evaluation. Nonetheless, when this is used as the alternative hypothesis, one finds an odds adjustment factor of 1658 in favor of BIT. More informed priors produce better odds in favor of the effect. For example, one can almost casually constrain the effect of human intention on random events by making the observation that casinos make steady profits despite the presumptive (and often quite obvious) intention and desire of gamblers to win. We may therefore presume that whatever the typical scale of intentional effects on binary probabilities, it is smaller than the half-percent house advantage, which seems to be the smallest value for a commonly played casino game. (Blackjack, played by optimal strategy but without card counting, has this house advantage level.) Taken as the bitwise alternative, this means a uniform prior probability density -0.005 < Ap < 0.005, which produces an odds adjustment of 1.66 x lo5 in favor of BIT. Finally, and most apropos, a previous meta-analysis of microelectronic PK experiments (Radin and Nelson, 1989) surveyed a substantial body of prior work in the field. Unfortunately this survey includes some of the data in the current analysis; however, when the overlapping data are removed from the previous observation, the remainder provides a prior probability for the bitwise alternate that leads to an odds adjustment of 5.5 x lo6 in favor of the effect. Acknowledgments The author gratefully acknowledges the advice and assistance of Werner Ehm of the Institut fiir Grenzgebiete fiir Psychologie und Psychohygiene (IGPP) in developing the analytical formalisms of section 5. Please direct reprints requests to York H. Dobyns, C-131 E-Quad, Princeton University, Princeton, NJ 08544-5263; e-mail [email protected]. The PEAR laboratory is funded by donations from the IGPP, the Lifebridge Foundation, the Ohrstrom Foundation, and Laurance Rockefeller. References Dobyns, Y. H., & Nelson, R. D. (1998). Evidence against decision augmentation theory. Journal of Scientific Exploration, 12, 23 1-258. Jahn, R. G., Dunne, B. J., Nelson, R. D., Dobyns, Y. H., & Bradish, G. J. (1997). Correlations of random binary sequences with pre-stated operator intention: A review of a 12-year program. Journal of Scientific Exploration, 11, 345-367. May, E. C., Utts, J. M., & Spottiswoode, S. J. P. (1995). Decision augmentation theory: Applications to the random number generator database. Journal of Scientific Exploration, 9,453-488. Nelson, R. D. (1 994). Effect size per hour: A natural unit for interpreting anomalies experiments. Technical Note PEAR 94003. Radin, D. I., & Nelson, R. D. (1989). Evidence for consciousness-related anomalies in random physical systems. Foundations of Physics, 19, 1499-1 5 14. Journal of Scientific Exploration, Vol. 14, No. 2, pp. 195-216,2000 0892-33 10100 O 2000 Society for Scientific Exploration The Ordering of Random Events by Emotional Expression 1 Center for Functional Research 2392 Mar East Street Tiburon, CA 94920 e-mail: [email protected] Abstract- The purpose of this experiment was to see whether any correlation could be found between the expression of emotions of patients in Reichian biopsychiatric therapy and the output of an electronic random event generator (REG). Videotaping and coding of therapy sessions were conducted in synchronization with the operation of an REG. Comparisons were made of REG output while patients spoke with neutral emotion and during the spontaneous expression of crying with anxiety, frustration or sadness, and anger. Statistical analysis revealed anomalous REG outputs during periods of emotional expression compared to periods of neutral talking. Periods of the expression of anger compared to those of crying with sadness andlor anxiety were significantly correlated with marked elevations and depression of REG output, respectively. Based on the concepts developed by Jahn and Dunne, as well as Reich, it is hypothesized that the observed effects are due to the establishment of resonance between the therapistlinvestigator, the emotionally expanding andlor contracting patient, and the REG. Keywords: REG-emotion-resonance Introduction In their experiments on human-machine interactions, Robert Jahn and Brenda Dunne (Jahn and Dunne, 1987; Jahn et al., 1997) demonstrate that the distribution of impulses generated by a random event generator (REG) can be anomalously, marginally shifted from normal either locally or at a distance by active mental intention. They also found that many operators appeared to impose a signature on the REG output, i.e., the pattern of output when a particular operator was operating displayed similar directional trends on most runs on the electronic REG and other experimental REG devices. The effect appeared to be equally strong even if the operator was thousands of miles away from the device. The experimenters also found that the REG output during baseline (null-intention) runs was anomalous in that the obtained variance of the REG output was significantly narrower than the theoretical expectation (Jahn and Dunne, 1987). Later studies indicated that the baseline anomalies were strongly correlated with operator gender (Dunne, 1998). These findings indicate that something other than conscious intention can apparently affect the REG and do so at a distance. 196 R. Blasband In their model, Jahn and Dunne define consciousness as all that one identifies as oneself: thought, emotions, physical substance, etc. They account for their anomalous finding by hypothesizing a nonelectromagnetic field, through which a state of resonance is established between the operator and the machine. Given the above findings, and assuming that Jahn and Dunne were correct in including emotions in their definition of consciousness, we reasoned that if other-than-conscious intentions can order the output of an REG, then it was possible that spontaneous, undirected, emotional expression might, in some way, do the same. In the form of psychiatric therapy employed by the author (Reichian biopsychiatry), intense emotional expression by the patient occurs fairly regularly. It was therefore hypothesized that an REG set up in the therapy office 10 feet distant from a patient would be anomalously affected during those moments when the patient would express emotions compared to those times when patients would be emotionally neutral. The present study is an initial survey designed to see whether correlations can, indeed, be found between REG output and emotional expression. I Literature Review Other than the quantum mechanical theoretical formulations of Jahn and Dunne (1987), there was no reason based upon the findings and theories of classical biophysics to expect any influence on the output of an REG by a spontaneously emoting individual many feet distant from the device. Although it was true, as stated above, that Jahn and Dunne had amply demonstrated that the output of the REG could be correlated with the conscious intention of operators, there was nothing in the literature at the time of beginning the experiments described in this paper that suggested effects of either local or nonlocal emotional output on such a device. Since then, the only relevant findings involve the effects of group expression on the REG (Nelson et al., 1996, 1998; Radin, Rebman, and Cross, 1996; Rowe, 1998; Schwartz et al., 1997). Further, a review of the mainstream literature on the biophysical basis of emotions shows that although one can demonstrate myriad physiological parameters that correlate with feelings and expressed emotions, all involved measurements are made either within or on the surface of the body (Panksepp, 1998). Indeed, except for experiments performed by Wilhelm Reich and Harold Burr, I could find no research involving biophysical and physiological parameters of biological or psychobiological functions where measurements were made nonlocally. Burr observed that an electrical field measured a small distance away from the surface of an unfertilized, biologically undifferentiated salamander egg appeared to have a determining effect on the establishment of the pattern of the future axis of the central nervous system (Burr, 1972). On the basis of his experimental observations, Burr, with F. S. C. Northrop, formulated an electrodynamic theory of life, maintaining that an electrical field was a primary prop- . Ordering Random Events by Emotional Expression 197 erty of protoplasm, sustaining pattern in the organism in the midst of physiochemical flux (Burr and Northrop, 1935). The work of the contemporary biophysicist Mae-Wan Ho at the Open University, London, complements many of Burr's findings. Her analysis of her own and other's experimental work supports the hypothesis that the "organizing embryonic field is global in character, right down to individual macromolecules, and that its major axis is electrodynamic in character" (Ho and Saunders, 1994). The radiation effects of electromagnetic (EM) fields generated by the acceleration of electrically charged inorganic, organic, and macromolecular ions within the body have been investigated experimentally and theoretically. The waves resulting from the movement of these ions through the body are variously emitted, decay, and are refracted and reflected at organ interfaces, producing interference patterns, which combine according to quantum theory's superposition principle. According to C. Zhang and F. A. Popp, these interference patterns create stable standing waves, which may have a great deal to do with holographic effects seen in the global functioning of the organism and its nonlocal treatment by such modalities as acupuncture (Rubik, 1995). But even in these studies involving EM fields, all measurements were made on or beneath the surface of the body. Thus, we see that there is at least some basis for the existence of a biofield effect based upon electromagnetism. However, although it is possible that emotional expression could affect an REG through perturbing an EM field emanating from the body or be some component of the elements that generate such a field, the findings and theory of electromagnetism provide no basis for expecting that such a field would extend much further than a few inches from the body. Further, and most important, as Jahn and Dunne note, the experimental phenomena documented in their work with the REG and remote viewing, especially those involving anomalies of time, cannot be explained by electromagnetic theory (Jahn and Dunne, 1987). A non-EM biophysical basis for the emotions was elucidated by Wilhelm Reich, the Austrian psychoanalyst, who devoted the latter part of his life to the investigation of the nature of life and life energy. In his clinical psychoanalytic and later vegetotherapeutic and orgone-therapeutic work, Reich proposed that emotions were a function of the patient either bioenergetically expanding toward or contracting away from the outer world (Reich, 1949). This amoeboidlike behavior confirmed an earlier postulate of Freud that never took serious root in later psychoanalytic theory. When Reich moved from the classical psychoanalytic technique of free association toward the more confrontive kinds of intervention of the method he called character analysis, he noted that patients more readily gave in to the expression of their previously blocked emotions. When this happened he observed spontaneous pulsatile (clonisms) movements of the patient's body, which at times included the entire torso. These movements were greatly amplified when Reich added massage directed at physical release of the patient's chronic muscular tension (muscular armoring) to his 198 R. Blasband therapeutic armamentarium. Reich found that as the characterological and muscular armoring softened in the course of therapy, patients reported feeling electrical currents and sensations of something streaming through their bodies. This was usually associated with an increase in general vagotonic tone, flushing of the skin, brightening of the eyes, contraction of the pupils, slowing of the heart, and an increase in pleasurable sensations at the skin. The opposite of this state of bioenergetic expansion was one of bioenergetic contraction, usually brought on by fear or anxiety, and characterized by a general autonomic sympathetic tone with pallor of the skin, narrowing of the eyes, dilatation of the pupils, acceleration of heart rate, and sensations of inner tension (Reich, 1942). In order to objectify these observations, Reich measured bioelectric charge on the skin surface of subjects in a variety of emotional states. He found that the subjective perception of anxiety or sadness was directly correlated ("functionally identical," to use Reich's term) with a contractive movement of bioelectricity away from the skin surface toward the bioenergetic core of the organism-autonomic neural plexes deep in the abdomen and pelvis. Anger, pleasure, and longing were correlated with an expansive movement of bioelectricity from the core out to the skin surface (Reich, 1937). These directions of movement were understood by Reich to be mediated by opposing domination of the two different branches of the autonomic nervous system, the parasympathetic in bioenergetic expansion and the sympathetic in bioenergetic contraction. A more recent attempt to replicate Reich's study using modem equipment confirmed in many respects Reich's findings (Braid and Dew, 1988). Reich found that a bioelectrical interpretation was not, however, sufficient to explain adequately all the phenomena observed in his bioelectrical studies. He then undertook a series of experiments on the sources of energy sustaining life, which ultimately suggested a nonelectromagnetic basis for living processes. In the course of his research, Reich reported experiments in which he postulated a field of bioenergy, orgone energy, surrounding and interpenetrating all living things. Reich's principle device for detecting this field, the orgone energy field meter (Reich, 1948, p. 125) could apparently detect the energy field of a lively human at distances up to 6 feet.' The effects of the spontaneous expression of emotions on the meter were not undertaken, as far as I know, although Reich did report that subjects who were more vegetatively alive (capable of the expression of intense emotions) could more readily affect the meter, compared to those who were vegetatively dead, a catatonic schizophrenic or a heavily armored obsessive compulsive neurotic. 'This device consisted of moveable facing metal plates, one of which was connected to the different pole of the secondary coil of an induction apparatus. A 40-watt bulb connected between the plates glows when the primary current is at a certain intensity. The proximity of something living to the upper plate affects the intensity of glow of the bulb. The more alive the object the more intense the glow. Ordering Random Events by Emotional Expression 199 Confirmation of Reich's psychiatric, biophysical, and physical findings have been reported in those few contemporary journals devoted to his work (The Journal of Orgonomy, Annals of The Institute for Orgonornic Science) but never seriously challenged by reports in the mainstream literature. Methodology Reichian biopsychiatry is a so-called depth therapy, whose aim is to free the patient from his characterological and muscular armoring, or blocks, thus permitting the free flow of life energy through the organism (Reich, 1949). In the process, emotions are spontaneously released. The depth of the emotional release is a function of many things, including the layer of the personality or character being addressed at that moment in the therapy, the rigidity of the character structure of the patient, whether hysterical or compulsive, etc., and the energetic charge of the patient in general and on that particular day. Patients usually engage in therapy for several to many years, motivated by the continuing improvement they experience in their sense of internal freedom and well-being, their increased capacity to feel pleasure, and a growing sense of inner strength, personal independence, and the capacity to accept greater responsibility for their lives. The technique of therapy involves attention to and interventions in the process of verbal and nonverbal interchange between the therapist and the patient plus detailed and consistent attention to the patient's characterological and muscular armoring. Characterological armoring is treated by the therapist's informing the patient through either verbal interventions or mimicry the artificial ways in which the patient appears and behaves. Muscular armoring is the functional somatic counterpart of characterological armoring. Chronic spastic tensions in the striated and smooth musculature are released by deep massage combined with encouraging the patient to express any emotion bound by the arrnoring. The first part of each session usually involves talking by the patient as helshe describes whatever is on hislher mind and verbal responses and interventions by the therapist. This is often, but not always, followed by having the patient, prone on the therapy couch, deeply sigh in order to build up a bioenergetic charge. This may, by itself, without any further interventions, be enough to trigger the overt expression of blocked feelings. When the therapist sees that there is no or little energetic movement, he may intervene by describing to the patient a characterologic attitude or state of bodily constraint or by direct systematic work on the musculature to release armoring. In therapy sessions during the experimental REG periods, the therapist intentionally took a more passive role than usual in order, as much as possible, to avoid adding an unnecessary variable to the experiment. This meant fewer verbal interventions relating to the patients' character and much less work than usual on the musculature. Patients participating as subjects in the experiment were rarely physically touched, and, in the few situations where it was 200 R. Blasband deemed necessary in order to advance therapy, not more than once during the session. Study 1 The experiment was conducted in the therapist's (R.A.B .'s) office, a renovated trailer located 30 feet away from his house, in a semirural setting in Northern California, at least 118 mile away from the nearest neighbor. Patients were videotaped during each session, while a computer time-synched to the camcorder collected REG data. Patients were informed only that the therapist wished to conduct an experiment with a random event generator, and that part of the experiment involved videotaping of their sessions. All patients gave their permission to proceed and accepted the experimental conditions with no observable inhibition throughout the course of the experiment. Twelve patients, ranging in age from 25 to 60 years, were initially selected from the therapist's full caseload because they had been coming to treatment regularly, most of them weekly, for at least 1 year prior to the study, and their superficial resistances, including distrust of the therapist and the therapeutic process, had been well resolved. Of the 12 patients three were men. Since it had been the author's clinical experience that emotional expression by men is much more difficult than for women, and in view of the fact that only three men were available for the study, it seemed best to limit the current study to women to optimize the possibilities of seeing some kind of correlation of emotional expression with the REG. It was anticipated that when the therapist's caseload included more men who could qualify for the experiment at a later time, that the current experiment would be repeated using only men.2Owing to a technical problem, the data from one of the two therapy sessions of a female patient was not recorded, so she was dropped from the study. Of the remaining eight patientlsubjects, characterological diagnoses included hysterical, phallic, and nonpsychotic catatonic schizophrenic character types, using Reich's (1949) and Baker's (1967) character typology. Videotaping was done with a Sony 8-mm camcorder unobtrusively placed in the office. The camcorder was set to record the time and date of the beginning of the session and to continuously record elapsed time. A portable random event generator, similar to those used in their field REG experiments (Nelson et al., 1996, 1998), was provided by the Princeton Engineering Anomalies Laboratory (PEAR) along with software to provide continuous REG recording to hard disk with built-in statistical and graphing capabilities. According to PEAR, in this device "the random event sequence is based on a low-level microelectronic white noise source which is amplified, limited, and ultimately compared with a precisely adjusted DC reference level. At any instant of time the probability of the analog signal equaling or exceeding the ref- 2Dunne(1 998) later reported gender differences in studies involving conscious intention on the REG. Ordering Random Events by Emotional Expression 201 erence threshold is precisely 0.5. This white noise signal is sampled 1,000 times per second, and the output of a comparator stage is clocked into a digital flip-flop, yielding a stream of binary events, 1 or 0, each with probability 0.5. This unpredictable, continuous sequence of bits is then compared with an alternating template using a logical XOR in hardware, and the matches are counted, thus precluding first-order bias of the mean due to short or long-term drift in any analog component values by inverting every second bit. The resulting sequence is then accumulated as bytes that are transmitted to a serial port of the computer, where they are read and converted to REG data by dedicated software" (Nelson, 1996). Built-in, fail-safe, and calibration components guarantee the device's integrity against technical malfunctions and environmental disturbances. Whereas the original PEAR experiments with conscious intention used a tripolar protocol, the field REG version used in this experiment had a single null-intention protocol. Data were fed into the computer by the REG in continuous 13-minute, 1,000-trial segments. Except for an indication that the computer was recording data, the screen was blank. Thus, investigator and patient were blind to any results emerging during the session. The device plus attached Zeos laptop computer was located 10 feet away from the patients, out of their line of sight. The camcorder and REG were started within 5 seconds of each other and within 15 seconds of the beginning of the therapeutic sessions. This permitted two or three REG segments to be recorded over the course of a 30- or 45-minute session. Calibration runs were made at intervals during the several months of the experiment when the office was unoccupied during the day and, at times, through the night. The task of the data analysis was to determine whether any correlation existed between the patient's overt emotional expression and the REG output. For these purposes we considered an emotionally neutral period of talking as one where there was no obvious elation, anger, anxiety, or depression being expressed by the patient as she spoke with the therapist at the beginning of each session. Segments registered as containing emotional expression were those where emotion was actually expressed by either spontaneous overt crying, screaming with fear, reaching out and/or sobbing with longing, or yelling and/or hitting in anger, all of which were seen during the course of the experiment. That is, we distinguished patients' subjective perceptions of emotional states from their overt expression of the state. Patients might feel like they would like to or were going to cry, for example, but this was distinguished from the actual expression of crying with overt sobbing. The former was not considered overt expression, but the latter was. Where the patient was not overtly expressing an emotion such as crying or anger while talking, but the therapist could clearly sense a strong undercurrent of sadness or anger, the segment was excluded from the data analysis. The therapist was highly experienced at making such estimations, having done so for over 25 years of clinical work. Periods of remarkable pleasure or joy were not seen in any patientlsubject during the experimental periods, although most patients reported considerable 202 R. Blasband relief following the full expression of a blocked emotion. It is usual in this form of therapy for a single emotion to dominate long periods of the session, and this held true for our experimental sessions: In nearly all therapy sessions the time period corresponding to a given 13-minute REG segment was clearly dominated by a single form of expression. This made it fairly easy to label the emotional qualities of most segments. For example, during an initial 5 to 10 minutes of discussion, a patient might evidence neutral affect, then, when lying down and gently sighing, might spontaneously begin to cry. This is not unusual in women in this kind of therapy. Usually sobbing (indeed, any emotion expressed in therapy) is expressed in pulses, that is, several periods of two to three minutes of sobbing separated by one or two minutes of simply sighing or verbalization without overt sobbing. Being constrained by the software to compute in 13-minute segments, we labeled the segment as one of crying despite the fact that the patient did not cry during every minute of the segment. The available software at the time of Part 1 of this study provided only cumulative results of output for each segment; therefore, we could not precisely extract that REG output that correlated with each emotional period during segments where mixed emotions were expressed. In those segments where mixed emotions, such as anger and sorrow, were sequentially expressed we counted the minutes associated with each emotion and labeled the segment according to the dominant emotion. Such segments were rare during the experimental periods. To minimize the possibility of feedback-driven effects on the REG by the investigator's intentionality, the data from all subjects were analyzed only at the completion of the experiment. The evaluation and labeling of the kinds of emotions expressed during each segment were made by the author. The procedure for matching independent and dependent variables was as follows: The times of the beginning and end of each REG segment were noted, then the dominant emotional expression corresponding to this interval was paired with the REG output during that segment. By observation, we established in Study 1 that we were dealing with five main, easily differentiated categories of behavioral expression-emotionally neutral talking; talking with emotion; sighing without emotional expression; overt crying with fear, anxiety, frustration, or sadness characterized by sobbing with tears; and anger characterized by yelling, and (often) hitting the couch, andlor kicking. One patient also expressed longing through part of a segment of a session: This was characterized by her crying with reaching out with the arms and verbalizations of "wanting mother." We limited our quantitative analysis to a comparison of REG outputs for segments of emotional expression that were most prevalent and clear cutcrying with sadness, fear, or anxiety (all labeled below as anxcry), and anger. REG outputs during segments of these emotional expressions were compared to outputs during neutral talking at the beginning of the sessions, and to each other, using one-way analysis of variance (ANOVA). Ordering Random Events by Emotional Expression Results 203 A. Calibration. The calibration sample totaled 979 13-minute segments over the test period. The mean number of counts per 200-sample trial over all the runs was 99.99 with a standard deviation (SD) of 7.06 indicating that the REG-computer setup was operating within the range expected theoretically (mean = 100, SD = 7.07) at a probability against chance of .40, consistent with previous calibrations at PEAR. A graph of 10,000 calibration trials (10 segments) is shown in Figure 1. The parabolic line above and below the baseline indicates displacement from the mean at the one-tailed 5% level of probability. One can see here that the REG takes a "random walk" about the baseline. Note that the graph crosses the parabolic boundary early on but quickly returns to and remains within the expected boundaries as the trials proceed. B. Experimental. The eight patientlsubjects selected for the study had a total of 39 therapy sessions during the experimental period in the spring of 1993.The number of recorded therapy sessions per subject varied considerably, ranging from one to eight. The 39 sessions yielded 76 13-minute REG segments. Of these, 33 were during periods of neutral talking, that is, talking without obvious emotion or emotional expression (neutalk on charts), 13 were during periods of anger, and 30 were during periods of crying. The following details from a few therapy sessions will give some idea of the -1414l I 1 I Number of Trials 0 3333 6666 10000 Fig. 1. Calibration. 204 R. Blasband kinds of behavior observed in therapy, the REG results, and patterns of movement in the REG output. We will begin with data from two subjects who expressed a single emotion throughout most of their sessions. Subject 6, a woman in her thirties, had a relatively unarmored character structure and was fluid and open in her expression of emotion. During her six experimental sessions she expressed essentially two kinds of behavior, unemotional (neutral) talking, and, after a few minutes of sighing, fear with screaming and deep sobbing. Figure 2, a cumulative graph of all of her sessions, shows an accumulating downward shift in the trial counts emitted by the REG, breaking through the border of the 5% probability parabola in several places, and terminating outside the envelope. In contrast, Subject 8, who was also relatively unarmored, spent most of her experimental sessions spontaneously raging, with loud yelling and hitting the couch. Figure 3 shows the REG output for her six test sessions. The raging of sessions 1 to 5 is associated with a cumulative rise in REG output until the sixth session when her anger gave way to crying. At this point the REG took a sudden drop. The data in Figure 4 are from a subject (11) who expressed mixed emotions during the sessions. The cumulative graph is for a single segment of 1,000 trials. It is important to note that in the graphical representation, we are most -1723 1 0 I I Subject 6 I \ I Number of Trials 4423 8666 13000 Fig. 2. REG output during anxious crying. Ordering Random Events by Emotional Expression 1095 205 r I Subject 8 - -- Number of Trials 2000 4000 6000 Fig. 3. REG output during anger. concerned with sharp shifts in direction of movement, either up or down, rather than where the shift is taking place in relation to baseline. Thus a sudden shift upward, even though it begins well below the baseline, represents a greater than average number of correlation counts generated by the REG. This subject showed great lability in emotional expression throughout all her sessions. Talking or the slightest amount of sighing would often spontaneously lead to the expression of intense crying alternating with anger. The REG output correlated well with her emotional lability, with strong trends both above and below baseline at different times, although the final cumulative distribution of counts is well within chance. Figure 4 shows the first of three 13-minute segments from her session of April 25, 1993. She spent the first 8 minutes of the session talking about recent events without much emotion, then when instructed to sigh, spontaneously trembled, cried, and felt fear. When talking, the REG output was nominal, wandering above and below baseline in a typical random walk. When she began to cry, however, (at 666 trials) the REG output rapidly dropped. All Subjects The basic unit of analysis for quantitative assessment in this study is the 13minute segment, which consists of 1,000 trials where each trial is scored by the R. Blasband -447 0 1 I Number of Trials 1 1 333 666 1000 Fig. 4. REG output during anxious crying. number of counts observed in 200 samples of the 50-50 random process. The trial counts are averaged for the 1,000 trials in the segment. For convenience, these segment means are then converted to z scores using the formula, z= (MN - lOO)lSE, where MN is the count mean for the 1,000 trials in the segment, and SE is the standard error of the mean, computed as the SD (trial standard deviation), averaged over 1,000 trials, divided by the square root of 1,000. These z scores were used for all the analyses. Table 1 shows the average REG scores ( z scores) by subject and emotional state. The compounded data revealed in the totals show that the REG output during the expression of emotion is significantly different from that when patients are talking with neutral affect. Also, there is a significant upward shift in REG output when anger is expressed and a significant shift downward when the patient is crying. As can be seen in the table, this pattern is consistent across all eight subjects. Study 2 The experimental setup was similar to that used in Study 1 except for a change in locale and the use of upgraded software from PEAR. The change in locale consisted of a move of the same trailer-office to a new location, a few miles away. The office was located approximately 50 feet from my home and the closest other dwelling was at least one-quarter mile away. The upgraded Ordering Random Events by Emotional Expression TABLE 1 ANOVA of REG Output vs. Emotion-Individual Subjects and Totals Subject Neutalk Anxcrying Anger F-Test 207 1 4 5 6 7 8 10 11 13 Total M n M n M n M n M n M n M n M n M n Note: The n refers to the number of segments. M refers to the mean z score for the various conditions for each subject. Each epoch is 1,000 trials, and each trial score is the number of hits in 200 attempts of a 50-50 random process. In each row, means that share a superscript do not differ significantly (by Fisher's PLSD multiple comparison test, 0.05 level). software permitted continuous generation of REG output, unconstrained by the earlier 1,000 trial limit, the use of computer function keys to mark events in real time, and much more precise statistical analysis of data segments by trials, timed to seconds, or as marked by function keys. Videotaping, therefore, was replaced by taking timed notes of events and pressing appropriate function keys to mark times of the beginning and end of periods of talking, sighing, emoting, etc. REG data were obtained during the treatment of nine female patients, two of whom had been subjects in Study 1. The number of therapy sessions per patient ranged from one to 12. The same criteria for the selection of subjects used in Study 1 were used in Study 2. In the new office, the REG was within arm's reach of the therapist, on the side away from the patient. The therapist could press the computer function keys and read computer time, but the REG readout was not displayed. Procedure II At the moment the patient entered the office, the REG was started. As the session progressed, appropriate function keys were pressed to mark the beginning and end of events of overt emotional expression. These events and the time as displayed on the computer were simultaneously noted by the therapist on a pad and by pressing a function key. The end of the session was marked by turning off the REG. I 208 R. Blasband The REG data were examined only at the end of the 8-month test period, which began in January 1995. To analyze the data, notable events were first listed for each patient, then the REG record was examined for that event time period. Calibration runs were conducted during periods throughout the experiment when the office was unoccupied. A one-way ANOVA was used to analyze the data. Results The mean of 299,514 calibration trials was 100.012 with a SD of 7.072, well within the parameters for calibration determined by the PEAR laboratory. The 86,789 trials produced during all patient sessions yielded 70 notable events fitting the description given in Study 1, above. Thirty events were of talking with neutral affect; 30 were of the expression of anxious, frustrated, or depressed crying; eight were of the expression of anger; and two were of the expression of longing. As in Study 1, the basic unit of analysis is the segment of REG output associated in time with each notable event. In Study 2, however, the segments consist of a varying number of trials. For each segment, a z score is computed in the same way as in Study 1, and these z scores are used for the analysis. The z scores for the segments associated with the 70 notable events are listed in Table 2. An analysis of the data in Table 2 is seen in Table 3. It shows a summary of an overall comparison of REG output during conditions of the expression of neutral talking, anxious crying, anger, and longing using ANOVA. Significant differences are seen between all conditions [ F (3,66) = 3.956, P = .012]. This was due to the anxious crying condition being significantly lower than all of the other conditions in pairwise multiple comparison analysis [by Fisher's protected least significant difference (PLSD) statistic]. We can see from Table 3 that in Study 2 there was no significant difference in REG output between periods of neutral talking and the expression of anger, but there was a significant difference between periods of neutral talking and crying and between anger and anxious crying. As in Study 1, the direction of REG output is upward with the expression of anger and downward with the expression of anxious crying. The compounded data reveal highly significant differences between all variables. Data associated with longing was not included in this analysis since it was not recorded in Study 1. Longing will, however, be discussed below. Table 4 shows the results of a 2 x 2 ANOVA for the combined results of Studies 1 and 2, using study as the first factor (1 vs. 2) and condition as the second factor (neutral talking, anxious crying, anger). This analysis shows that the two studies did not differ significantly [F(1, 138) = 0.032, p = .8575]. Nor is there a significant study by condition interaction [F(2, 138) = 0.814, p = .4453]. It is therefore valid to pool the two studies for a combined analysis. Table 5 shows a comparison of the REG output for the different states of emotional expression using the pooled data from Studies 1 and 2. Ordering Random Events by Emotional Expression TABLE 2 Data for 70 Events, Study 2 N 209 Mean Neutral talking 1061 263 1399 261 929 1430 844 1127 157 362 686 416 210 255 158 155 548 1428 391 625 62 25 2 795 619 853 185 661 507 843 300 Anxious crying 2234 2263 422 1658 1288 894 1643 529 1619 1681 929 1352 1715 1089 1732 1482 1007 3154 328 1258 2118 1632 R. Blasband TABLE 2 (continued) Data for 70 Events, Study 2 N Mean SD Z -1.660 -0.121 -1.698 0.180 0.865 -1.197 -0.132 -0.935 1962 111 513 222 865 1595 1184 2283 Anger 99.735 99.919 99.470 100.086 99.869 99.788 99.973 99.862 100.188 100.136 99.848 100.476 100.233 100.163 100.488 100.381 100.620 100.1 10 6.988 7.675 7.064 7.492 7.186 7.290 7.082 7.046 7.418 6.979 6.761 6.687 6.536 7.738 7.144 6.832 7.007 7.214 622 863 683 607 45 1 338 162 452 Longing 0.663 0.563 -0.563 1.659 0.699 0.423 0.878 1.144 1.886 0.357 463 527 This analysis shows that when all of the data are considered there are significant differences between all the studied states of emotional expression and that anxious crying is associated with a downward shift of REG output and anger is associated with an upward shift in REG output. This may be illustrated by the graph of cumulative REG output shown in Figure 5. Discussion Although this study was designed simply to explore the possible relationship between emotional expression and REG output and to develop working hypotheses for future experimentation, the results from both studies appear remarkably robust. The data indicate that overt emotional expression on the part TABLE 3 Study 2-Comparison of Four Conditions, All Data Group Neutral Anxcry Anger Longing Count Mean SD SE 0.148 0.233 0.237 0.777 30 30 8 2 0.137' -0.437" 0.696' 1.127' 0.810 1.276 0.671 1.099 Note: The count is the number of separate segments for each condition of emotional expression. Means with different superscripts differ significantly at the 5% level. Means that share a superscript do not differ significantly (by Fischer's PLSD multiple comparison test, 0.05 level). Ordering Random Events by Emotional Expression TABLE 4 ANOVA, Combined Results of Studies 1 and 2 - 21 1 Source Study (A) Condition (B) AB Error DF 1 2 2 138 Sum Sq. 0.024 32.626 1.221 103.506 Mean Sq. 0.024 16.313 0.610 0.750 F test 0.032 21.750 0.814 P 0.8575 0.0001 0.4453 of female patients correlates significantly with shifts in the distribution of trial scores generated by an electronic random event generator. Furthermore, we find a highly significant correlation between the direction of the shift and the kind of emotion expressed: Periods where the patient was crying with fear, anxiety, frustration, or sadness were associated with a marked downward shift in REG output, whereas periods where the patient was expressing anger were associated with a marked upward shift. In attempting to understand the basis for this phenomenon, we are faced with several serious problems, the first of which is that there is no satisfactory understanding in traditional mechanistic biophysical terms of what an emotion I 0 10 Anxious Crying: Z-score 4.47 20 30 40 50 60 Thousands of Trials Fig. 5. Combined cumulative deviation of REG output. 212 R. Blasband TABLE 5 Comparison of Conditions, Studies 1 and 2 Pooled Group Neutral talking Anxious crying Anger Count Mean SD 0.722 1.039 0.658 SE 0.091 0.134 0.144 63 60 21 0.218" -0.547b 0.793' Note: The count is the number of separate segments included. Means with different superscripts differ significantly at the 5% level. Means that share a superscript do not differ significantly (by Fischer's PLSD multiple comparison test, 0.05 level) is, any more than there is an understanding of what the mind is. Second, even if we understood the biophysical basis of emotions in traditionally accepted terms, i.e., as originating and functioning strictly within the physical body, there is no known basis for understanding how emotions could influence a random event generator either locally or at a distance. A tentative, but incomplete, functional explanation is suggested, however, by combining Jahn and Dunne's concept of resonance in man-machine interactions and Reich's concept of bioenergetic pulsation in emotional expression. Jahn and Dunne (1987) hypothesize that consciousness operates primarily on a wave-mechanical basis and that the degree of resonance between the operator and the device is a function of the degree of superposition of the wavelike properties of the two. When resonance is established, they state, "molecular experiential patterns can arise whose observable characteristics differ significantly from the simple sum of their individual behaviors." At this point we must assume that human-machine resonance as Jahn and Dunne have defined it can be established whether or not there is conscious intention of the subject (operator, patient) toward the machine. Their field REG studies, cited above, indicate that this is possible. As noted earlier, Reich proposed that anxiety and pleasure are antithetical functions with respect to the phenomenon of total organismic pulsation. For several reasons, Reich found that bioelectricity per se could not satisfactorily account for all the phenomena seen in his experiments (Reich, 1948). His proposition of a vital, nonelectromagnetic, bioenergetic force called orgone energy did, however, satisfy the necessary requisites for understanding the phenomena, and he recast his formulations in its terms. According to Reich, states of anxiety or depression were functionally identical to a contractive movement of orgone energy toward the core, pleasure as an expansion of energetic excitation toward the skin surface, and anger as expansion toward the skin surface, but stopping at the musculature. The results of the present studies indicate that, with respect to direction of REG output, anger and anxietylsadness are antithetical emotional expressions. We have found that the emotions described by Reich as expansive significantly correlate with an upward shift in REG movement, corresponding to a positive or constructive correspondence to the binary output with the regularly alternating template sequence. Those described as contractive are correlated Ordering Random Events by Emotional Expression 213 with a downward shift in output or a destructive correspondence with the template. Both indicate increased order in the nominally random process. This finding is supported by an analysis of the REG output when patients were longing, an expansive emotion according to Reich. We found that longing, despite the fact that its expression in the study involved crying, was in all instances correlated with marked, highly anomalous upward shifts in REG output. Although there were only two events where longing was expressed (Study 2), their average z score was 1.27 (SD of 1.099). Establishing that the expressed emotions correspond to direction of REG response and that this fits Reich's hypotheses with respect to bioenergetic movement within the physical boundaries of the organism does not, however, explain the nonlocal effect on the REG. And, assuming that many of the same forces are operative in emotional expression and conscious intentionality, we cannot easily invoke electromagnetism as an explanatory mechanism since Jahn and Dunne found similar results in their local and nonlocal experiments, as well as in experiments performed atemporally (Jahn and Dunne, 1987). The issue is further complicated by the fact that directionality in REG output with the device used in this experiment is completely arbitrary: Once the original signal is electronically rectified, upward and downward shifts of the REG output do not mean that they are generated by physically greater or fewer noise source pulses. So, as much as one might be tempted to hypothesize some expansive physical force being emitted during, say, the expression of anger, which then secondarily causes an increased generation of electrical impulses in the REG, such a causal mechanism would be impossible given the lack of linkage between the original generation of electronic signals by the REG and the final direction of its output. The fact remains that we simply do not know enough at this time to satisfactorily explain these phenomena. We can, however, explore the following hypothesis for future experimentation. First we must answer the oft-asked question of the role of the experimenter in determining the outcome of the experiment. On beginning the first study, I was aware that I brought to it certain assumptions and intentions. These were my known beliefs related to the subject prior to beginning Study 1: 1. Conscious intention and passive participation in an experimental setup could effect anomalous changes in the output of machines. 2. Information could be anomalously transmitted from one person to another independent of time and distance. 3. Life processes, including the generation and expression of emotions, were not the result of a highly sophisticated mechanical concatenation of dead parts, but rather the expression of a spontaneously pulsatile life energy functioning within a membrane. 4. This same energy functioned at large, external to living systems, where it provided a medium from which electromagnetic impulses could emerge and through which they were transported. To my mind it 214 R. Blasband was possible, though definitely unproven, that it also served as the medium for the transmission of psychokinetic impulses, telepathic information, and other anomalous manifestations of consciousness. 5. If conscious intention and passive attention could anomalously affect an REG, so might spontaneous emotional expression. There was no conscious belief that opposite directions of REG output would be found to correlate with certain kinds of emotions, although I was certainly preconsciously aware of Reich's findings, noted above. (By preconscious I mean information that is accessible to conscious awareness, although not necessarily conscious at the time.) 6. At the beginning of Study 2, I was consciously aware of the bidirectionality of REG performance in correlation with emotional expression. I believe that my knowledge and biases prior to undertaking the investigation may have been a significant factor in obtaining our results, My current working hypothesis is that the investigator, patient, and REG cofunction in a state of resonance and that the REG output is a manifestation of the functional unity of the triad. REG anomalies in group situations, where members of the group exerted no conscious intention toward the device, were seen in PEAR studies reported by Nelson et al. (1996; 1998) and also independently by Radin, Rebman, and Cross (1996), Schwartz et al. (1997), and Rowe (1998). In their discussions of the possible cause/source of the observed anomalies, Nelson, Radin, and Schwartz postulate that the interacting participants of the group may generate a consciousness field "to which the REG responds via an anomalous decrease in the entropy of its nominally random output." (Nelson et al., 1996). Noting that the results of their (PEAR) benchmark REG and remote viewing experiments indicated a lack of dependence of the effects on time and distance, and assuming that the anomalous effects found in the group REG studies derive from the same basic phenomenon as the laboratory experiment, Nelson et al. conclude that "no conceptual models based on currently known physical fields with their usual 11r2 dependencies and very limited advanced and retarded signal capabilities are likely to suffice." They go on to suggest that in view of these facts and the further finding that, "the basic effects are analytically tantamount to small changes in the elemental binary probabilities underlying the otherwise random distributions," that the "anomalies may be more inforrnational than dynamical in their physical character." 3de Quincey states that such concepts as quantum field potentials and field consciousness "have far more to do with the nature of time and probabilities than with space. So-called quantum fields are not actually fields in any spatial sense. They are abstract mathematical descriptions of matrices of probabilities (of tendencies for certain events to occur). It is only the representations of such probabilities that take on the characteristics of fields. Probabilistic events, as tendencies of events to occur, are temporal-perhaps even psychological. In the end, statements of probability are statements about psychological expectations" (de Quincey, 1999). 216 R. Blasband Nelson, R. D., Jahn, R. G., Dunne, B. J., Dobyns, Y. H., & Bradish, G. J. (1998). FieldREG 11: Consciousness field effects: Replications and explorations. Journal of Scientific Exploration, 12,425-455. Panksepp, J. (1998). Affective neuroscience. Oxford, England: Oxford University Press. Radin, D. I., Rebman, M., & Cross, M. P. (1996). Anomalous organization of random events by group consciousness: Two exploratory experiments. Journal of Scientific Exploration, 10, 143-169. Reich, W. (1937). Experimental investigation of the electrical function of sexuality and anxiety. Republished in The Impulsive Character. Koopman, B . G. (Trans.) New York: Meridian, 1974. Reich, W. (1942). The function of the orgasm. New York: Meridian. Reich, W. (1948). The cancer biopathy. New York: Orgone Institute Press. Reich, W. (1 949). Character analysis. New York: Orgone Institute Press. Rowe, W. D. (1998). Physical measurements of episodes of focused group energy. Journal of Scientific Exploration, 12, 569-583. Rubik, B. (1995). Can western science provide a foundation for acupuncture? Alternative Therapies, 1, 41-43. Schwartz, G. E. R., Russek, L. G. S., Zhen-Su S., Song, L. Z. Y. X., & Xin, Y. (1997). Anomalous organization of random events during an international quigong meeting: Evidence for group consciousness or accumulated qi fields? Subtle Energies & Energy Medicine, 8, 55-65. Journal of Scientific Exploration, Vol. 14, No. 2, pp. 217-231,2000 0892-3310100 O 2000 Society for Scientific Exploration Energy, Fitness, and Information-AugmentedElectromagnetic Fields in Drosophila melanogaster Ditron LLC, R 0. 70 Box Excelsior, MN 55331 Department of Materials Science and Engineering, Stanford University Stanford, CA 94305 Abstract-Exposure of developing larvae to a few specific electromagnetic fields (EMFs) and information-augmented EMFs modified (a) the expression of larval development time, a genetically based trait relevant to development and whole organism fitness, and (b) a measure of energy metabolism, the [adenine triphosphateladenine diphosphate] ([ATP]l[ADP]) ratio in isofemale strains of Drosophila rnelanogaster. The study represents a compilation of approximately 10,000 larvae and 7,000 adults counted. The specific EM frequencies used in this study, 5.0, 7.3, 8.0, and 9.3 MHz at output power levels in the approximately 1 microwatt range, were produced by two small electronic devices of physically identical nature, but one was intentionally imprinted with specific information. Exposure periods varied from 4 hours to one life cycle. Larval development time was significantly shortened (approximately 10%) and the [ATP]/[ADP] ratio significantly increased (approximately 30%) in a Faraday cage without the EMFs compared to a Faraday cage with the specific EMFs. The Faraday cage represents a shielded environment that facilitates exposure to both fewer and specific electromagnetic frequencies. Larval development time results for development in the laboratory environment, which represents exposure to background EMFs of various frequencies, were intermediate. The information-augmented EMFs also gave intermediate results. Overall, there were no significant effects observed for the other measured fitness components-third instar larval weight, adult survival, and surviving adult weight. We discuss a thermodynamic model to account for our results and general bioelectromagnetic effects and attribute the change in development time to EMFs modifying electron transport chain activity and the [ATP]/[ADP] ratio via the influence of the EMFImagnetic vector potential upon electron availability and nicotinamide adenine dinucleotide levels. Keywords: fitness experiments-[ATP]/[ADP] ratio-Faraday cagesinformation-augmented and normal EMFs, theoretical models 218 M. J. Kohane & W. A. Tiller 1. Introduction Bioelectromagnetic studies have advanced considerably in the past decade (Goodman, Greenbaum, and Marron, 1995). For example, studies of nontherma1 effects on cells of the immune system from exposure to electromagnetic fields (EMFs) in the extremely low frequency range ( ~ 3 0 0 indicate that Hz) stimulatory, inhibitory, and no field exposure effects exist even for identical field parameters. The results depend upon the degree of cellular activation and the physical and biochemical boundary conditions experienced (Eichwald and Walleczek, 1996). Furthermore, low frequency EMFs influence specific RNA transcripts in human cells and transcription in Drosophila melanogaster cells (Goodman, Wei, and Henderson, 1989; Goodman et al., 1992), and Ho et al. (1992) have shown that weak static magnetic fields cause abnormalities in first instar larvae of D. melanogaster. A number of models have been proposed to account for EMF effects on biological systems. Ho et al. (1992) suggested that the weak static magnetic fields they studied must affect some cooperative process involved in pattern determination during critical stages of early Drosophila development. On the basis of the immune cell experiments, Eichwald and Walleczek (1996) suggested that external EMFs interact with cellular systems at the level of intracellular signal transduction pathways, specifically, Ca-signaling processes. Nossol, Buse, and Silny (1993) reported magnetic field influences on the in vitro redox activity of cytochrome oxidase activity. Additionally, weak EMFs stimulate adenine triphosphate (ATP) synthesis and alter Na,K-ATPase activity (Blank, Soo, and Papstein, 1995; Lei and Berg, 1998). Thus, magnetic fields may influence cellular energy metabolism, and Menendez (1996) has suggested that an electromagnetic coupling process may explain the proton translocation mechanism in cellular energy transfer. A significant association has been observed between stress, larval development time and aspects of energy metabolism, the cofactor nicotinamide adenine dinucleotide (NAD), and the [ATP]/[ADP] ratio in D. melanogaster (Kohane, 1988, 1994). In the present paper, we use the theoretical and experimental approach presented in these earlier papers and expand it to incorporate the intention-imprinted electronic device (IIED) techniques of Dibble and Tiller (1999) and investigate the hypothesis that both normal and informationaugmented EMFs may influence fitness and energy metabolism. We study fitness and the [ATP]/[ADP] ratio under nonstressful nutrient conditions, in the presence and absence of small electronic devices that produce EMFs of frequencies much higher than those used in previous studies. We assess these frequency effects using exposure periods from 4 hours to one life cycle in order to detect EMF effects that may not be observed at more conservative levels of EMF frequency and exposure period. Our results indicate that larval development time is significantly shortened and the [ATP]/[ADP] ratio significantly increased in a Faraday cage without the devices compared to a Faraday cage with the devices. The Faraday cage Electromagnetic Fields in D. rnelanogaster 219 represents a shielded environment with respect to electromagnetic (EM) radiation and facilitates exposure to fewer frequencies in the absence of devices and specific EM frequencies in the presence of devices. In addition, the IIEDs gave significantly better results on larval development time than did the unimprinted control devices. Overall, there are no significant effects for the other fitness components assessed, third instar larval weight, adult survival, and surviving adult weight. Thus, a reduction in exposure to EMFs increases one component of fitness, suggesting that EMFs may act as a biological stress (Goodman, Greenbaum, and Marron, 1995; Smith, 1996). Although we acknowledge the fact that EMFs may modify the larval environment (e.g., the food), we attribute the effects of the specific EMFs in this study to the modified [ATP]/[ADP] ratio as a consequence of altered NAD levels, electron availability, and electron transport chain activity. Finally, we discuss a thermodynamic model that may rationalize our observed fitness and energy changes and general bioelectromagnetic effects. 2. Experimental Methods (a) Strains We studied larvae obtained from two isofemale strains, Strains 1 and 2 of Kohane (1994). Nonstressful food was used for strain culture and experiments, and the food composition was as follows: 36 g agar, 108 g sugar, 72 g dry yeast, 10 ml propionic acid, and 24 ml nipagen in 2,000 ml water (Kohane, 1987). Separate constant temperature rooms (18"C, 55% relative humidity) were used for (a) device storage, (b) strain culture and unexposed adult culture, and (c) experiments. (b) Faraday Cages A standard Faraday cage consisted of a copper mesh screen enclosing a certain spatial volume. It is electrically grounded so the EM waves of wavelength larger than the mesh size, which impinge on the screen, will leak off to ground and only minimally penetrate to the interior space. Thus, the interior space has a greatly reduced EM integrated power density in the wavelength range larger than the copper mesh spacing. The one layer of copper mesh cages (dimensions: 40 x 40 x 30 cm) used here can be expected to reduce the EM field strength by a factor of approximately 10. (c) Electronic Devices Our experiments used two electronic devices in order to assess exposure to multiple frequencies and a single frequency as follows: (dl), a triple oscillator device producing frequencies of 5.0, 8.0, and 9.3 MHz; and (d2), a single oscillator device, producing a frequency of 7.3 MHz. Additionally, we studied two categories of EMFs produced by these devices. The first category 220 M. J. Kohane & W. A. Tiller involved devices (dl, o) and (d2, o), which had not been exposed to human informational influences. The second category involved devices (dl, j) and (d2, j), which had been exposed to human informational influences (see below). Thus, (dl, o) and (dl, j) and (d2, o) and (d2, j) constituted physically identical pairs of devices that differed only in the fact that one of each pair, (dl, j) and (d2, j), respectively, had been exposed to the human informational influence. The devices were individually wrapped in aluminum foil and stored in separate Faraday cages and were fabricated to be identical to those produced by Clarus Corporation (La Costa, CA). The triple oscillator device was powered by line voltage to 9V DC, and the single oscillator device was previously charged for 24 h by line voltage to 9 V DC and used with battery power. ( d ) Intentions The actual imprinting procedure was as follows: (a) The device was placed along with its current transformer on a table around which the imprinters sit; (b) Four people (two men plus two women) who were all accomplished meditators, coherent, inner-self managed and readily capable of entering an ordered mode of heart function (Tiller, 1997) and sustaining it for an extended period of time, sat around the table ready to enter a deep meditative state; (c) A signal was then given to enter such an internal state, to cleanse the environment and create a sacred space for the intention, then, a signal was given by one of the four to put attention on the tabletop objects and begin a mental cleansing process to erase any prior imprints from the device; (d) After 3 or 4 minutes, another signal was given to begin focusing on the specific prearranged intention statement for about 10-15 minutes; (e) Next, a final signal was given to shift focus to a closing intention designed to seal off the imprint into the device and minimize the leakage of the essential energylinformation from the devices. This completed the process, so the four people withdrew from the meditative state and returned to their normal state of consciousness. The specific intention was "to activate the indwelling consciousness of the device (d, j) so as to increase the concentration of NAD plus the activity of the available enzymes, dehydrogenases and ATP synthase in the mitochondria so that production of ATP is significantly increased relative to that produced in the unimprinted device (d, o)." This chemical transformation process in the cells of the fruit fly larvae was expected to significantly influence their fitness, which would manifest itself via a reduced larval development time for these larvae because they have a larger pool of ATP to work with (Kohane, 1994). (e) Fitness We conducted four similar experiments over a 6-month period and assessed EMF effects on the fitness of the strains using the above devices and different exposure periods. The experiments are summarized in Tables 1 and 2. We measured fitness at l8OC using the procedures given in Kohane (1988; 1994). Electromagnetic Fields in D. melanogaster TABLE 1 Summary of Fitness Experiments Treatment and replicate number Experiment 1 2 3 4fz Strain" 2 1 1 1 ~ e v i c e ~C dl d2 dl dl 16 16 15 15 Fc 15 15 14 15 d, j d 15 1.5 15 16 d, o d 16 16 15 16 Exposure period e 4hours 4hours 4days Lifecycleplus 4 days Date February 1997 February 1997 July 1997 July1997 Note: Fitness components assessed were larval development time, adult survival, and surviving adult weight. "Strain 1 and strain 2 refer to two isofemale strains. (dl) refers to a triple oscillator device producing frequencies of 5.0,8.0, and 9.3 MHz; and (d2) to a single oscillator device, producing a frequency of 7.3 MHz. The output power of the devices at the exposure distances is expected to be less than 1 pW. Categories of EMFs produced by these devices were as follows: (a) (d, o), devices which had not been exposed to human informational influences; (b) (d, j), devices which had been exposed to human informational influences-an intention concerned with significantly increasing the [ATP]/[ADP] ratio and decreasing larval development time. (F) refers to culture in a Faraday cage without a device, and (C) refers to culture in the laboratory environment. * (d, 0) and (d, j) refer to culture in a Faraday cage with a single device. "Experiments were conducted at 18OC. Experimental vial cultures involved 30 larvae (0-4 h old) transferred to a single vial containing nonstressful food. f ~ oExperiments 1-3, larvae were derived from unexposed adults, and for Experiment 4, larvae r were derived from exposed adults. gThe [ATP]/[ADP] ratio and third instar larval weight were assessed in Experiment 4 (see Materials and Methods). Experiments were conducted at 18OC since development time differences have been detected at this temperature for the strains studied here (Kohane, 1994). The treatments investigated in the experiments were as follows: (a) (C)culture in the random EMF environment of the laboratory; (b) (F)-culture in the relatively reduced EMF environment of the Faraday cage without a device; (c) (dl, o) and (d2,o)-culture in the relatively reduced EMF environment of the Faraday cage in the presence of a device that had not been associated with human informational influences; and (d) (dl, j) and (d2, j)-culture in the relatively reduced EMF environment of the Faraday cage in the presence of a device that had been associated with human informational influences. A single replicate involved 30 larvae (0-4 h old) transferred to a single vial containing nonstressful food (see above). For each experiment, all vials were established within a 3-hour period. Vials were transferred to Faraday cages and the cages were placed, at the same time, immediately next to each other on the same bench in a constant temperature room. Treatment (C) involved vials concurrently transferred to a tray, which was placed on the lid of treatment (F). We used vial cultures as a thermocouple and could not detect temperature variation between treatments in each experiment. Exposure of vials to devices in Faraday cages was achieved as follows: The 222 M. J. Kohane & W. A. Tiller TABLE 2 Results for Each Experimental Condition Described in Table 1 Treatment Variable a C F j o Experiment 1 16.276 (0.170) 2 15.504 (0.097) 3 15.964 (0.315) 4 16.507 (0.087) ~xperiment~ 1 0.917 (0.833) (0.933) 2 0.900 (0.842) (0.915) 3 0.933 (0.830) (0.938) 4 0.867 (0.833) (0.967) Experimentc 1 1.140 (0.049) 2 1.155 (0.045) 3 1.267 (0.099) 4 1.254 (0.047) Experiment 4d 1.478 (0.157) Assaye NAD 1.305 (0.046) Pure water 0.486 (0.035) -- "Larval development time (T,, in days). Means and standard deviations are given in parentheses. Adult survival given as proportions for an input of 30 larvae. Medians and 25 and 75 percentiles are shown in parentheses for adult survival since the data distribution curves for each batch were not normal. The 25 percentiles are given before the 75 percentiles. 'Surviving adult weight (mg). Means and standard deviations are given in parentheses. dLar~al weight in mg. Means and standard deviations are given in parentheses. '[ATP]/[ADP] ratio. Means and standard deviations are given in parentheses. respective device was placed in the center of a Faraday cage and vials/bottles were transferred to cages and placed around the perimeter. The device was then removed at the end of a specific time period and larval development proceeded until adults eclosed. The device was approximately 15 cm from the vials/bottles on average, where the output power of the devices in their specific frequency ranges is expected to be less than 1 microwatt. Vials were monitored daily and surviving adults were collected and weighed. Surviving adult weight was calculated for each vial as the weight of the number of flies surviving divided by the number of flies surviving and is given in mg. Larval development time is given as T,, the time taken for half of the surviving adults to emerge. Electromagnetic Fields in D. melanogaster 223 In Experiments 1-3 (Table l ) , larvae were derived from unexposed adults. In Experiment 4 (Table l), larvae were derived from exposed adults of Strain 1 as follows: Replicate bottle cultures were exposed to all treatments-(C), (F) (dl, j), (dl, o) for one life cycle. Emerging adults were synchronously collected from the respective cultures and used to generate larvae for experimental vials. These were then exposed to the same treatment as the parents using a 4day exposure period for devices. ( ' Energy Levels During Larval Development The energy assay follows Kohane (1994) and primarily detects electron transport chain activity. We measured the [ATP]/[ADP] ratio in third instar larval homogenates where the glucose-hexokinase reaction, ADP-regenerating reactions, and glycolysis and the citric acid cycle were not functional or were severely impaired (Kohane, 1994). Supplemental dH,O does not alter the [ATP]/[ADP] ratio in these homogenates. Supplemental 0.0 1 M NAD increases the [ATP]/[ADP] ratio as a consequence of altered electron transport chain activity (Kohane, 1994). Kohane (1 994) stored larval homogenates supplemented with NAD and dH,O on ice for a reaction time of 7 minutes to facilitate metabolic activity, prior to determination of the [ATP]/[ADP] ratio. In order to assess treatment, fitness, and the [ATP]/[ADP] ratio in the same experiment, we established experimental vials for energy assays concurrently with Experiment 4 of Table 1. We also adopted a more conservative approach and stored the homogenates, supplemented with NAD and dH,O, in their respective treatments for a longer reaction time of 40 minutes. Thus, the assay assesses in vivo and in vitro effects of EMFs on the [ATP]/[ADP] ratio (KOhane, 1994). We avoided age-dependent effects on energy metabolism by preparing larval homogenates on the same day as follows: 15 third instar larvae were collected, transferred to microcentrifuge tubes, weighed (mg), and homogenized in 250 p1 ice-cold dH,O to which 250 p1 of ice-cold 0.01 M NAD or dH,O were immediately added. The solutions were mixed and transferred to each treatment for a reaction time of 40 minutes to facilitate metabolic activity. The microcentrifuge tubes were then freeze-clamped in liquid nitrogen. ATP, ADP, and AMP (adenine monophosophate) were extracted using 4.2 M formic acid and 4.2 M ammonium hydroxide (Kohane, 1994). They were quantified using an automated Isco high performance liquid chromatography (HPLC) apparatus, a Vydac column (302IC4.6), and a preprogrammed gradient from 0.025 M sodium phosphate monobasic pH 2.8 to 0.5 M sodium phosphate monobasic pH 2.8 (modified from Kohane, 1994). There were eight replicates per treatment for both supplemental NAD and dH,O. All extractions were completed within a 4-hour period. Replicates were stored at 2-4°C and assayed in random order during a 3day period. 224 M. J. Kohane & W. A. Tiller 3. Results ( a ) EMFs and Fitness We have used visual inspection of Figure 1 to assess the homogeneity of the treatment results across experiments. Considering larval development time, (C) and (F) were not homogeneous and rankings for (C) were as follows: (C2) < (C3) I (Cl) < (C4). The rankings for (F) were as follows: (F3) < (F2)1 (F4) I (Fl). (d, j) was essentially homogeneous. Considering (d, o), the rankings were as follows: (o 1) = (02) = (03) < (04). All treatments appeared to be homogeneous across experiments for adult survival. Considering surviving adult weight, treatments were not homogenous across experiments. Rankings were as follows: (C) - (Cl) I (C2) < (C4) 1 (C3); (F) - (Fl) = (F2) < (F3) = (F4); (d, j) - (jl)= (j2) < (j4) < (j3); and (d, 0)- (01) = (0 2) < (04) < (03). Finally, larval weight was only assessed in Experiment 4. On the basis of the observed heterogeneity, we have considered treatment comparisons for each experiment and have used visual inspection of Figure 1 to assess these comparisons. In our experiments, comparisons between (d, o) and (d, j) indicate the effects of exposure to the two categories of EMFs-the intention effect. The presence of (d, j) significantly decreased larval development time in comparison to (d, o) in all experiments. We did not observe a significant device effect for adult survival, surviving adult weight, and larval weight. Comparisons between (C) and (F) indicate the effect of a reduction in exposure to random EMFs-the Faraday cage effect. The presence of the Faraday cage significantly decreased larval development time in comparison to (C) in all experiments. We did not observe a significant Faraday cage effect for adult survival and larval weight. Considering surviving adult weight, a significantly higher weight was observed in Experiments 3 and 4 for (F) in comparison to (C). Comparisons between (F) and (d, o) and (d, j) indicate the effects of the addition of specific EMFs to the Faraday cage environment-the oscillator effect. In all experiments, the presence of the Faraday cage without a device (F) significantly decreased development time in comparison to both (d, j) and (d, 0). As noted above, the effect of (d, j) was to decrease development time in comparison to (d, 0). We did not detect any significant comparisons for both adult survival and larval weight. Considering surviving adult weight, no treatment effects were observed for Experiments 1, 2, and 3. However, in Experiment 4 the treatment rankings were as follows: (F) > (d, j) = (d>o). Electromagnetic Fields in D. rnelanogaster Treatment and Experiment Treatment and Experiment Treatment and Experiment Treatment and Experiment Fig. I . EMFs and fitness-Notched box plots are given for the experiments described in Table 1 as follows: (a) larval development time (the time taken for half of the surviving flies to emerge, T,, in days); (b) adult survival for an input of 30 larvae, given as a proportion; (c) surviving adult weight (mg); and (d) larval weight (Experiment 4 only, mg). The x-axis gives the treatment (C, F, j, and o) for each experiment (1, 2, 3, and 4) as described in Table 1. Note: A notched box plot provides a simple graphical summary of a batch of data and implements confidence intervals on the shown median values. The boxes are notched at the median and return to full width at the lower and upper confidence interval values. If the intervals around two medians do not overlap, then one can be confident at about the 95% level that the two population medians are different. Outside values are represented by an asterisk and far outside values by an open circle. 226 M. J. Kohane & W. A. Tiller (b) EMFs and the [ATP]/[ADP]Ratio Results for the [ATP]/[ADP] ratio in the presence of both NAD and dH20 are given in Figure 2. Again, we have used visual inspection of the figure to assess the comparisons described above. Comparisons between (d, j) and (d, o) indicated significant differences for both assays, with a higher [ATP]/[ADP] ratio observed for (d, j). Comparisons between (F) and (C) indicated a dependence on the assay as follows: (F) > (C) for NAD and (F) = (C) for dH20.Comparisons between (F) and (d, j) and (d, o) indicated that (F) produced a higher [ATP]/[ADP] ratio than (d, j), which was greater than (d, 0). Differences between treatments were more evident for supplemental NAD than dH20. Finally, the Pearson correlation coefficient for the [ATP]/[ADP] ratio and larval development time in Experiment 4 was -0.856. 4. Discussion A reduction in exposure to random laboratory EMFs and two categories of EMF frequencies shortened larval development time and increased the [ATP]/[ADP] ratio in D. melanogaster. The magnitude of this effect appeared to be modified by the category of EMFs assessed, and we suggest that this may occur via modified and information-augmented EMFs (see below). This is another example of intention imprinting having a significant effect on physical reality (Dibble and Tiller, 1999). Differences observed for the [ATP]/[ADP] ratios were not large, but were significant, indicating that the EMFs studied here modified energy metabolism. Although our results for developing larvae Treatment and Assay Fig. 2. EMFs and the [ATP]/[ADP] ratio in the presence of NAD and dH,O as assayed in Experiment 4 of Table 1. The x-axis gives the treatment (C, F, j, and o) for each assay (1: NAD and 2: dH,O). Electromagnetic Fields in D. melanogaster 227 1 I I indicated a decrease in the [ATP]/[ADP] ratio in the presence of EMFs, other studies have shown that EMFs (15, 50 Hz) may stimulate both ATP synthesis and cell proliferation in Corynebacterium glutamicum and Sacchharomyces cerevisiae (Lei and Berg, 1998; Mehedintu and Berg, 1997). This contrast may be a consequence of the higher EMF frequencies studied here, the use of the Faraday cage, and the fact that both the biological system under study and its status influence the effects of EMFs (Goodman, Greenbaum, and Marron, 1995). Larger treatment differences were observed for the [ATP]/[ADP] ratio and supplemental NAD, in comparison to supplemental dH,O (Figure 2). This result was in accord with Kohane (1994) where the [ATP]/[ADP] ratios were also higher, as a consequence of the shorter reaction time. As previously noted, supplemental NAD increases electron transport chain activity and the [ATP]/[ADP] ratio in larval homogenates. Hence, EMF effects on energy metabolism may be greater when the electron transport chain is active. EMF effects on energy metabolism may manifest both in vivo and in vitro, and we will assess details of this in future experiments. A higher [ATP]/[ADP] ratio was observed for the treatments that produced a shorter development time, providing additional evidence for a relationship between energetics and fitness (Parsons, 1997; Watt, 1985). Our results also suggest that that EMFs may function as a biological stress (Goodman, Greenbaum, and Marron, 1995; Smith, 1996), possibly via change in energy metabolism, since exposure to EMFs lengthened larval development time, which is associated with total fitness: Earlier sexual maturity leads to higher fitness (Cole, 1954; Hoffmann and Parsons, 199 1; Parsons, 1997). We detected significant effects on larval development time, but not on third instar larval weight, adult survival, and, in general, surviving adult weight. In this regard, Veicsteinas et al. (1996) studied the development of chicken embryos exposed to an intermittent EMF of 50 Hz. Statistical comparisons between exposed and sham-exposed values did not show significant differences for the variables studied-extracellular membrane components and histological examinations of brain liver and heart during development, egg fertility and egg weight, body weight of chickens at 90 days from hatching, and histological analysis of body organs. Furthermore, Morganato et al. (1995) concluded that continuous exposure to a 50 Hz magnetic field during 70% of the life span of rats prior to sacrifice did not significantly alter growth rate; the morphology and histology of liver, heart, lymph nodes, testes, and bone marrow. Hematology, hematochemistry, and the neurotransmitters dopamine and serotonin were also unaffected by the exposure. The significant results observed here for larval development time may be a consequence of higher EMF frequencies, the Faraday cage, and the biological system under study (see above). The decrease in larval development time observed here did not appear to be due to a trade off with the measured fitness components but may also arise as a trade off with other life-history traits (Hoffmann and Parsons, 1991; Odum, 228 M. J. Kohane & W. A. Tiller 1985). In the present study, we assessed EMF frequency effects under nonstressful nutrient conditions and a relatively low larval density (Kohane, 1988, 1994). Fitness differences are enhanced under stress, and, therefore, we may both detect differences for additional fitness components and clarify the surviving adult weight results under more stressful nutrient conditions (Kohane, 1994). We also note that the EMF effects may be due to modification of the larval environment, in particular the food, and we will assess this idea in future experiments studying EMF effects on larval energy metabolism in vitro. Treatment results across the four experiments were not homogeneous for larval development time and adult survival. Of interest were the following: (a) the relative homogeneity of the effect of (d, j) on larval development time suggesting that the intention may produce a constrained larval development time and (b) the shorter development time observed for all treatments in Experiment 3, in comparison to Experiment 4, suggesting that exposure period may influence this fitness component. Considering surviving adult weight, significant differences were only observed in Experiments 3 and 4, where the Faraday cage effect produced the highest weights (F). The results for Experiment 4 suggest that cross-generational exposure to EMFs may affect fitness in ways that differ from single generation exposure. However, in general there did not appear to be specific patterns to the heterogeneity. Hence, at this time we cannot conclude much about strain, specific device (single or triple oscillator), and exposure period, and we will assess these aspects in detail in future experiments. The thermodynamic basis for an EMF effect on biological systems is a consequence of the influence of the standard electrochemical potential upon biological processes. Standard thermodynamic theory (Tiller, 1990) shows us that the electrochemical potential, q i , of species j in medium A, is given by Here, p, = standard state chemical potential, T = temperature, k = Boltzmann's constant, a = chemical activity, ( a = y, where c = concentration and y = activity coefficient), e = electron charge, z = valence, V = electrostatic potential, E = electrical permittivity, m = magnetic permeability, I? = electric field, and H = magnetic field. One finds that zJ = +1 and -1, respectively, for the proton and electron. Thus, the last term represents the applied EMF contribution to the standard electrochemical potential and, hence, to various biological systems. The particular physical variable or quantum potential that may be important in our study is the magnetic vector potential (MVP), denoted A , which creates an electric field E, and magnetic field (Jackson, 1962; Kraus and Carver, 1973; Tiller, 1993, 1997). Therefore, we attribute the EMF and augmented EMF effects on larval development time to the influence of the MVP upon the activity of the electron transport chain activity and the [ATP]/[ADP] ratio as follows: the MVP diminished electron transfer to the electron transport chain from the NAD pool and cellular electron availability. The diminished electron Electromagnetic Fields in D. melanogaster 229 transfer modified the redox potential causing lowered electron transport through the electron transport chain and decreased proton pumping and energy availability. This effect may have decreased the proton motive force across the inner mitochondria1 membrane, subsequently modifying cytosolic phosphorylation status (Brand and Murphy, 1987; Kohane, 1994). The above approach suggests that the MVP may (a) modify NAD levels and (b) interact with the electron transport chain at positions beyond electron availability. We will assess these hypotheses in future experiments studying EMF effects on larval energy metabolism in vitro. Discussion of the results concerning the categories of EMFs and the possibility of augmenting EMFs requires that one look to a deeper understanding of the quantum vacuum or so-called empty space. This has been established by quantum theory as a chaotic, virtual particle sea of boundless energy (energy density equivalent to approximately glcc) at the quantum relativity level. An interaction occurs between this virtual particle sea and the fundamental particles of physical matter to ultimately determine the magnitude of p, in Equation 1 (Feynman and Hibbs, 1965; Lee, 1988; Tiller, 1997). Change in characteristics of the physical vacuum should, in turn, change the ground state energy of fundamental particles, atoms, molecules, and biological moieties (Tiller, 1993). These changes may manifest in biological systems as a consequence of the relationship between the ground state energy of fundamental particles (or charge carrying species) and the standard state electrochemical potential (p, in Equation 1). Tiller (1997) suggested that directed human intention is capable of altering characteristics of the physical vacuum. Thus, if a human intention can shift these characteristics even a tiny amount, the ground state energy of the electron would be appreciably altered. Thus, the primary effect of intention on biological systems may occur by modification of this ground state energy for the charge carrying species and, thus, the standard state electrochemical potential for atoms containing such particles. Therefore, pi, in Equation 1 may be affected by directed human intention. The magnetic vector potential is thought to be involved with human intention (Tiller, 1993, 1997). If we call the incremental change in A associated with a specific focused intention, AA , then standard electrodynamic equations yield changes A E and A H (Jackson, 1962; Kraus and Carver, 1973) and the intention effect may both enter physical reality via Equation 1 and influence energy metabolism and fitness. Although others have proposed a kinetic rate-limiting step for EMF effects on biological systems (Eichwald and Walleczek, 1996, see above), a sound basis also exists for a thermodynamic potential change as the rate-limiting step. The final apportioning of the total effect between thermodynamic and kinetic categories awaits further research. In conclusion, the process path from EMFs to a biological effect in this work is likely to be via an influence of the magnetic vector potential on electron transport chain activity. The intentionaugmented EMF effect is likely to just insert contributions (Ap, and AA) into 230 M. J. Kohane & W. A. Tiller Acknowledgments We wish to acknowledge Professor W. B. Watt, Department of Biological Sciences, Stanford University, for technical help and the loaning of facilities and equipment. We also acknowledge Durance Corporation and Ditron LLC for financial assistance and laboratory facilities. References Blank, M., Soo, L., & Papstein, V. (1995). Effects of low frequency magnetic fields on Na,K-ATPase activity. Bioelectrochemistry and Bioenergetics, 38,267-273. Brand, M. D., & Murphy, M. P. (1987). Control of electron flux through the respiratory chain in mitochondria and cells. Biology Review, 62, 14 1-193. Cole, L. C. (1954). The population consequences of life-history phenomena. Quarterly Review of Biology, 29, 103-107. Dibble, W. E., Jr., & Tiller, W. A. (1999). Electronic device mediated pH changes in water. Journal of Scientific Exploration, 13, 155-176. Eichwald, C., & Walleczek, J. (1996). Activation-dependent and biphasic electromagnetic field effects: Model based on cooperative enzyme kinetics in cellular signaling. Bioelectromagnetics, 17, 427-435. Feynman, R. P., & Hibbs, A. R. (1965). Quantum mechanics and path integrals. New York: McGraw-Hill. Goodman, E. M., Greenbaum, B., & Marron, M. T. (1995). Effects of electromagnetic fields on molecules and cells. International Review of Cytology, 158, 279-338. Goodman, R., Wei, L. X. J., & Henderson, A. (1989). Exposure of human cells to low-frequency electromagnetic fields results in quantitative changes in transcripts. Biochemica et Biophysica Acta, 1009, 216-220. Goodman, R., Weisbrot, D., Uluc, A., & Henderson, A. (1992). Transcription in Drosophila melanogaster salivary gland cells is altered following exposure to low-frequency electromagnetic fields. Bioelectromagnetics, 13, 11 1-1 18. Ho, M. W., Stone, T. A., Jerman, I., Bolton, J., Bolton, H., Goodwin, B. C., Saunders, P. T., & Robertson, F. (1992). Brief exposure to weak static magnetic fields during early embryogenesis cause cuticular pattern abnormalities in Drosophila larvae. Physics in Medicine and Biology, 37, 1171-1179. Hoffmann, A. A., & Parsons, P. A. (1991). Evolutionary genetics and environmental stress. Oxford, England: Oxford University Press. Jackson, J. 0 . (1962). Classical electrodynamics. New York: John Wiley and Sons. Kohane, M. J. (1987). Fitness relationships under different temperature regimes in Drosophila melanogaster. Genetica, 72, 199-210. Kohane, M. J. (1988). Stress, altered energy availability and larval fitness in Drosophila rnelanogaster. Heredity, 60, 273-281. Kohane, M. J. (1994). Energy, development and fitness in Drosophila melanogaster. Proceedings of the Royal Society, Series B, 257, 185-191. Kraus, J. D., & Carver, K. R. (1973). Electromagnetics. New York: McGraw Hill. Lee, T. D. (1988). Particle physics and introduction to field theory. London: Harwood Academy. Lei, C., & Berg, H. (1998). Electromagnetic window effects on proliferation rate of Corynebacterium glutamicum. Biolectrochemistry and Bioenergetics, 45, 261-265. Mehedintu, M., & Berg, H. (1997). Proliferation response of yeast Sacchharomyces cerevisiae on electromagnetic field parameters. Biolelectrochemistry and Bioenergetics, 45, 261-265. Menendez, R. G. (1996). An electromagnetic coupling hypothesis to explain the proton translocation mechanism in mitochondria, bacteria and chloroplasts. Medical Hypotheses, 47, 179-182. Morganato, V., Nicolini, P., Conti, R., Zecca, L., Veicsteinas, A., & Cerretelli, P. (1995). Biological effects of prolonged exposure to ELF electromagnetic fields in rats: 11. 50 Hz magnetic fields. Bioelectromagnetics, 16, 343-355. Nossol, B., Buse, G., & Silny, J. (1993). Influence of weak static and 50 Hz magnetic fields on the redox activity of cytochrome-C oxidase. Bioelectromagnetics, 14, 361-372. Odum, E. P. (1985). Trends expected in stressed ecosystems. Bioscience, 35,419-422. Electromagnetic Fields in D. melanogaster 23 1 Parsons, P. A. (1997). Stress-resistance genotypes, metabolic efficiency and interpreting evolutionary change: From living organisms to fossils. In Bijsma R., & Loeschcke V. (Eds.), Environmental stress, adaptation and evolution (pp. 291-305). Basel, Switzerland: Birkhauser Verlag. Smith, 0. (1996). Cells, stress and EMFs. Nature Medicine, 2, 23-24. Tiller, W. A. (1990). The science of crystallization: Microscopic interface processes. London: Cambridge University Press. Tiller, W. A. (1993). What are subtle energies? Journal of Scientific Exploration, 7, 293. Tiller, W. A. (1997). Science and human transformation: Subtle energies, intentionality and consciousness. Walnut Creek, California: Pavior Publishing. Veicsteinas, A., Belleri, M., Cinquetti, A., Parolini, S., Barabato, G., & Molinari Tosatti, M. P. (1996). Development of chicken embryos exposed to an intermittent horizontal sinusoidal 50 Hz magnetic field. Bioelectromagnetics, 17, 41 1-424. Watt, W. B. (1985). Bioenergetics and evolutionary genetics-opportunities for new synthesis. American Naturalist, 1275, 11 8-143. Journal of Scientific Exploration, Vol. 14, No. 2, pp. 233-255,2000 0892-33 10/00 O 2000 Society for Scientific Exploration A Dog That Seems to Know When His Owner Is Coming Home: Videotaped Experiments and Observations 20 Willow Road London NW3 I TJ, UK e-mail: ars @dircon.co.uk 173 Kay Brow Ramsbottom, Bury, BLO 9AI: UK Abstract- Many dog owners claim that their animals know when a member of the household is about to come home, showing their anticipation by waiting at a door or window. We have investigated such a dog, called Jaytee, in more than 100 videotaped experiments. His owner, Pam Smart (P.S.) traveled at least 7 km away from home while the place where the dog usually waited for her was filmed continuously. The time-coded videotapes were scored blind. In experiments in which P.S. returned at randomly selected times, Jaytee was at the window 4% of the time during the main period of her absence and 55% of the time when she was returning ( p c .0001). Jaytee showed a similar pattern of behavior in experiments conducted independently by Wiseman, Smith, and Milton (1998). When P.S. returned at nonroutine times of her own choosing, Jaytee also spent very significantly more time at the window when she was on her way home. His anticipatory behavior usually began shortly before she set off. Jaytee also anticipated P.S.'s return when he was left at P.S.'s sister's house or alone in P.S.'s flat. In control experiments, when P.S. was not returning, Jaytee did not wait at the window more and more as time went on. Possible explanations for Jaytee's behavior are discussed. We conclude that the dog's anticipation may have depended on a telepathic influence from his owner. Keywords: dog-anticipation-telepathy-human-animal bonds Introduction Many dog owners claim that their animal knows when a member of the household is about to come home. Typically, the dog is said to go and wait at a door, window, or gate while the person is on the way home (Sheldrake, 1994, 1999a). Random household surveys in Britain and the United States have shown that between 45% and 52% of dog owners say they have noticed this kind of behavior (Brown and Sheldrake, 1998; Sheldrake, Lawlor, and Turney, 1998; Sheldrake and Smart, 1997). 234 R. Sheldrake & P. Smart Dog owners often ascribe their animal's anticipations to telepathy or a sixth sense, but there could be more conventional explanations: First, the dog could be hearing or smelling its owner approaching. Second, the dog could be reacting to routine times of return. Third, it could be responding to subtle cues from people at home who know when the absent person is returning. Fourth, the animal may go to the place at which it waits for its owner when the person is not on the way home; the people at home may remember its apparent anticipation only when the person returns shortly afterward, forgetting the other occasions. Thus the phenomenon could simply be an artifact of selective memory. In order to test these possibilities, the dog should be capable of reacting at least 10 minutes in advance, the person to whom the dog responds should come home at nonroutine times, the people at home should not know when this person is coming, and the behavior of the dog should be recorded in such a way that selective memory can be ruled out (Sheldrake, 1994). This recording of the dog's behavior can be done most effectively by means of time-coded videotape. In this paper we describe a series of videotaped experiments and observations with a dog called Jaytee, belonging to Pamela Smart (P.S.). Jaytee's Anticipatory Behavior P.S. adopted Jaytee from Manchester Dogs' Home in 1989, when he was still a puppy, and soon formed a close bond with him. She lived in Ramsbottom, Greater Manchester, in a ground-floor flat, adjacent to the flat of her parents, William and Muriel Smart, who were retired. When she went out, she usually left Jaytee with her parents. In 1991, when P.S. was working as a secretary in Manchester, her parents noticed that Jaytee used to go to the French window in the living room almost every weekday at about 4:30 p.m., around the time she set off to come home. Her journey usually took 45-60 minutes, and Jaytee would wait at the window most of the time she was on her way. Since she worked routine office hours, the family assumed that Jaytee's behavior depended on some kind of time sense. P.S. was laid off from her job in 1993 and was subsequently unemployed. She was often away from home for hours at a time and was no longer tied to any regular pattern of activity. Her parents usually did not know when she would be returning, but Jaytee still continued to anticipate her return. His reactions seemed to occur around the time she set off on her homeward journey. In April 1994, P.S. read an article in the British Sunday Telegraph about the research Rupert Sheldrake (R.S.) was doing on this phenomenon (Matthews, 1994) and volunteered to take part. The first stage in this investigation was the keeping of a log by P.S. and her parents. Between May 1994 and February 1995 on 100 occasions she left Jaytee with her parents when she went out, and they made notes on Jaytee's reactions. P.S. herself kept a record of where she Experiments With a Return-Anticipating Dog 235 had been, how far she had traveled (usually at least 6 km and sometimes 50 km), her mode of transport, and when she had set off to come home. On 85 of these 100 occasions, Jaytee reacted by going to wait at the French window in the living room before P.S. returned, usually 10 or more minutes in advance. When these data were analyzed statistically, a linear regression of Jaytee's waiting times against P.S.'s journey times showed that the times when Jaytee began waiting were very significantly ( p < .0001) related to the times that P.S. set off (Sheldrake and Smart, 1998). It did not seem to matter how far away she was. Jaytee's anticipatory reactions usually began when P.S. was more than 6 km away. He could not have heard her car at such distances, especially against the background of the heavy traffic in Greater Manchester and on the M66 motorway, which runs close to Ramsbottom. Moreover, Mr. and Mrs. Smart had already noticed that Jaytee still anticipated P.S.'s return when she arrived in unfamiliar vehicles. Nevertheless, to check that Jaytee was not reacting to the sound of P.S.'s car or other familiar vehicles, we investigated whether he still anticipated her arrival when she traveled by unusual means: by bicycle, by train, and by taxi. He did (Sheldrake and Smart, 1998). P.S. did not usually tell her parents in advance when she would be coming home, nor did she telephone to inform them. Indeed, she often did not know in advance when she would be returning after shopping, visiting friends and relations, attending meetings, or an evening out. But it is possible that her parents might in some cases have guessed when she might be coming, and then, consciously or unconsciously, communicated their expectation to Jaytee. Some of his reactions might therefore be due to her parents' anticipation, rather than depending on some mysterious influence from P.S. herself. To test this possibility, we carried out experiments in which P.S. set off at times selected at random after she had left home. These times were unknown to anyone else. In these experiments, Jaytee started to wait when she set off, even though no one at home knew when she would be coming (Sheldrake and Smart, 1998). Therefore his reactions could not be explained in terms of her parents' expectations. Our first series of investigations involved the recording of Jaytee's reactions in a notebook and, hence, necessitated a subjective assessment of his behavior. In this paper we describe a preplanned series of 12 experiments with randomly chosen return times in which Jaytee's behavior was recorded throughout the entire period of P.S.'s absence on time-coded videotape. We also discuss four videotaped experiments with randomly chosen return times carried out with Jaytee at our invitation by Wiseman, Smith, and Milton (Sheldrake, 1999b; Wiseman, Smith, and Milton, 1998). In addition, we describe 95 videotaped observations of Jaytee's behavior in three different environments. We made these observations to find out more about the natural history of the dog's anticipatory behavior, On these occasions, P.S. did not return at randomly selected times, but rather at times of her 236 R. Sheldrake & P. Smart own choosing. She went out shopping, visiting friends or members of her family, attending meetings, or visiting pubs and returned when she felt like it. Her journeys varied in distance between 7 and 22 km away from home. They took place at various times of the day or evening and followed no routine pattern. When she left Jaytee with members of her family, they were not informed when she would be returning, and she usually did not know in advance herself. On 50 occasions, Jaytee was left on his own. We also carried out a series of 10 control observations in which Jaytee was filmed continuously on evenings when P.S. was not returning home, or was returning unusually late. Methods In these experiments, when P.S. went out she left Jaytee either with her parents, William and Muriel Smart; or alone in her own flat in Ramsbottom, Greater Manchester, next door to her parents' flat; or with her sister, Cathie MacKenzie, in the nearby town of Edenfield. Having left Jaytee, P.S. traveled a minimum distance of 7 km. She recorded in a notebook the details of where she had been, when she set off to come home, how long her journey took, and her mode of transport. In some cases she traveled in taxis or in cars belonging to her sisters or friends, but in most cases she traveled in her own car, since we had already established that Jaytee's anticipatory behavior still occurred when she traveled in unfamiliar vehicles, and hence could not be explained in terms of the dog hearing her car (Sheldrake and Smart, 1998). While P.S. was out, Jaytee's visits to the window and his absences from it were monitored continuously on videotape. The videotaping procedure was kept as simple as possible, so that the filming of Jaytee could be done routinely and automatically. The video camera was set up on a tripod and left running continuously in the long-play mode with a long-play film, with the time-code recorded on it. In this way up to 4 hours of continuous observation was possible without anyone needing to attend to the camera. P.S. switched the camera on just before she left and switched it off when she returned. Because of the need to keep Jaytee's visits to the window under continuous surveillance, all experiments involved absences of less than 4 hours. The camera pointed at the area where Jaytee usually waited. In both P.S.'s parents' flat and in P.S.'s own flat (a ground-floor flat adjacent to her parents') this was by the French window in the living room, through which he could see the road outside where P.S. drew up and parked her car. In P.S.'s sister's house, Jaytee jumped up onto the back of a sofa from which he could see out of the window. Experiments With Randomly Selected Return Times In a preplanned series of 12 experiments with randomly selected return times, Jaytee was left at P.S. parents' flat and neither P.S. nor her parents knew Experiments With a Return-Anticipating Dog 237 in advance when she would be returning. In all these experiments, P.S. traveled in her own car. P.S. was beeped on a telephone pager when it was time to set off home. On most occasions, the random selection of the times and the beeping of P.S. were done by R.S., who was in London, over 300 km away. On two occasions (on November 19, 1996 and July 1, 1997) the selection of random times and the beeping was done by another person in London who was unknown to P.S. and Jaytee. These beep times were within a prearranged period, between 45 and 90 minutes long. This period commenced 80 to 170 minutes after P.S. had gone out. The beep window was then divided into 20 equal intervals, and one of these was selected at random by throwing a die three times, to determine the page, row, and column in standard random number tables (Snedecor and Cochran, 1967). Reading downward from this point looking at the first two digits of each random number, the first pair of digits between 01 and 20 determined the time at which the beep was to be given. Three of the 12 experiments were carried out in the afternoon, with beeps at 2:22, 3:04, and 3:36 p.m.; the remaining experiments were carried out in the evening with beeps at a range of times between 8:09 and 9:39 p.m. Observations in Diflerent Environments We carried out a preplanned series of 30 observations in P.S.'s parents' flat between May 1995 and July 1996. Seven of P.S. absences were in the daytime, at various times in the morning and afternoon, with P.S.'s times of return ranging from 11:13 a.m. to 3:36 p.m. Twenty-three were in the evening, with P.S. returning at a range of times between 7:30 and 10:45 p.m. The length of her absences ranged from 85 to 220 minutes. In P.S.'s parents' flat we also carried out a preplanned series of 10 control experiments on evenings when P.S. was not returning or coming home unusually late. Her parents were not informed that she would not be returning during the 4-hour period that the videotape was running. This series of observations was made between July and November 1997, during the period when we were doing experiments with randomly selected return times. We also carried out a preplanned series of 50 observations in P.S.'s own flat, where Jaytee was left on his own, between May 1995 and September 1997. On 15 of these occasions, P.S. went out and returned in the morning, with times of return ranging from 9:59 to 11:57 a.m.; on 34 occasions she returned in the afternoon, at a range of times between 12:20 and 4:50 p.m.; and on one occasion she returned in the evening, at 9:27 p.m. The length of her absences ranged from 81 to 223 minutes. The five observations at P.S.'s sister's house were conducted between October 1995 and June 1996, two in the morning and three in the evening, with absences ranging from 93 to 199 minutes. 238 R. Sheldrake & P. Smart Analysis of Videotapes and Tabulation of Data The videotapes were analyzed blind by Jane Turney and/or Dr. Amanda Jacks, who did not know when P.S. set off to come home or other details of the experiments. Starting from the beginning of the tape, they recorded the exact times (to the nearest second) when Jaytee was in the target area near the window and made notes on his activities there: for example that he was barking at a passing cat, sleeping in the sun, or sitting looking out of the window for no apparent reason. In cases where the same tape was scored blind by both people, the agreement between their records was excellent, showing occasional differences of only a second or so. (Although the scoring was carried out blind, when the end of the tape was reached and P.S. was seen entering the room, the judges then knew at what time she had arrived, and hence were no longer blind. But by this time the data had all been recorded and were not subsequently altered.) Some of the videotapes were also scored independently by P.S. and R.S. to see how well their records corresponded to each other and to the blind scores by Jane Turney of Amanda Jacks. Again the agreement was excellent, with occasional differences of only a second or two. For the tabulation of the data, two methods were used. First, all the visits of Jaytee to the window were included, even if he was there for reasons that seemed to be unconnected with his anticipatory behavior, for example if he was simply sleeping in the sun, barking at passing cats, or watching people unloading cars. In this way any selective use of data was avoided, although the data were noisy because they included irrelevant visits to the window that had nothing to do with P.S.'s returns. Second, these visits to the window that seemed to have nothing to do with Jaytee's anticipatory behavior were excluded. This set of data was cleaner but more dependent on subjective assessments. However, since these assessments were done blind they should not have involved any systematic bias. Statistical Analysis We used two main methods of analyzing the data, both of which were preplanned. The first provided a simple way of averaging and comparing different experiments. For each experiment, the percentage of the time that Jaytee spent by the window was calculated for three periods: The first 10 minutes after P.S. got into her car and started traveling homeward (the return period). In the case of experiments with randomly selected return times, this return period was deemed to begin at the time P.S. received the beep signaling that she should set off. All homeward journeys lasted at least 13 minutes. Thus Jaytee's reactions in the last 3 or more minutes of P.S.'s journey were omitted from the analysis in case he could have been responding to the sounds of her car approaching. In fact most journey times were more than 15 minutes long, so more than five minutes of Jaytee's behavior were omitted. In cases where the jour- Experiments With a Return-Anticipating Dog 239 ney time lasted 23 minutes or more, the percentage of time for the combined first and second 10-minute periods of the return journey was also calculated, and a separate statistical analysis was carried out for comparison with the usual method involving only 10-minute return periods. 2. The 10-minute period prior to the return period (the prereturn period). 3. The time when P.S. was absent prior to the prereturn period (the main period). Because the experiments varied in length, the length of the main period ranged between 50 and 200 minutes. The percentage of the time that Jaytee spent by the window in these three periods was analyzed statistically by a repeated-measures analysis of variance (ANOVA), and comparisons of pairs of periods were made using the pairedsample t test. The second method of analyzing the data also involved 10-minute return periods, but the main period was also divided up into 10-minute intervals, defined in relation to the time at which P.S. was beeped to come home. The total number of seconds that Jaytee spent by the window in each of these 10-minute periods was then plotted on graphs. In cases where P.S.'s return journey lasted 23 minutes or more, data for two 10-minute return periods are shown on the graphs, representing the first 20 minutes of her homeward journey. A statistical analysis of the time-course data (including all visits to the window) was carried out for us by Dr. Dean Radin using a randomized permutation analysis (RPA) (Good, 1994; Hjorth, 1994). For each data set, he calculated the correlation between time at the window versus the 10-minute segment number of the original data for all experiments (as plotted in the graphs in Figure 4). These correlations showed strong positive trends. The RPA calculations made the assumption that under the null hypothesis, Jaytee should have spent about the same amount of time at the window in each of the 10-minute periods. The z scores were formed as z = (original correlation - average permutated correlation)/(standard deviation of perrnutated correlations), based on 500 random permutations. The RPA tests converged very rapidly; typically only about 100 random permutations were needed, so the estimated z scores with 500 permutations were quite accurate. Results Experiments With Randomly Selected Return Times The overall results summarized in Figure 1 show that Jaytee was at the window far more when P.S. was on her way home than during the main period of her absence. When all Jaytee's visits to the window were included in the analysis (Figure IA), he was at the window for an average of 55% of the time during the first 10 minutes of P.S.'s return journey, as opposed to 4% of the time during the main period of P.S.'s absence. During the 10-minute prereturn period he was at the window 23% of the time. These differences were highly significant statistically (repeated-measures ANOVA, F value [df 2, 221 = 20.46; 240 R. Sheldrake & P. Smart main period. pre-return. return. Fig. 1. The average percentage of time spent at the window by Jaytee during the main period of P.S.'s absence (main period) during the 10 minutes prior to her setting off to come home (prereturn), and during the first 10 minutes of her homeward journey (return). Standard errors are indicated by bars. (A) Data for all visits to the window. (B) Data excluding irrelevant visits. p c .0001; paired-sample t test comparing main period with return period p = .0001). When Jaytee's irrelevant visits to the window were excluded from the analysis, the general pattern was very similar (Figure lB), but the percentage of time at the window was of course somewhat lower. In the main period Jaytee spent 0.5% of the time by the window, in the prereturn period 18%, and in the return period 54%. The significance of these differences was higher than when all Jaytee's visits were included (repeated-measures ANOVA, F value [df 2,221 24.36; p = 3 x lop6). In six out of the 12 experiments, P.S.'s return journeys took more than 23 minutes and hence included two 10-minute periods rather than just one. In the analysis shown in Figures 1A and B, only the first 10-minute return period was included. When both 10-minute return periods from these experiments were included in the analysis, the average percentage of time at the window during the return period increased from 55% to 61% when all visits to the window Experiments With a Return-Anticipating Dog 241 were included, and from 54% to 59% when irrelevant visits were excluded. The statistical significance of the differences was even higher than before [repeated-measures ANOVA, F values (df 2,22) 25.43 and 29.03 respectively]. The increased percentage of time the Jaytee spent at the window during the 10-minute prereturn period was statistically significant (paired-sample t test comparing main period with prereturn period for the data included all visits to the window, p = .04). The difference between the prereturn and return periods was very significant (p = .0009). However, Jaytee did not visit the window in the prereturn period in all experiments, but only in seven out of 12. The detailed time courses for all 12 beep experiments are shown in Figure 2. The graphs show the duration of all Jaytee's visits to the window in each 10minute period, both with and without the exclusion of irrelevant visits. In one of these experiments, Jaytee did not go to the window at all, but in all the others he was at the window for the highest proportion of the time when P.S. was on her way home. In six of these experiments, P.S. was beeped to come home in the first half of the beep window (early beep) and in the other six she was beeped in the second half (late beep). Inspection of the graphs shows that Jaytee responded in the prereturn period in only two of the early-beep experiments, whereas he did so in five of the late-beep experiments (three when irrelevant visits to the window were excluded). Thirty Ordinary Homecomings In order to observe how Jaytee behaved under more or less natural conditions, we made a preplanned series of 30 videotapes of Jaytee at P.S.'s parents' flat while P.S. went out and about. She returned at times of her own choosing, ranging from 11: 13 a.m. to 10:45 p.m., with absences ranging from 85 to 220 minutes. P.S. did not tell her parents when she would be returning, and usually she did not know in advance herself. The overall results are shown in Figure 3A. The general pattern is clear. On average, Jaytee was at the window for the highest proportion of the time (65%) in the return period, when P.S. was on her way home. He was at the window 31 % of the time in the 10-minute prereturn period, and only 1 1 % of the time during the main period of her absence. These differences were highly significant statistically ( p < .0001). The paired-sample t test (two-tailed) showed that the difference between the main period and return period was significant a t p < .0001, between the prereturn and return period at p = .008, and between the main period and prereturn period at p = .0009. A number of interesting details are hidden by this averaging process. First of all, although on 24 occasions Jaytee spent more time at the window when P.S. was on her way home, on six occasions he did not. On five (all in the evening) he did not go to the window at all during the first 10 minutes of her homeward journey. On the sixth (in the morning) he did so for only 10 seconds. On some of these occasions he was unusually inactive and may have been exhausted 242 R. Sheldrake & P. Smart EARLY BEEPS 1 ' " 400 300 4 g 200 500 19/11/96 600 7 - 25/3/97 500 -100 0 r l n l r ~ ~ ~ 2 4 6 8 10 12 14 16 18 20 22 period s l ~ u 100 0 s ~ a - "' .................... .. . .,......... ... . , ,, ,, ,,,, , , ,,, , . -100' ~ 8 0 f---------, r~ r '8 1 ~ . 1q = 'I l z I m l n l l l 2 4 6 8 10 12 14 16 18 20 22 period -50 0 0 2 4 6 8 10 12 14 16 18 20 22 period 0 2 4 6 8 10 12 14 16 18 20 22 period 600: 500 400 11/2/97 450 1 8/10/97 L - , 300 -2 8 200 % 100 0 1 0 0 r l ~ l ~ l ~ - 0 2 4 6 8 10 12 14 16 18 20 22 period l q l ~ l ..."................... -.......-.."..* 1 8 350 300 ,250 -E 200 150100 50 , u ) I H K H M O 0 1 ~ -50 I Ir 0 2 - - ~ * ~ s ~ r ~ ~ r - ~ 8 ~ 9 ~ 4 6 8 10 12 14 16 18 20 22 period Fig. 2. The time courses from all 12 experiments in which P.S. came home at randomly selected times in response to being beeped. The ordinate shows the total number of seconds that Jaytee spent at the window in each 10-minute period, the abscissa the series of 10-minute periods defined in relation to the time at which P.S. was beeped to come home. Data for all Jaytee's visits to the window, including irrelevant visits, are indicated by circles, and data from which irrelevant visits have been excluded are indicated by squares. The beep Experiments With a Return-Anticipating Dog 243 LATE BEEPS 0 2 ' 0 ° : 7/5/97 "coo 8 8 in - 1 4 6 8 period . ~ 8 ~ ~ z 10 12 14 16 18 20 22 0 2 4 6 8 10 12 14 16 18 20 22 period O4 6 - 29/8/97 ZOO loo: 0 - <\ o& 8 ~ $ 8 i ................................................................. ~ r ~ " , 2 --000a>Of>i)f>i)-0. 0 -2-4-6 ~ -100 0 2 4 6 8 10 12 14 16 18 20 22 period -8 -10 0 -- I ~ ~ - I ~~ I r , ~ I -- I ~ ' I ~ I ' ~ ~ ~ 2 4 6 8 10 12 14 16 18 20 22 period 600 500 0 - 1/7/97 400 v, C 300 8 8 200 100 0 -100 --w t - - , - - r ~ ~ 8 ~ r ~ ~ . ~ 0 2 4 6 8 10 12 14 16 18 20 22 period 500- 21/9/97 450 400 350 300 4 250 200 Z 150100 50 ............. . 0 -. I * I . I . I , -5 0 0 2 4 6 8 10 12 14 16 18 20 22 - 8 - 8 ~ - ~ . ~ - L ~ . ~ period Fig. 2 (continued). window is indicated by a line with two arrowheads, and this represents the period during which P.S. could have received the signal to come home. Experiments with beeps in the first half of the beep window (early beeps) are on the left, and those with beeps in the second half of the beep window (late beeps) are on the right. The points for the 10-minute periods immediately following the beep during which P.S. was returning are indicated by filled circles or squares. il 244 R. Sheldrake & P. Smart .- ' g60 50 30 20 10 0 daytime evening i40 main period pre-return return main period pre-return return 100 90 80 C 120 100 D first 10 second lo third 10 , 70 2 60 :: 30 20 10 0 main period pre-return return 0 normal noisy g, 2 80 a 60 40 20 0 main period pre-return return f a 70 :: main period pre-return return long medium short 40 30 20 10 0 Fig. 3. Percentage of time spent at the window by Jaytee during the main period of P.S.'s absence, during the 10 minutes prior to her setting off to come home (prereturn), and during the first 10 minutes of her homeward journey (return). Standard errors are indicated by bars. (A) Averages from 30 ordinary homecomings. (B) Comparison of experiments in the daytime (7) and in the evening (23). (C) Comparison of normal experiments (23) and noisy experiments (7) in which Jaytee was at the window for more than 15% of the time during the main period of P.S.'s absence. (D) Comparison of the first, second, and third groups of 10 experiments. (E) Comparison of long (13), medium (9), and short (8) experiments. after long walks or sick. But irrespective of the reasons for his unresponsiveness, the fact is that he did not show his usual signs of anticipation on six out of 30 occasions. Second, in the daytime Jaytee was generally more active and alert than in the evening, and on average he was at the window more (Figure 3B). There was more activity outside for him to watch, and on sunny days he tended to snooze by the window in the sunlight. Third, the effect of data noise on the pattern of Jaytee's response can be examined directly by comparing noisy experiments with normal experiments (Figure 3C). Noisy experiments were defined as ones in which Jaytee spent more than 15% of the time at the window in the main period. By this criterion Experiments With a Return-Anticipating Dog 245 seven out of the 30 experiments were noisy. Most noisy experiments occurred in the daytime when there was much activity outdoors that Jaytee went to the window to watch. Also, on sunny days he tended to lie down by the window in the sun and go to sleep. Nevertheless, in both normal and noisy experiments Jaytee was at the window least in the main period, more in the prereturn period and most when P.S. was actually returning. These differences were highly significant for both normal and noisy experiments analyzed separately ( p = .0004 and p = .0001, respectively). Fourth, the question of whether Jaytee's pattern of response changed with time can be examined by comparing the average of the first 10 experiments (from May to September 1995) with the second (from September 1995 to January 1996) and third batches of 10 experiments (from January to July 1996). The pattern was similar in all three groups (Figure 3D). Finally, the length of time that P.S. was away from home varied considerably. Did Jaytee behave in a similar way when she returned after short absences and after longer ones? To explore this question, we have divided the data up into three groups: long, medium, and short absences, defined respectively as 180 minutes or more, 110-170 minutes, and 80-100 minutes. The general pattern in all three groups was similar, but in the short absences the experiments were noisier, and Jaytee showed more anticipation in the prereturn period (Figure 3E). Since Jaytee was at the window most in the final period, when P.S. was on the way home, could it be that Jaytee simply went to the window more and more when P.S. was out? If he did so, he would automatically be at the window most in the final period whatever the length of the experiment, and more in the penultimate period that in the previous periods. The going to the window more and more hypothesis can be tested by looking in more detail at the average time courses of long, medium, and short experiments in Figure 4. This figure shows data from all the experiments, as well as from the normal experiments after the exclusion of the minority of noisy experiments, which tended to obscure the usual pattern. The data in Figure 4 show that Jaytee's waiting at the window occurred soonest in the short experiments, later in the medium experiments and latest in the long experiments. In other words, Jaytee's behavior was more closely related to P.S.'s impending return than to the amount of time that had elapsed since she went out. If Jaytee had simply gone to the window more and more as time went on, there should have been little or no difference between the time he spent there in the long, medium, and short experiments in any given period. This can be tested statistically. (In the following analyses, all the data were included, with no exclusions of noisy experiments.) When P.S. was returning in the short experiments in period 8, Jaytee was at the window a significantly higher proportion of the time than in period 8 of the medium and long duration experiments (by a factorial analysis of variance, p = .004). Likewise, Jaytee spent a significantly higher proportion of the time at the window when P.S. was on the way home in the medium experiments in ' 0 ° 350 1 LONG U all -0- normal - - all 3 --D- normal ~ . % X b % period S S 8 E !. .P 8 ~ o T + - all 3 -0- normal -5 0 , , , , , , . . period 8 Fig. 4. The time courses of Jaytee's visits to the window during P.S.'s long, medium, and short absences. The horizontal axis shows the series of 10-minute periods (pl, p2, etc.). The vertical axis shows the average number of seconds that Jaytee spent at the window in each 10-minute period. Data for all 30 experiments are shown, as well as data for normal experiments after the exclusion of the seven noisy experiments. The last period shown on the graph represents the first 10 minutes of P.S.'s return journey (ret), the point for this is indicated by a filled circle or square. The bars show standard errors. Experiments With a Return-Anticipating Dog 247 period 11 than in period 11 of the long absences, when she would not be returning for more than another hour ( p = .003). In a randomized permutation analysis (RPA), including all experiments, the observed time courses were tested against the null hypothesis that Jaytee should have spent about the same amount of time at the window in each of the 10-minute periods. The probabilities that the observed pattern of data differed from the null hypothesis by chance were Long experiments Medium experiments Short experiments Combined p < .0008 p < .01 p < .008 p < .000003 Jaytee S Behavior When PS. Was Not Returning In order to study Jaytee's behavior when P.S. was not coming home, we filmed him at P.S. 's parents' flat on 10 evenings when P.S. was either spending the night away from home or coming home at least one hour after the filming period had terminated. Figure 5 shows the average time he spent at the window in the series of 10-minute periods between 6:30 and 10:OO p.m. In these control observations, Jaytee made a number of visits to the window for a variety of reasons, as usual, but he did not go to the window more and more as the evening went on. Observations on Jaytee at PS. 's Sister's House P.S. sometimes left Jaytee at her sister's house, and here too he usually went to the window when she was coming home. P.S. did not tell her sister when she would be returning, but her sister usually knew when she was on her way because of Jaytee's behavior. period Fig. 5. Time spent by Jaytee by the window on evenings when P.S. was not coming home. The first of the 30 10-minute periods was from 6:30 to 6:40 p.m., the last from 9:50 to 10:OO p.m. The figures shown are averages from 10 evenings. The bars show standard errors. 248 R. Sheldrake & P. Smart In this house, in order to look out of the window Jaytee had to balance himself on the back of a sofa. Unlike the situation in P.S.'s parents flat and in her own flat, Jaytee could not wait by the window comfortably and rarely stayed for long. Nevertheless, in a series of five videotaped experiments, his general pattern of response (Figure 6A) was similar to that in P.S.'s parents' flat (Figure 3), although the percentage of time spent at the window was lower, the variability was greater and differences were not statistically significant. Observations on Jaytee Left on His Own We carried out a preplanned series of 50 videotaped experiments in which Jaytee was left by himself in P.S.'s own flat while she went out. The overall pattern (Figure 6B) was similar to that in P.S.'s parents' flat (Figure 3) and her sister's house (Figure 6A). The differences were significant statistically (repeated-measures ANOVA, p < .01; paired-sample t test comparing the main period with return period, p < .005).But the average proportion of the time at the window was lower than in P.S.'s parents' flat. A closer analysis of the data revealed that Jaytee showed two different patterns of response. In most of the tests (35 out of 50) Jaytee did not go to the window when P.S. was on her way home. In fact he made few or no visits to the main period pre-return return main period pre-return return Fig. 6. Percentage of time spent by the window by Jaytee during the main period, prereturn, and return periods. The bars show standard errors. (A) In P.S.'s sister's house (average of five experiments). (B) Alone in P.S.'s flat (average of 50 experiments). Experiments With a Return-Anticipating Dog 249 window during the entire time she was absent. One reason may be that the view from the window was largely obscured by a bush, so there was not much scope for watching activities outside, although it was still possible to see the road on which P.S. approached in her car. By contrast, in 15 out of 50 experiments (30%), Jaytee behaved much as he did at P.S.'s parent's flat and showed his usual anticipatory waiting while P.S. was preparing to come home and while she was on her way. An Independent Replication During the course of our research with Jaytee, at our invitation Wiseman, Smith, and Milton carried out four experiments with Jaytee, three at P.S.'s parents' flat and one at her sister's house. During these experiments, Wiseman filmed Jaytee while Smith accompanied P.S. and returned with her at randomly selected times in cars unfamiliar to Jaytee (Wiseman, Smith, and Milton, 1998). In all three experiments at P.S.'s parents' flat, the pattern of response was very similar to the pattern we observed, with Jaytee at the window most when P.S. was returning. Using the same definition of the main, prereturn, and return periods used in Figure 1, the average proportion of the time that Jaytee spent at the window was 4% in the main period, 48% in the prereturn period, and 78% in the return period. The differences between the periods were significant (by repeated-measures ANOVA, p = .02; comparison of the main period with return period by the paired-sample t test, p = .03). When the time courses were plotted following the same method used in our Figure 2, they showed a very similar pattern (Figure 7). Wiseman, Smith, and Milton recorded Jaytee's behavior only during the experimental period during which P.S. could have been asked to go home, and have no data on his behavior during the preceding period, up to 90 minutes long, from the time that P.S. left home until the beginning of the experimental period. This is the main difference between the graphs from Wiseman, Smith, and Milton's experiments and our own. In Wiseman, Smith, and Milton's experiment at P.S.'s sister's house, the first time Jaytee went to look out of the window for no apparent reason coincided with P.S. setting off to come home. In spite of these striking effects, Wiseman, Smith, and Milton (1998, 2000) portrayed their results as a refutation of Jaytee's ability to anticipate P.S.'s returns. They arrived at this conclusion by the use of narrow and arbitrary criteria for Jaytee's signal, based on his going to the window for no apparent external reason for a brief period (less than a minute in one experiment, and for at least two minutes in the others). They disregarded the rest of their own data and did not plot graphs. Unfortunately Wiseman, Smith, and Milton based their criteria not on the waiting behavior of Jaytee that we had already observed and documented on more than 100 occasions before they carried out their tests (Sheldrake and R. Sheldrake & P. Smart 0 2 4 6 8 10 period 12 14 16 18 350- 13/6/95 200 - b -50 ~ I I ~ I ~ ~ ~ ~ ~ ~ 0 2 4 6 8 10 period 12 14 16 18 -100 0 2 4 6 8 10 period 12 14 16 18 Fig. 7. The time courses from the three experiments conducted by Wiseman and Smith with Jaytee at P.S.'s parents' flat. The data are taken from Wiseman, Smith, and Milton (1998); the graphs are plotted in the same way as those in Figure 2 and show the total amount of time that the dog spent at the window in successive 10-minute periods, defined in relation to the randomly selected time at which P.S. was told to return home. The final point on each graph, indicated by a filled circle, represents the first 10 minutes of P.S.'s return journey. Experiments With a Return-Anticipating Dog 251 Smart, 1998), but rather on a claim made by the media about Jaytee's behavior. They showed, unsurprisingly, that statements on popular television shows are sometimes oversimplified. Ironically, the way their own skeptical conclusions were publicized in the media provided several striking examples of misleading claims (Sheldrake, 1999b, 2000). Discussion Nomal Explanations of Jaytee's Behavior The data presented in this paper imply that Jaytee's waiting by the window when his owner is coming home cannot be explained in terms of any of the following hypotheses: 1. Routine. Jaytee's anticipatory behavior when P.S. was coming home occurred at various times in the morning, afternoon, and evening and did not depend on a routine time of return. This was apparent in the series of 30 ordinary homecomings (Figures 3 and 4) as well as in our experiments with randomly selected return times (Figures 1 and 2; see also Sheldrake and Smart, 1998). The data from the experiments of Wiseman, Smith, and Milton (1998) with randomly selected return times replicate and confirm our own findings (Figure 7). Moreover, in control observations when P.S. was not coming home Jaytee did not start waiting at a particular time (Figure 5). 2. Hearing a familiar vehicle. In many experiments, Jaytee's anticipatory behavior was already apparent in the prereturn periods (Figures 2 , 3 , 4 , and 6) before P.S. had actually set off in a vehicle, and hence before he could have heard any characteristic sounds. When she was actually traveling home, Jaytee was waiting at the window when the vehicle was at least 7 km away, and in some cases more than 25 km. Although dogs can hear higher pitches than human beings, their general sensitivity to noise levels is similar to that of people (Munro, Paul, and Cox, 1997; Shiu, Munro, and Cox, 1997). It is not possible that Jaytee could have heard the sounds of familiar cars at such distances against all the background noises of Greater Manchester, and in a manner independent of the direction of the wind. Moreover, Jaytee also waited for P.S. in a similar way when she was traveling in taxis or other unfamiliar vehicles (Sheldrake, 1999a; Sheldrake and Smart, 1998), an effect replicated by Wiseman, Smith, and Milton (Figure 7). 3. Picking up clues from people at home. P.S. did not tell her parents or her sister when she would be coming home, and often did not know in advance herself. But perhaps in some of P.S.'s ordinary homecomings, her parents or her sister might have guessed approximately when she would return and consciously or unconsciously communicated their expectation to Jaytee. But this possibility cannot account for Jaytee's behavior in the trials with randomly se-lected return times (Figures 1,2, and 7) nor when he was alone (Figure 6B). 4. Selective memory or selective reporting of data. The video recordings permitted all Jaytee's visits to the window to be recorded, and the data presented in this paper include all the visits he made, even when these were obviously 252 R. Sheldrake & P. Smart related to events going on outside, such as cats passing the window, or when he was sleeping by the window in the sunlight. The videotapes were analyzed blind by people who did not know the details of the experiments. Hence there was no scope for selective memory or selective reporting of data. The data from the experiments conducted with Jaytee by Wiseman, Smith, and Milton (1998) also show the same pattern of behavior by Jaytee as our own experiments (Figure 7). 5. Jaytee going to the window more and more the longer his owner was absent. The data in Figure 4 and the statistical analysis described above show that Jaytee's visits to the window were not explicable in terms of his going there more and more the longer P.S. had been absent. Nor did he go to the window more and more as time went on in the control experiments (Figure 5). His waiting by the window was related to P.S.'s returns, rather than to the length of time she had been away from home. The Possibility of Telepathy Jaytee seemed to be detecting P.S.'s intention to come home in a way that could not be explained in terms of any of the normal hypotheses considered above. Perhaps he was responding to her intentions or thoughts telepathically. The hypothesis of telepathy would not only agree with Jaytee's waiting behavior when P.S. was actually on her way home, but it could help to explain why Jaytee began to spend more time at the window before she set off. In reallife situations when P.S. returned home at nonroutine times of her own choosing, Jaytee's anticipations regularly began in the prereturn period, before she started driving home (Figures 3,4, and 6; see also Sheldrake and Smart, 1998). This pattern of behavior is in good agreement with the telepathic hypothesis because prior to getting into a car and driving, or being driven, P.S. was forming the intention to go home and preparing to do so. If Jaytee was responding telepathically to her intention to return, he would be expected to show this anticipation before she actually got into the car. But Jaytee also showed signs of anticipation in the experiments when P.S. returned at randomly selected times, before she received the signal to go home (Figures 1 and 2). How could he have anticipated when P.S. was going to be beeped? It is perhaps conceivable that Jaytee was telepathically picking up R.S.'s intention to beep P.S. from over 300 km away, but we do not take this possibility very seriously. On one occasion (on July 1, 1997) the beeping was done not by R.S. but by someone neither P.S. nor Jaytee had met, and Jaytee still responded in advance (Figure 2). It is also perhaps conceivable that Jaytee had a precognition of when P.S. would be beeped. But this would involve introducing another paranormal hypothesis in addition to the telepathic hypothesis. It is more economical to consider a possible explanation in terms of telepathy from P.S. In all the experiments with randomly selected return times, P.S. knew that she would be beeped to come home within a particular time period. Ideally, her Experiments With a Return-Anticipating Dog 25 3 mind would have been entirely engaged with other concerns until the beep came. But unavoidably she was sometimes thinking about the signal to go home before it came, especially if it came toward the end of the period of time in which she knew she would be beeped. Jaytee might have picked up these anticipatory thoughts, just as he seemed to respond to a fully formed intention to go home. If Jaytee was indeed responding to P.S.'s expectation that she would soon be receiving the signal to return, then this anticipatory effect would be expected to show up more when the beep came toward the end of the period in which she knew she would be beeped than at the beginning. In four out of six of the trials in which P.S. was beeped in the first half of the beep period (early beep), Jaytee did not show any anticipation prior to P.S. setting off (Figure 2). By contrast, there were signs of anticipation in all but one of the late-beep trials. The exception was a trial in which Jaytee did not go to the window at all throughout the entire experiment. Thus Jaytee's anticipation of the beep signaling P.S.'s return may have been related to her own anticipation of the beep, which tended to be greater the later the signal came. A similar anticipation of P.S.'s setting off occurred in the experiments conducted by Richard Wiseman and Matthew Smith (Figure 7). Here again, Jaytee's early response may well have taken place in response to P.S.'s anticipation. While she was with Smith waiting for him to tell her when to return, she found it impossible not to think about going home. Smith himself knew when they were going to set off because the randomly determined time had been set in advance (Wiseman, Smith, and Milton, 1998). He could well have communicated his anticipation to P.S. unconsciously, for example through an increasing tenseness as the predetermined time approached. Moreover, in all three experiments, the randomly selected return time was in the second half of the experimental period, corresponding to the late beeps in our own experiments (Figure 2B). This increasing anticipation by P.S. that she would soon be going home as the experimental period progressed was an unavoidable feature of the experimental design adopted both by ourselves and by Wiseman, Smith, and Milton. Why D d Jaytee Sometimes Not React to PS.'s Returns? i In all our series of experiments with Jaytee, on some occasions he did not show his usual anticipatory behavior. In our preliminary series of 100 observations, he failed to do so on 15 occasions. On some of these occasions he was tired after long walks; on some he was sick; on others he was distracted by a bitch in heat in a neighboring apartment (Sheldrake and Smart, 1998). But in a few cases there was no obvious reason for his failure to react. In our series of 12 experiments with randomly selected return times, he did not go to the window at all in one experiment (Figure 2). In the series of 30 ordinary homecomings, he did not show his anticipatory behavior in six experiments. When Jaytee was left in P.S.'s flat on his own, his lack of anticipatory behav- 254 R. Sheldrake & P. Smart ior was usual rather than exceptional. On most occasions he did not go to wait for her at the window or indeed visit the window at all. Nevertheless on 15 out of 50 occasions he showed his usual pattern of anticipation, waiting at the window when P.S. was returning. Thus he seemed capable of anticipating P.S.'s returns when he was on his own, but did not usually do so. Why not? Our guess is that it was a matter of motivation. His waiting at the window while P.S. was on her homeward journey may have been more for the sake of communicating his anticipation to members of P.S.'s family, as if to tell them she was on her way. When there was no one to tell, he was less motivated to wait at the window. Nevertheless, he sometimes did it anyway. The difference in his behavior in P.S.'s own flat and in her parents' was a matter of degree. In both places, he sometimes waited by the window when P.S. was returning, and sometimes failed to wait there. In P.S.'s parents' flat the ratio of occasions on which he waited to those he did not was around 80:20, whereas when he was alone in P.S.'s own flat it was 30:70. Evolutionary Implications The hypothesis that some dogs, such as Jaytee, can anticipate their owners' arrivals telepathically obviously needs to be tested further. We have already obtained comparable results with several other dogs. Similar anticipatory behavior is said by many animal owners to occur with other domesticated species, especially cats, parrots, and horses (Brown and Sheldrake, 1998; Sheldrake, 1999a; Sheldrake, Lawlor, and Turney, 1998; Sheldrake and Smart, 1997), and there is a need for experimental research on anticipatory behavior by animals of these species. It would also be worth investigating whether animals in the wild seem to know when members of their group are coming home: for example, do wolf cubs waiting at their den show signs of anticipation before the return of adults with food? Although parapsychologists and psychical researchers have conducted much research on person-to-person telepathy (for a review, see Radin, 1997), there has very little previous research on person-to-animal or animal-to-anima1 telepathy (Sheldrake, 1999a). If it turns out that telepathic communication does indeed occur among nonhuman animals, then this would imply a biological and evolutionary origin for person-to-person telepathy and would enable this paranormal phenomenon to seem more normal, at least in the sense that it is biological and has an evolutionary history. Acknowledgements We are grateful to Muriel Smart, the late William Smart, and Cathie MacKenzie for their invaluable cooperation in this research; to Amanda Jacks and Jane Turney for their analysis of the videotapes; and to Dean Radin for carrying out the randomized permutation analysis. We thank the Lifebridge Foundation and the Institute of Noetic Sciences for financial support. Experiments With a Return-Anticipating Dog 255 References Brown, D. J., & Sheldrake, R. (1998). Perceptive pets: A survey in north-west California. Journal of the Society for Psychical Research, 62, 396-406. Good, P. (1994). Permutation tests: A practical guide to resampling methods for testing hypotheses. New York: Springer Verlag. Hjorth, J. S. U. (1994). Computer intensive statistical methods. New York: Chapman and Hall. Matthews, R. (1 994, April 24). Animal magic or mysterious sixth sense? Sunday Telegraph, p. 17. Munro, K. J., Paul, B., & Cox, C. L. (1997). Normative auditory brainstem response data for bone conduction in the dog. Joumal of Small Animal Practice, 38, 353-356. Radin, D. (1997). The conscious universe: The scientific truth of psychic phenomena. San Francisco: Harper. Sheldrake, R. (1994). Seven experiments that could change the world: A do-it-yourselfguide to revolutionary science. London: Fourth Estate. Sheldrake, R. (1999a). Dogs that know when their owners are coming home. New York: Crown. Sheldrake, R. (1999b). Commentary on a paper by Wiseman, Smith and Milton on the "Psychic pet" phenomenon. Journal of the Society for Psychical Research, 63, 306-3 11. Sheldrake, R. (2000). The 'psychic pet' phenomenon. Journal of the Society for Psychical Research, 64, 126-128. Sheldrake, R., Lawlor, C., & Turney, J. (1998). Perceptive pets: A survey in London. Biology Forum, 91, 57-74. Sheldrake, R., & Smart, P. (1 997). Psychic pets: A survey in north-west England. Journal of the Society for Psychical Research, 61, 353-364. Sheldrake, R., & Smart, P. (1998). A dog that seems to know when its owner is returning: Preliminary investigations. Journal of the Society for Psychical Research, 62, 220-232. Shiu, J. N., Munro, K. J., & Cox., C. L. (1997). Normative auditory brainstem response data for hearing threshold and neuro-otiological diagnosis in the dog. Journal of Small Animal Practice, 38, 103-107. Snedecor, G. W., & Cochran, W. G. (1967). Statistical methods. Ames, IA: Iowa State University Press. Wiseman, R., Smith, M., & Milton, J. (1998). Can animals detect when their owners are returning home? An experimental test of the 'psychic pet' phenomenon. British Joumal of Psychology, 89,453-462. Wiseman, R., Smith, M., & Milton, J. (2000). The 'psychic pet' phenomenon: A reply to Rupert Sheldrake. Journal of the Society for Psychical Research, 64, 46-49. Journal of Scientific Exploration, Vol. 14, No. 2, pp. 257-274,2000 0892-33 10/00 O 2000 Society for Scientific Exploration ESSAY What Can Elementary Particles Tell Us About the World in Which We Live?' Department of Physics Texas A & M University, College Station, TX 77843-4242 e-mail: bryan @physics.tamu.edu Abstract-What kind of space-time do we live in? Does it extend beyond the four dimensions of ordinary space and time? In physicists' efforts to explain the origin of elementary Dirac particles, namely the 12 kinds of quarks and leptons, they find that they are driven to as many as 7 or 8 additional spacelike dimensions. They generally assume that the extra dimensions are curled up into tiny circles or generalizations thereof, with diameters of the order of meter. These models tend to be quite complex. However, I will argue that it may be much easier to model these quarks and leptons if we assume that the extra dimensions are flat, that is, stretch out to infinity. What keeps quarks and leptons (and us) from drifting off into the higher dimensions may be a local "well" in space (a soliton) generated by the particles' field equations. Furthermore, only four extra dimensions may be needed. If there really are four extra dimensions besides the ordinary four, then why don't we see them? It may be that many people who have had an out-ofbody or near-death experience (NDE) have seen the extra dimensions. For example, the "tunnel" in the NDE may lead to another local universe like our own, only situated in another "well" in the extra dimensions. In the model that I will describe, quarks and leptons, which are accelerated to sufficiently high energies, can escape our local space-time "well" and travel freely in eight dimensions, as our consciousnesses seem to be able to do. Could it be that large configurations of these particles might even constitute spaceships, the UFOs that seem to come out of nowhere? Higher dimensions may also provide avenues for information transfer ascribed to ESP. Keywords: elementary particle physics-extra dimensions-psychic phenomena 1. Introduction: Elementary Dirac Particles and Extra Dimensions Consider the ordinary hydrogen atom, pictured in Figure 1 as Niels Bohr saw it, with the electron orbiting the proton. If you blew up the atom to be as large as the tip of your little finger, then the tip of your little finger, scaled up like- This paper is an extension of a talk given at the annual meeting of the Society of Scientific Exploration, held at the University of Virginia, Charlottesville, VA, May 27-30, 1998. ' Ronald A. Bryan electron a proton Fig. 1. Electron orbiting proton in the Bohr Model of the hydrogen atom. wise, would be as large as Saturn. The electron was the first elementary particle to be discovered (by J. J. Thomson in 1897, in a gas discharge tube, similar in principle to a neon sign). Today we know that there are (at least) six kinds of electrons: the e, v,, p , v,, z, and v,; namely, the electron, the electron neutrino, the muon, the mu neutrino, the tau, and the tau neutrino. These particles are like brothers and sisters, in that they all obey the Dirac equation of quantum mechanics. (The Dirac equation is a kind of theoretical quantum device that generates particle waves from space-time.) This means that these particles can also appear as antiparticles (e.g., the anti-electron is the positron) and all have an intrinsic spinning motion with angular momentum A12 (where A is Planck's constant divided by 2n), which can point either up or down. The e, p, and rare identical (with the same negative electric charge) except that they have different masses. The neutrinos are identical, too, except for their masses, which might be extremely small (we aren't sure) (Particle Data Group, 1998). (Neutrinos have no charge and at typical reactor energies can travel 25 light-years in lead before deflecting.) These six kinds of electrons are called leptons. Leptons are incredibly small. If you blew up any one of them to be as big as the tip of your finger, then the tip of your finger, increased proportionally, would be much bigger than our solar system and stretch out toward to the nearest star, or perhaps even beyond (we don't know) (Particle Data Group, 1998). The other constituent of the hydrogen atom, the proton, is far smaller than it appears in Figure 1. Nevertheless, it contains three quarks, as shown in Figure 2. Elementary Particles the proton Fig. 2. The quarks in the proton. These quarks occupy as little volume as the electron does, although they are much heavier. They were discovered at the Stanford Linear Accelerator Center around 1968 in electron-proton scattering experiments. Other kinds of quarks have been discovered since then, and now we knob of six types. They have the following amusing names: up, down, charm, strange, top, and bottom. They too are like brothers and sisters in that they all obey the Dirac equation of quantum mechanics, exist in particle or antiparticle mode, and have two possible spin directions. They are like cousins to the six kinds of electrons. They differ from these leptons in that they have never been seen alone; they only occur in triplets, like the proton, or in quark-antiquark pairs, like the pi meson. They are well described by a theory known as Quantum Chromodynamics. I have plotted the leptons and quarks versus their rest-mass energies in Figure 3. (For example, the electron's rest-mass energy, as indicated on the vertical axis, is M C = 0.5 MeV = 0.5 Million electron Volts, corresponding to a par~ ticle-mass M = 9 x kg; c = speed of light.) A pair of quarks and a pair of leptons comprise each generation, and there are three generations (Particle Data Group, 1998) (see Figure 3). Because there are so many kinds of leptons and quarks, many physicists tried to model them as bound states of smaller, more elementary particles, variously called rishons, preons, subquarks, etc. The idea was that just as the hydrogen atom has (infinitely) many excited states, perhaps the quarks and leptons are just different configurations of two or three smaller particles. However, these subparticle models turned out to be about as complicated as the quarks and leptons that they were supposed to predict, and in fact, predicted far too many other kinds of particles that have never been seen. Ronald A. Bryan M C ~ lSf gen 2nd 3rd EW gen gen bosons 10,000 MeV 1000 MeV 100 MeV 10 MeV 1 MeV 0.1 MeV 0.01 MeV 0.001 MeV - - I i 1 0.00001 MeV 1". photon Fig. 3. The three generations of quarks and leptons plotted versus rest-mass energy; weakisospin pairs are joined. Quarks come in three "colors" (not shown), and leptons in one. The electroweak bosons, W', and W-, and (massless) photon are also plotted. The masses of the neutrinos are not known, but their experimental upper limits are indicated as horizontal bars. For comparison, the rest-mass energy of the proton is 938 MeV. These days, most physicists trying to model quarks and leptons turn to higher dimensions. They basically generalize a gravitational theory proposed by Kaluza and Klein in the 1920s, where these authors added another spacelike dimension to the three of space and one of time in an attempt to unify electromagnetism with gravity (see, e.g., Lee, 1984). This fifth dimension did not stretch out to infinity like ordinary dimensions but was taken to be a circle. In fact, an incredibly small circle. The radius had to be the order of cm for the theory to predict the correct charge for the electron (Miller, 1980). To help visualize these higher dimensions, let's consider some projections. For definiteness, first consider ordinary three-dimensional space. I'll portray it in Figure 4 as a large block, but I will think of space as stretching out to infinity in each direction. (Actually, this is a model of space, i.e., Cartesian space; one assumes that space can be divided into little cubes, each identified by three coordinates.) Elementary Particles 261 Fig. 4. Three-dimensional Cartesian space. Now let's eliminate one spacelike dimension, say the (vertical) y-dimension, leaving just the horizontal x-z plane, as in Figure 5. Fig. 5. Two-dimensional Cartesian space. Let's eliminate the x dimension too. This leaves just one dimension, z, still understood to stretch out to infinity in both directions (Figure 6). Fig. 6. One-dimensional Cartesian space. Now let's add an extra dimension, a synthetic dimension not present in ordinary space-time, call it 2 (see Figure 7). (I will embellish all extra dimensions, Fig. 7. Two-dirnenslvnal Cartesian space consisting of one ordinary dimension (z) and one extra dimension (x). 262 Ronald A. Bryan such as 2, with a tilde.) Here, we have recovered a plane, but it is not the usual kind of plane because leptons and quarks, and we ourselves, cannot ordinarily make excursions in the i direction. In Kaluza-Klein theory (Lee, 1984), the extra dimension is mapped onto a tiny circle, as I mentioned above. This is illustrated in Figure 8 for the case of the extra dimension and one ordinary dimension, say z. It is like rolling the i - z plane up like a sheet of paper to form a tube. The extra dimension is shown in Figure 8 projected on a circle on the left and is denoted 2. This theory was only meant to reproduce gravity and electromagnetism. However, since the 1920s, 11 quarks and leptons have been discovered in addition to the electron. Fig. 8. One ordinary dimension and the extra Kaluza-Klein dimension 2 , which appears as a minuscule circle of radius of the order of m. Now, it may be that gravity has no more to do with the mass spectrum of quarks and leptons than it has to do with the energy-level spectrum of the hydrogen atom (in contrast to a basic assumption of Kaluza-Klein and string theory). Nevertheless, the extra dimension can still be useful, as it provides natural space for additional elementary particles. Indeed, physicists have routinely employed two extra dimensions for decades to describe a feature of elementary particles called weak isospin. To understand this feature, picture an electron spinning about an axis in ordinary space. This spinning motion is a quantum-mechanical version of the spinning motion of, say, a ball, or the earth about an axis joining the north and south poles, as pictured in Figure 9. The electron is very small, of course, if it has any size at all. We confirm that it has this spinning motion (predicted by the Dirac equation) from the way that it interacts with other particles. The direction of the axis about which the electron is spinning can be recorded by two angles, say 8 and 4, or equivalently, by the two coordinates of a small area on the surface of a large sphere (say, of unit radius) where the electron's extended spin axis touches the sphere. The latter is pictured in Figure 10. Now just as the electron has a spinning motion in ordinary space, it also behaves in reactions with other particles as if it had an additional spinning motion in a synthetic space (i.e., in extra dimensions). Physicists call this additional spin weak isospin. In Figure 11, I have sketched the electron's weak isospin, s", directed to a small area on the synthetic sphere (called the weakisospin sphere or the isospin sphere). The surface of the isospin sphere can be considered to be two extra dimensions. In a classical (i.e., non-quantum-mechanical) theory, each electron pointing to a different point on the isospin sphere would correspond to a different particle in ordinary space because it would behave differently in its inter- Elementary Particles 263 S Fig. 9. Earth spinning about its north-south axis. unit sphere L Fig. 10. Unit sphere concentric with electron, and small area on unit sphere where the electron's spin axis touches. Ronald A. Bryan Fig. 11. Electron at center of (unit-radius) weak-isospin sphere; electron's weak-isospin axis, directed at small area on two-dimensional surface of sphere. s", actions with other particles. Thus, the surface's two extra dimensions would apparently turn one particle into an infinite number of different kinds of particles as seen in ordinary space. However, according to the rules of quantum mechanics, the electron may point only in two directions in ordinary space (up and down), and similarly it may point only in two directions in isospin space. Thus, isospin only doubles the number of electron types. The other member of the electron's doublet is the electron neutrino, v,, and its spin is usually taken to point "up." The electron's weak isospin is then taken to point "down." The e - v, doublet is plotted with all of the other doublets in Figure 3, and plotted alone in Figure 12. Let us suppose that we are constructing a theory of elementary particles, and we incorporate weak isospin to double the number of electrons. Fine, but there are six Dirac-particle doublets in Figure 3. How can we account for all six doublets? Well, if we are going to take the two-dimensional surface seriously and say that the electron doublet "lives" on that surface, then I think that it is only reasonable to take the volume within the spherical surface seriously too and say that the electron lives in the entire volume. In this way, I introduce another artificial dimension, the radial dimension, which brings the number of extra dimensions to three. This, as it turns out, gives enough space for an extended Dirac equation (a Dirac equation in 4 + 3 = 7 dimensions) to generate the up and down quarks as well as the electron and its neutrino, that is, enough space to generate the whole first generation. The e - v, and up-down doublets are plotted in Figure 3 and repeated below in Figure 13. Elementary Particles 0.01 MeV OC 1 MeV .O 0.0001 MeV 0.00001 MeV 1 I t 1 1 tve Fig. 12. Plot of e - v, weak-isospin doublet versus rest-mass energy. Upper experimental limit of electron neutrino's mass indicated by horizontal bar. Although I have suggested that the Dirac particle is trapped within the spherical shell as a marble might rattle around inside a tin can, in fact a softer repelling surface is required to enable the Dirac equation to produce the kind of multiplets that experimentalists actually see. This surface allows the particle to penetrate somewhat and turns out to be what is called a harmonic- 0.1 MeV 0.01 MeV 0.001 MeV 0.0001 MeV 0.00001 MeV t I 1 1 1 lve I Fig. 13. Up and down quarks, and electron-neutrino and electron, grouped as weak-isospin doublets versus rest-mass energy, M C ~ . These particles constitute the first generation. 266 Ronald A. Bryan oscillator "well." Mathematically, it is as if the Dirac particle were connected to the center of the isospin sphere with a rubber band. As I will explain in the next section, we can generate all three generations of quarks and leptons with a Dirac equation if we increase the number of extra dimensions from three to four and again restrain the Dirac particle with a harmonic-oscillator force. This is what I consider the particle-physics evidence that we live not in four, but rather in 4 + 4 = 8 dimensions (Bryan, 1998, 1999). It is reasonable to suppose that if four extra dimensions exist within a spherical shell, then these dimensions should exist outside the shell as well, perhaps stretching out to infinity in all extra directions. As far as I know, there is no elementary-particle-physics evidence (yet) for extra dimensions stretching out beyond the shell. However, there may be anecdotal evidence: the widely reported out-of-body experiences (Monroe, 1971), near-death experiences (NDEs) (Moody, 1975), existence of discarnate entities (Roberts, 1972), encounters with UFOs (Mack, 1994), and cases of extrasensory perception (Targ and Puthoff, 1977) suggest long-ranging extra dimensions. I will elaborate on this in the last section of this paper. In the next section, I give in more technical language the evidence that I have found for four extra dimensions. I will also discuss some limitations of the model. Readers not interested in the details can skip to the last section with little loss in continuity. 2. Some Technical Details To reproduce the quarks and leptons, it is not at all necessary to take the radius of the extradimensional sphere to be as small as the Kaluza-Klein radius. A radius of the order of 10-l9m will suffice. If the surface of the sphere were impenetrable, then quantum mechanics (in this case, a Schrodinger equation acting in the three extra dimensions) would permit a particle to be trapped provided that its mass were m, or m, or m, ... This spectrum of possible massstates is sketched in Figure 14 as horizontal lines, drawn inside the walls of the "square well." (The masses' subscripts merely signify that they are distinct.) mass rigid surface of isospin-sphere \\ m3 2" A / rigid surface of isospin-sphere ml O C radial distance in extra dimensions Fig. 14. Quantum-mechanical mass-levels of a particle trapped within a hard spherical shell (a "square well") in three extra dimensions. Elementary Particles 267 If the hard-shell square well is replaced by a harmonic-oscillator well (which means that the Dirac particle can "push" the wall out somewhat), then the trapped particles appear as 1 , 3 , 6 , 1 0 , . multiplets with allowed masses m,, m,, m, ..., as plotted in Figure 15. (Again, the subscripts merely denote distinct masses.) The 1 comprises one particle, the 3 comprises three different kinds of particles, etc. These multiplets exhibit what is known as SU(3) symmetry. The triplet 3 has the symmetry of quarks (quarks come in three "colors") and the singlet 1 has the symmetry of leptons (leptons are "co1or1ess~'). The Schrodinger equation can also factor in the two isospin states that come with each multiplet to create a quark doublet and a lepton doublet. The 3 doublet and the l doublet then have the same quantum numbers as the physical u - d and the e v, doublets, respectively, as in Figure 3 or 13. These two doublets, then, constitute the first generation, or family. mass \A/! A harmonic-oscillator well distance from center of extra dimensions Fig. 15. Particle mass levels predicted by symmetric harmonic-oscillator well in three extra dimensions; SU(3) multiplicities 1 , 3 , 6 indicated. Note, however, that the harmonic-oscillator well also generates 6s, as well as 10s, etc. (not shown). These sextuplets and decuplets would be new kinds of Dirac particles, neither quarks nor leptons. Such particles have not been seen. However, astrophysicists see indirect evidence that at least ten times as much matter exists in the universe than has been observed with our optical and radio telescopes. For instance, stars somewhat beyond the edge of the Milky Way circle our galaxy at surprisingly high speeds, indicating that they are being gravitationally pulled toward the center of the galaxy by much more matter than is visible to our instruments. The 6s, 10s, and so on predicted by my model might be the source of this so-called "dark matter." I have indicated how one generation of quarks and leptons can be generated by a harmonic-oscillator force acting on weak-isospin doublets. However, there are two more families to be accounted for. As I mentioned earlier, these can be generated by increasing the number of extra dimensions from three to four. A symmetric harmonic-oscillator well in four dimensions will yield SU(4) multiplets with multiplicities 1,4,10.... In an SU(4) = SU(3) x U(1) de- 268 Ronald A. Bryan composition, the 1 stays a 1, the 4 breaks up into a 1 plus a 3, the 10 breaks up into a 1 plus a 3 plus a 6, etc. This results in the weight diagrams shown in Figure 16. Quarks and leptons suggested by these quantum numbers are indicated. In Figure 16, the quantum numbers, fi = 0, 1, 2, correspond to increasing mass levels. The chart is divided into two halves, with the weight diagrams on the left corresponding to particles with weak-isospin quantum number = l/z (isospin "up") and the diagrams on the right to particles with weak-isospin vPls = -% (isospin "down"). If the restraining force is that of a pure harmonic oscillator, then the multiplets go on past to fi = 2 infinity. I have calculated these particle multiplets using a quantum-mechanical Dirac equation in eight dimensions with a symmetrical harmonic-oscillator potential acting in the four extra dimensions (Bryan, 1998, 1999). The equation separates into two equations, a standard free-particle Dirac equation operating in ordinary four-dimensional space-time and a second Dirac equation acting in the four extra dimensions. An extremely large mass, M , (not any particle's rest mass), is added to the potential so that the Dirac equation in the extra four dimensions reduces approximately to a type of Schrodinger equation. This is used to predict the particles noted in Figure 16. Now some caveats: The Dirac equation of my model does not quite preserve SU(4) symmetry, or the SU(3) symmetry of the quarks, so it does not agree isospin b b ~ ~ " isospin "down" Fig. 16. SU(4) multiplets generated by a symmetrical harmonic-oscillator well acting in a Schrijdinger equation in four flat extra dimensions. Elementary Particles 269 perfectly with Quantum Chromodynamics. Also, as can be seen in Figure 15, the model predicts masses rising at a constant rate, whereas experimentally, particles' masses seem to increase exponentially, as in Figure 3. However, "Higgs-Yukawa" terms might perhaps be introduced to generate the correct masses, as is done in the Standard Model. [The Standard Model is a gauge field theory, which fits all of the known elementary-particle data with some 19 adjustable constants. It was invented by Sheldon Glashow (1961), Steven Weinberg (1967), and Abdus Salam (1968). For a less technical description of the Standard Model, see Coughlan and Dodd (199 1) and references listed therein.] There are two general classes of particles in nature, the Dirac particles (and their antiparticles) that I have been describing, and bosons. The W', W, 2?, and photon are bosons, plotted along with the Dirac particles in Figure 3. They mediate the forces between the Dirac particles. My model treats Dirac particles as real particles but only takes bosons in account when they interact with Dirac particles. Work is underway to include bosons as real dynamical entities. There has to be a mechanism to keep the particles in place in the extra dimensions, so they do not drift off and disappear from our universe. In most models, localizing poses no problem because the assumed extra dimensions are tiny circles or generalizations of circles of the order of m in diameter, so the particles have hardly any place to go. However, in models with infinitely extended extra dimensions, such as mine, some trapping mechanism has to exist. For the time being, I simply introduce a potential well "by hand" to trap the particles. However, in 1983 Rubakov and Shaposhnikov had already considered the possibility that the extra dimensions might be infinite in extent, and they solved the particle-trapping problem with an elegant mechanism, namely a soliton, generated by the nonlinear field equations of their model (Rubakov and Shaposhnikov, 1983). However, they only considered the case of a single extra dimension. For this case, they trapped a boson in the well sketched in Figure 17. v, Fig. 17. Potential well V in one extra dimension generated by soliton (kink) in 4 ! ~theory; wave ~ function of Dirac particle trapped on the soliton is denoted Po. - 270 Ronald A. Bryan The width of the well is determined by the mass parameter, m. Note that the well does not rise indefinitely, but only as high as 2m2. This particular well traps just one kind of particle, at the level m 2 . Particles whose mass-squared exceeds 2m2 lie in the continuum and can propagate freely in Rubakov and Shaposhnikov's extra dimension 2 . If the Rubakov-Shaposhnikov model could be extended from the one extra dimension to four extra dimensions, then it might provide the entrapment that I have simulated with the harmonic-oscillator well. Furthermore, a RubakovShaposhnikov-type well would provide a physically natural way to limit the number of generations to just the three that have been seen, because the well rises only a finite amount. Particles with energies exceeding the top of the well would travel freely in all eight dimensions and would not be listed in Figure 3. 3. Some Speculations I have outlined a model based on four extra dimensions stretching out to infinity, in addition to the usual four dimensions of space and time (Bryan, 1998, 1999). The quarks and leptons in this picture are nevertheless trapped in a very small region in these extra dimensions, about 10-l9 m in diameter (which also happens to be the upper limit of the size of the electron in ordinary space). Presumably I could modify the model so that the extra dimensions themselves were little circles 10-l9 m in diameter. However, the mathematics is much simpler if the extra dimensions stretch out indefinitely, and so, by Occam's razor, I choose such a model. ("The upper dimensions are just like the lower dimensions.") But if the extra dimensions do stretch out to infinity, then why haven't we seen them? Well, perhaps some of us have. It may be that some of the people who have had an out-of-body experience (Monroe, 1971) or an NDE (Moody, 1975) have seen and perhaps traveled into the extra dimensions. If we take ordinary space-time and reduce it to just one spacelike dimension, z, as in Figure 6, and then take the four extra dimensions of my model and reduce them to two extra dimensions, say i and j , then the particles are trapped in a kind of tube in both the ordinary and the extra dimensions, as illustrated in Figure 18. The particles are free to move back and forth along the z Fig. 18. Eight-dime_nsiona_l space-time reduced to one ordinary dimension ( z ) and two extra dimensions (xand y ); quarks and leptons are s_hown_as quantum-mechanical wave packets, trapped in a harmonic-oscillator well in the X and y dimensions. Elementary Particles 271 axis but are constrained by the walls of the tube (representing the harmonic-0scillator well) from going off in the 2 or y direction. Because we are made of quarks and leptons, we are also trapped in the tube according to this model. In one real dimension, we might be represented by the string of quarks and leptons, as illustrated in Figure 19. Fig. 19. Human body depicted as.a string of quarks and leptons in one ordinary dimension, confined in a tube in two extra dimensions. On the other hand, our consciousnesses may not be trapped in the tube and perhaps are not fundamentally constrained at all (Roberts, 1972), as suggested by Figure 20. [For an interesting article that hints how a consciousness might control a physical body using energies no greater than those of visible photons (of the order of electron volts), see Firsoff (1975).] Fig. 20. Human consciousness sketched as a kind of wave function, not necessarily confined to human body (depicted in one dimension as a string of quarks and leptons). Now, if there is one tube representing our universe, then there is probably another tube somewhere else representing another universe. There might be an uncountable number of tubes, or universes. Perhaps one of the other tubes is very close to our tube. Then a "tunnel" might make it possible for a human consciousness to drift over to the other tube, and so entering, see a whole new universe. This is illustrated in Figure 21. Perhaps upon entering the other universe, the consciousness encounters the "light" and other consciousnesses 272 Ronald A. Bryan Fig. 21. Tunnel facilitating departure of human consciousness from our universe to another universe in higher dimensional space-time. (Moody, 1975). The other universe need not have the same number of dimensions as our universe, and particles there could be pure energy without mass. As I mentioned earlier, if a particle has sufficient energy, then it might escape our universe (the tube) and propagate freely in all of the dimensions, ordinary and extraordinary. For example, if an electron and a positron were each accelerated to sufficiently high energies in a linear accelerator, then upon colliding they might escape the tube. This is illustrated in Figure 22. Perhaps the next generation of accelerators, in particular the LHC proton accelerator, being constructed at CERN in Geneva, Switzerland, will be able to accelerate particles to sufficiently high energies to escape. Fig. 22. Electron (e-) and positron (e') accelerated in linear accelerator to sufficiently high energies to escape our universe and travel in the extra dimensions. It may well be that if there are indeed long extra dimensions, then particles abound in that space. Probably they can bind to one another, just as particles in our four-dimensional space-time can bind to one another. If such particles exist, then perhaps they can be fashioned into UFOs. The UFOs would not be able to actually enter our space, as they would probably blow up upon falling into our "potential well." However, they might be able to approach very closely (perhaps even as close as a millimeter). Associated consciousnesses might Elementary Particles 273 easily penetrate our space then. This might be the basis for reports of alien abduction (Mack, 1994). Finally, information might be able to propagate in ways not possible in ordinary space-time. For example, ordinary electromagnetic signals sent from submarines are quickly absorbed by the sea water, but if another kind of signal could be sent through the wall of our "tube" into the extra dimensions, then the signal might propagate quite freely in these higher dimensions, to finally return to a receiver somewhere else within our tube (Targ and Puthoff, 1977). This is illustrated in Figure 23. sender receiver Fig. 23. Signal leaving sender in ordinary space-time, travelling through higher dimensions and returning to receiver in ordinary space-time. 3.I . Note Added Recently there has been considerable interest in the possibility that the extra dimensions, although not flat or infinite in extent, might still be circles of millimeter size. Only gravity waves would range that far out. All other particles would be confined to distances of the order of m in the extra dimensions. Arkani-Hamed, Dimopoulos, and Dvali (1998) and Antoniadis et al. (1998) have proposed models of this sort. These models have passed a gauntlet of tests but many challenges remain. See, for example, Banks, Dine, and Nelson (1999). Acknowledgments I would like to thank Hal Puthoff and Robert Butts for much encouragement over the years. I would also like to thank Robert Monroe and John Mack for useful conversations, and Sunny Nash for helpful criticism. References Antoniadis, I., Arkani-Hamed, N., Dimopoulos, S., & Dvali, G. (1998). New dimensions at a millimeter to a fermi and superstrings at a TeV. Physics Letters B, 436, 257. Arkani-Hamed, N., Dimopoulos, S., & Dvali, G . (1998). The hierarchy problem and new dimensions at a millimeter. Physics Letters B, 429, 263. Banks, T., Dine, M., & Nelson, A. E. (1999). Constraints on theories with large extra dimensions. Journal of High Energy Physics (electronic journal), 06(1999), 0 14. Bryan, R. A. (1998). Are quarks and leptons dynamically confined in four flat extra dimensions? Nuclear Physics B, 523, 232. 274 Ronald A. Bryan Bryan, R. A. (1999). Are the Dirac particles of the Standard Model dynamically confined states in a higher dimensional flat space? Canadian Journal of Physics, 77, 197 (hep-phl9904218). Coughlan, G. D., & Dodd, J. E. (1991). The ideas of particle physics: An introduction for scientists. Cambridge, UK: Cambridge University Press. Firsoff, V. A. (1975). Life and quantum physics. In Oteri, Laura (Ed.), Quantumphysics andparapsychology (p. 109). New York: Parapsychology Foundation. Glashow, S. L. (1 961). Partial-symmetries of weak interactions. Nuclear Physics, 22, 579. Lee, H. C. (Ed.). (1984). An introduction to Kaluza-Klein theories. Singapore: World Scientific. Mack, J. E. (1994). Abduction: Human encounters with aliens. New York: Scribner's. Miller, J. G. (1980). Kaluza and Klein's five-dimensional relativity. In Marlow, A. R. (Ed.), Quantum theory and gravitation (p. 221). New York: Academic Press. Monroe, R. A. (1971). Journeys out of the body. New York: Doubleday. Moody, R. A., Jr. (1975). Life after life. Atlanta, GA: Mockingbird. Particle Data Group (1998). Review of particle physics. European Physical Journal C, 3, 1. Roberts, Jane (1 972). Seth speaks: The eternal validity of the soul. Englewood Cliffs, NJ: Prentice-Hall. Rubakov, V. A., & Shaposhnikov, M. E. (1983). Do we live inside a domain wall? Physics Letters B, 125, 136. Salam, A. (1968). Weak and electromagnetic interactions. In Svartholm, N. (Ed.), Elementary particle theory: Relativistic groups and analyticity (pp. 367-377). Stockholm: Almqvist & Wiksell. Targ, R., & Puthoff, H. (1977). Mind-Reach: Scientists look atpsychic ability. New York: Dell. Weinberg, S. (1967). A model of leptons. Physical Review Letters, 19, 1264. Journal of Scientific Exploration, Vol. 14, No. 2, pp. 275-279,2000 0892-33 10/00 O 2000 Society for Scientific Exploration ESSAY Modern Physics and Subtle Realms: Not Mutually Exclusive ROBERT D. KLAUBER 1100 University Manor Drive, 38B, Fai@eld, IA 52556 e-mail: [email protected] Abstract- One facet of the change in worldview ushered in by the quantum mechanical revolution was DeBroglie's discovery that all particles are actually wavelike and that because of this, a plurality of particles can occupy the same region of space, at the same time. This well accepted and empirically validated principle is explored in the context of the quantum field theory of force fieldlparticle coupling. It is then shown that subtle (nonphysical) realms could readily exist without being in any way contradictory to, or inconsistent with, modern physics. Keywords: interactions-coupling-subtle realms-four forcesother universes There is a common misconception, held by many scientists and nonscientists alike, that the laws of physics preclude the existence of nonphysical entities and any concomitant metaphysical realms. This viewpoint, as it turns out, is a vestige of pre-quantum mechanical scientific thinking and in no way represents a constraint imposed on reality by the postclassical physics of our modern age. A general pre-twentieth century scientific adage (even an axiom) held that no two objects could occupy the same place, at the same time. It was therefore implicit that apparitions and similar entities having the property of coexisting in time and space with physical structures such as doors and walls could not possibly exist. DeBroglie's discovery of the wavelike nature of matter changed that perspective dramatically. Today, physicists regularly deal with wave functions of leptons, quarks, photons, and the like, which overlap and share identical regions of space and time. Just as two waves rolling over the ocean heading in opposite directions can pass through each other unscathed, though occupying for a time the same area of the water surface, so too can two subatomic wavelparticles pass through one another unaltered, coexisting for a time in the same space. Trapped particles can, in fact, share a common "trap" indefinitely. If two such wavelparticles jointly occupy a particular region of space-time and do not interact with each another, then neither changes in any way. Often, however, they do interact, and such interaction can change their energies, momenta, charge, and other properties. Interactions are mediated by force fields 276 Robert D. Klauber between the particles, and these force fields carry properties such as energy, momentum, and charge from one particle to the other. In quantum field theory these force fields, the carriers of the properties between particles, are actually propagating waves. Because waves and particles are essentially one and the same, we commonly refer to the force fields as particles, or more precisely, as virtual particles. They are called virtual (in contrast to the real particles whose interactions they mediate) in part because they are singularly evanescent. For example, a first real particle such as an electron can emit a virtual particle such as a photon, which subsequently is absorbed by a second real particle such as a second electron. The two electrons change energy and momenta (i.e., each recoils from the other), and we can measure those changes. The photon on the other hand exists only very briefly, long enough for it to carry the appropriate amounts of energy and momentum from electron one to electron two. We can never measure the photon physically and so distinguish it from the real particles by'calling it virtual. According to our current understanding, all forces are mediated by such virtual particles. But particles are really waves, and these "wavicles" make up the entire universe. The reason our universe isn't simply an uninteresting collection of independent waves continually passing through one another unimpaired and immutable is that the various waves are coupled to one another via interactions (forces). The wavelike particles making up your hand do not pass through an object such as a door because the electrons (waves) in your hand and the electrons in the door interact; that is, they continually exchange copious numbers of virtual photons that effectively push the door away when the hand "touches" it. Without this interaction (the coupling of particles in the hand to other particles in the door), the hand would simply pass directly through the door, never feeling the sensation of touch and in fact never knowing the door was there. It turns out that there are four interactions, or forces, known to modern physics. Two of these-the electromagnetic and gravitational interactionsare familiar in our macroscopic world, and two-the strong and weak forcesare predominantly subatomic. We presently believe that each of these four forces is mediated by a different type of virtual particle. The photon mediates the electromagnetic force; the graviton, the gravitational force; the gluon, the strong force; and intermediate vector bosons, the weak force. It is important to recognize that we know a particle exists (actually that anything at all exists) only because of the coupling (interactions) between particles. For example, an electron interacts with an electron detector by exchanging virtual photons with that detector. A detection signal occurs only because the electron being detected is coupled via the electromagnetic force to the electrons in the electronic circuitry of the detector. Similarly, if we feel a door with our hands, or perceive through any of our senses, it is only because the particles in our sense organs are coupled to the particles transmitting particular properties (information) from the object we perceive. If there is no coupling, there is no perception. Modem Physics and Subtle Realms 277 A real world exemplar of this principle is the neutrino, a particle that has no electric charge and that therefore is not coupled to any charged particle via the electromagnetic force. A human skin cell or a particle detector that responds to virtual photons (i.e.,is coupled to the electromagnetic force) could have many neutrinos passing through it but would never register a thing. Similarly, neutrinos have no coupling with the strong force. So a detector that might be sensitive to virtual gluons would likewise be transparent to, and unable to detect, neutrinos. The various particles in creation are coupled in different ways via different combinations of the four forces. For example, the electron has electromagnetic, gravitational, and weak coupling, but not strong coupling. Quarks are coupled to all four forces. Neutrinos, because they are massless, or extremely close to massless, have gravitational coupling that is far too small to measure, and hence they effectively possess only weak coupling. This singular characteristic of neutrinos makes them not only interesting, but also particularly relevant to the theme of this article. Note that the only way we can detect neutrinos is via the weak force. But the weak force is so named because it is feeble. In fact, it is so extremely feeble that more than 200 billion neutrinos have passed through your thumbnail in the time it has taken to read this sentence, yet you felt nothing. The weak force (the only way your nervous system could have detected the neutrinos' presence) is so slight that none of those neutrinos interacted with a single atom in your nail. In fact, the only way neutrinos are actually detected in experiments is by using huge volumes of matter over long periods of time. In typical experiments, a mere handful of actual interactions is detected over many months. This near imperceptibility of weakly interacting neutrinos makes them almost ghostlike. They pass through matter virtually without our being aware of their presence. More remarkably, another property of the weak force may make certain neutrinos even more tenuous and even less a part of what we consider our universe. The weak force is restricted to particles physicists designate as having "lefthanded chirality." In oversimplified terms, one can think of an electron, neutrino, or quark as spinning, typically with the spin aligned in the direction of travel (velocity). Consider that the spin can be thought of as either clockwise (right-handed) around that direction or counterclockwise (left-handed). Peculiar as it may seem, only left-handed neutrinos couple with the weak force. Right-handed ones are immune, and hence transparent, to its effects. So only a left-handed neutrino could interact via the weak force with another particle such as a quark, electron, or other neutrino. The key point is this: Right-handed electrons and quarks exist. We know because they have been detected via the electromagnetic force. But we can not detect right-handed neutrinos in such a way because they do not interact electromagnetically. Because we can not detect right-handed neutrinos weakly, there is essentially no way to know if these particles even exist. Yet, there could be untold trillions of them passing every minute through each of us and 278 Robert D. Klauber through every known detector. If left-handed neutrinos are almost ghostlike, then right-handed neutrinos are fully so. Consider then that conscious beings in our universe are aware of each other, the rest of the universe, and at least some aspects of their own selves only because of interactions between the particles/waves of which physical objects are made. As noted, these interactions, as far as we know, are limited to four. Consider further the possible existence of a new family of diverse particles, similar to right-handed neutrinos in that none of them interacts via any of the four forces dominating our reality. This new family could consist of a limited number of types, each of which fills our known universe in immense numbers leading to significant densities. Consider further that this family might have three or four or five different interactions of its own, coupling its members in various ways. This family and its set of interactions could then behave in generally similar fashion to our own family of particles and force fields, although it would have unique types of interactions manifesting as a complexity and chemistry all its own. It might evolve, grow, and perhaps even produce intelligent beings. And it would never be detected by any of us-at least through our physical senses. We would coexist in the same space and time, yet because all quantum waves in that system would pass unperturbed through, and without perturbation of, our system, we would live our lives oblivious to this other independent cosmos. If there is one such other family, why not many? In fact, why not a great many? The universe certainly favors unimaginably large numbers. If, as we suppose, there are an uncountable number of galaxies (including those beyond our horizon of visibility) and as many theorists propose, an uncountable number of other possible universes, then why not an uncountable number of other independent particle families? In the very place where you now sit, there may now also sit a plethora of other sentient beings, some of whom might also be pondering the sensory limitations of their particular version of quantum field theory. In this context, the proposition that a heaven or hell coexists in space with us might start to seem rather plausible. So might reports of close UFO encounters in which alleged advanced civilizations seem capable of manipulating and moving between physical and nonphysical realms. The list readily expands to near-death tunnels, spirits, angels, auras, astral planes, other "dimensions," and various other concepts relegated by many mainstream scientists to the arena of fantasy. When certain individuals assert they perceive such things, perhaps the proper scientific response should be investigation, rather than the more common practice of disparagement and dismissal. Something in these peoples physiologies may be somehow coupled, in presumably delicate fashion, to one or more other worldly force fields. We know individual consciousness and its attendant physical body interact in ways we still do not fully understand. Could that same consciousness not also 7 Modern Physics and Subtle Realms 279 1 interact, in still less understood ways, with all but impalpable, but nontheless equally real, trans-physical bodies? In concluding, we note that we certainly have not proven that subtle realms actually exist. Yet we must bear in mind that in the long history of mankind's numerous metamorphoses in paradigm, the universe has repeatedly surprised us by being far more extraordinary and expansive in every regard than we had previously imagined (or even, as some have said, than we can imagine). Given such a history, it would seem prudent to proceed carefully and without prejudice in matters of purported metaphysical nature and draw conclusions based on empiricism alone. In particular, no proponent of materialism should ever denounce as scientifically indefensible claims made by others regarding the possible existence of nonphysical realms. As we have seen, modem physics imposes neither a limit on the probability for existence of such transcendental worlds, nor restrictions on their nature, total number, or ultimate extent. Journal of Scientific Exploration, Vol. 14, No. 2, pp. 281-299,2000 0892-3310100 O 2000 Society for Scientific Exploration BOOK REVIEWS The User Illusion: Cutting Consciousness Down to Size by Tor Nonetranders. Translated by Jonathan Sydenham. New York: VikingIPenguin Putnam, 1998. xii + 467 pp. $29.95 (c). ISBN 0-670-87579- 1. This book is not a turgid treatise in defense of objectivity in the so-called science wars. It does not discredit the emerging field of consciousness studies, and it does not call for the reinstatement of consciousness as the leading taboo topic in psychology. It is not a public broadside promoting biological materialism and reductionism. And it does not cut consciousness down to size by advocating either polarity of epistemology-realism or idealism. Here is what this book is about. The original title of this book in Danish, Maerk Verden, means "mark the world." This aptly summarizes Nonetranders' thesis, namely, that human perceptual processes carve out-or mark off-small segments of the physical world and represent them to our conscious awareness as reality. To do this, our sensory systems must filter and discard enormous quantities of information. The author terms this discarded information "exformation" (e.g., the first paragraph of this review is similar to exformation). In contrast, the English title, The User Illusion, is a somewhat oblique reference to a concept developed in the early days of the personal computer, which describes the intuitively designed interfaces between the owners of personal computers and their machines. Unfortunately, this is not fully explained until page 290. In a footnote, Norretranders rebuts effectively the criticism that the so-called user illusion is nothing more than the latest metaphor in a long line of technologically based ones about the mind. The legs of his thesis are four empirically verifiable neuropsychological phenomena: (a) the extremely limited "bandwidth" of our conscious awareness in the here and now, (b) the undervalued role of subliminal perception in everyday life, (c) the half-second delay between sensory inputs and conscious awareness, and (d) the backward referral in time of subjective experience. Yet, given these factors, the Danish title seems preferable, because it summarizes how consciousness establishes physical reality from a neuropsychological point of view. This book about consciousness does not include any discussion of consciousness-related anomalies. This is unfortunate, considering the light that the four phenomena cited above could shed on extended human abilities. Some readers will be vexed by the following elements: (a) The New York Times' Op-Ed style of editorializing about social implications woven through some of the later chapters, (b) the fact that the author is a prominent intellectual and journalist, but not an academician, and (c) the author's method of 282 Book Reviews scholarship, which impresses me as a good example of Polanyi's notion of "personal knowledge." This book is worthy of being read, because it is well-written, the chapter notes, bibliography, and index are edifying, and specialists and nonspecialists will find the content satisfying. On the balance, The User Illusion embodies an expression that Einstein reportedly used during his Princeton years whenever he was working on a particularly difficult problem, "I will a little think." Arnold L. Lettieri, Jr. PEAR Laboratory C-131, Engineering Quadrangle Princeton University Princeton, NJ 08544 Lamarck's Signature by Edward J. Steele, Robyn A. Lindley, and Robert V. Blanden. Reading, MA: Perseus Books, 1998. xxi + 286 pp. $25.00 (c). ISBN 0-7382-0014-X. Epigenetic Inheritance and Evolution by Eva Jablonka and Marion J. Lamb. New York: Oxford University Press, 1995, x + 346 pp., $60.00, (c). ISBN 019-854062-0. The incredible richness of the biosphere challenges us with obvious questions: What is the origin of diversity among plants and animals? Are natural kinds static, or do they change over geological time? What are the forces causing these changes, and are they directed toward specific ends? Is there any drive toward increasing complexity, and if so, what is the relationship of this process to entropy? What is the interplay between heredity, embryonic development, and ecology? A theory that addresses these issues is needed as a cornerstone of biology. We now know a significant amount about the processes at work during the morphogenetic events that turn a fertilized egg into an embryo and then an adult. In large part, these processes are directed by the genome-information that is instantiated by patterns within DNA molecules. This information is passed from parent to offspring and carries instructions telling molecular machines how to assemble and modify a growing embryo, as well as instructions for building the machines themselves. It is now appreciated that the genome is not simply "beads on a string" of DNA; functionally, it consists of logic elements, subroutine modules, addresses for information retrieval and modification, error-correction mechanisms, hierarchical organization, etc. (see Shapiro, 1991). The inheritance of acquired characteristics theory was a natural way to at- Book Reviews 283 tempt to explain cases of marvelous adaptation of organisms to features of their environment. This idea is most often associated with Lamarck, though it had adherents both before him and after him. Moreover, Darwin himself thought that inheritance of acquired characteristics did occur. This theory has the benefit of providing easily believable explanations. For example, the camel has rough pads on its knees because constant kneeling in gritty sand causes a thickening of the skin in that area. Unfortunately, in this most simplistic form, this theory faces a number of problems. For example, it would seem that most changes acquired by an organism within its lifetime are due to injury, disease, and age and are not beneficial to pass on to offspring. Likewise, it is often pointed out that thousands of years of circumcision have not produced a tendency for males to be born without foreskins. "Weismann s Barrier" provides a "genetic chastity belt around the reproductive organs" (Lamarck's Signature, p. 165). In most animals, offspring arise from a very small subset of the adult organism's cells-the germ line, whose cells are segregated long before adulthood, so changes occurring to somatic body cells are not passed on. It is possible to attempt to circumvent such objections. For example, Epigenetic Inheritance and Evolution discusses interesting complications such as the possibility of existing germ cells being subject to selection and competition; these processes may well be affected by environmental factors, simulating environmental inheritance when applied over the endogenous variation among germ cells. However, there is a much more basic reason why large-scale morphological changes wrought on the adult body cannot be passed on to offspring. The crucial factor is embryonic development. It is now known that embryos arise as the result of the complex interactions of chemicals generated in response to information contained in the DNA. The genetic material does not appear to contain descriptions of the outcomes of embryogenesis. The existence of reverse transcription (turning mRNA back into DNA), and even the possibility of reverse translation (turning a protein into mRNA), is simply not enough. The process of development is unidirectional; the laws of biochemistry are what drive the generation of form (such as a limb or nervous system) from basic ingredients such as proteins. There is no way to reverse the process and compute how to adjust every protein, including structural ingredients as well as controlling machinery, to result in a particular feature, such as a longer neck. The process of generation of fractal images shares this property. Given a simple mathematical function such as z = z2 + c, it is easy to produce fractal images of great complexity and beauty. The nature of chaos theory makes it impossible to reverse the process. Given a different image, it is simply not possible to know how to alter the function to produce the desired image. The same is true of embryonic development. However, there are limited cases in which inheritance of acquired characters does appear to occur; these are discussed in fascinating detail in Epigenetic Inheritance and Evolution. For example, plants do not all have a germ linelsoma 7 284 Book Reviews distinction. In plants where a new organism can regenerate from any part of the adult, changes made to adult body parts can indeed become part of the offspring. Likewise, protozoa (single-cell organisms) can pass on altered features of their cytoskeleton, for example, because they reproduce by fission of the adult organism. Bacteria can acquire genetic material from the environment (in the form of "minichromosomes" such as episomes), and because these also reproduce by fission of the adult, such changes can become a permanent part of the reproducing line. Given the powerful reasons why Lamarckian mechanisms are unlikely to play a role in large-scale morphological change, what alternatives remain? The Darwinian theory is a conclusion drawn from three facts, all of which were known long before Darwin's day. These basic axioms are as follows: (a) Offspring resemble their parents more than they resemble unrelated individuals (equal traits are hereditary), (b) he transmission of characters from parents to offspring occurs with high but imperfect fidelity, and (c) resources (such as food, mates, territory, etc.) are limiting because of the exponential growth of populations during times of plenty. The genius of Darwin was to realize (and support by painstaking observation) the claim that these three facts, when joined together, form a powerful system for explaining evolutionary change. Because of the intense competition for limited resources, even small advantages lead to increased chances for survival; thus, beneficial errors in hereditary transmission will accumulate and eventually dominate the population over geological timescales. This basic scheme, when carried over a sufficient number of individuals, over sufficiently great timescales, is an immensely powerful way to leverage random changes into adaptive results. This algorithm is so generally useful that it has been applied to explain the fine-tuning of everything from thoughts and concepts in human culture (Dawkins, 1976) to basic physical laws of the universe (Smolin, 1997) to computer programs that solve problems for which we have no algorithmic solution (Koza, 1992). In biology, as well as in applications of genetic strategies for other sciences, the basic issue is to define the unit of selection (genes versus organisms; e.g., Dawkins, 1982) and the fitness function, a way of assessing the quality of individuals. It should be noted that the three basic axioms of modem evolutionary theory are not controversial. What may be questioned, although it usually isn't, is the issue of whether they are sufficient to explain the incredible richness of the biosphere. The success of mimicking the Darwinian strategy to design computer programs (known as Genetic Programming, Holland, 1992) suggests that our intuition far underestimates the creative power of this process. At the same time, progress in the field of Genetic Algorithms research seems to have slowed considerably on the hard problems, and it remains an open question as to whether a closer look at the mechanisms of development (the layer between genotype and phenotype) will be sufficient to successfully apply evolutionary strategies to those difficult problems. Book Reviews 285 The idea that random changes can be harnessed by any mechanism to result in everything from viruses to human beings is repugnant to some, and a wellknown group of antievolutionists exists. They have argued that physical processes in and of themselves are not sufficient to result in the appearance of complex biological organisms. It is important to note that Lamarckian inheritance is of no use whatsoever to religious fundamentalist or dualist thinkers in arguing against the sufficiency of material evolution. A theory of environmentally caused changes that are inherited is just as mechanistic as neo-Darwinism; it is simply another way to explain adaptation without recourse to a grand design or creator. Of course, some will argue at this point that it is God who directs which somatic changes are carried into the genome. This move is useless, as then one might just as well posit that God controls what mutations occur in DNA, to later be subject to Darwinian selection. These are all interesting issues and are dealt with in these two volumes. Epigenetic Inheritance and Evolution is an excellent book for the amateur or professional biologist. It begins by reminding the reader that heredity is not transmission of characters but of instructions: A pure line of animals has identical genetic instructions, but the manifestation of their characteristics can be influenced environmentally. This book is packed with interesting special cases in which Lamarckian-like inheritance occurs. It covers inheritance systems other than the genome (epigenetic inheritance systems), such as episomes, phages, retroviruses, mitochondria, cytoskeleton (cilia, basal bodies, entrioles), chromatin marking, methylation, X-inactivation, mobile elements in DNA, etc., and discusses in excellent detail their role in sex, evolution, multicellularity, speciation, meiotic drive, hybrid inviability, etc. It has enough detail to satisfy, and it makes great reading for anyone who is interested in the details of real biology. Also included are viral shuttling of genetic info between organisms (horizontal inheritance), imprinting, parental age effects, and maternal control of gene expression. Issues such as segregation of germ line in various organisms and interactions between epigenetic systems and the genome are covered as well. Despite these fascinating examples of alternative types of inheritance, the authors point out early on (p. 20) that somatic changes are not put back into the germ line. Lamarck's Signature is also a very good read, aimed at more of a layperson audience. It is part of the Frontiers of Science series (edited by Paul Davies)a set that contains several other excellent books. This book doesn't cover as many examples of nonstandard inheritance as Epigenetic Inheritance and Evolution. The first six chapters are a good introduction to basic evolution and Lamarckism, molecular and cell biology, and immunology; they cover the basics in good detail: DNA, RNA, transcription, editing, viruses, recombination, reverse transcription, enzymes, etc. Only the last two chapters are about Lamarckism, and the first few set up the background. Lamarck's Signature focuses on the idea that the immune system instantiates an example of a Lamarckian inheritance mechanism. A cornerstone of the 286 Book Reviews book is the fact that the immune system is able to specifically react to a practically infinite number of antigens, which are not known in advance. It does this by a selectionist scheme, whereby specific pieces of DNA are randomly mutated and those cells containing mutations, which allow them to bind antigens effectively, proliferate. The basic idea of this book is that DNA mutation is induced within a subset of lymphocytes as an attempt to further optimize the binding of those molecules to the antigen. This in itself is an example of the environment's effects on changes in DNA (even though the mutation events themselves are still random). Moreover, by the action of reverse transcriptase (an enzyme, provided by endogenous retroviruses, that turns mRNA back into DNA) and homologous recombination within germ cells, such mutations can be inserted into the germ line and become inherited. The authors summarize, "Rapid random mutation involving the replacement of the original DNA sequence coupled to selection for a better one (encoding an antigen-binding site with higher affinity than the original antibody) effectively results in the appearance of 'directed mutation,' the complete antithesis of random mutation" (p. 162). The two books complement each other nicely and are recommended for anyone who wants a thorough, professional, well thought out discussion of often neglected aspects of biological inheritance. Epigenetic Inheritance and Evolution is more technical, containing lots of in-depth examples and references. Larnarck's Signature is probably accessible to a wider audience; it has a very good introduction to relevant issues, but expert readers will want to skip to the last few chapters, which focus on Lamarckian inheritance in the immune system. The word Lamarckian tends to arouse strong feelings. It should be noted that one might pick up either book expecting revolutionary models of the emergence of complex animal forms by non-Darwinian means. If so, one would be disappointed. Upon reading both books, it becomes clear that there are lots of examples in biology of simple, single-cell systems (and some plants) where inheritance of acquired characters does indeed occur, and there are lots of complex effects in higher animals where inheritance of acquired characters is mimicked by other processes. What is not found in these books (and nowhere else that I am aware of) is any evidence for the truly controversial claim of Lamarckism-that changes occurring to the adult bodies of organisms can be transmitted to offspring in such a way as to make them better adapted to their environments. The bottom line is that all of the controversy and excitement regarding Lamarckism and its "challenge to Darwinian orthodoxy" is not only about environmental influences on the genome-we know these exist-and the idea that stress or other environmental inputs can affect mutation rates, etc., isn't very frightening or damaging to the conventional paradigm. The truly threatening idea about Lamarckism is the question of teleology: Can the environment direct changes that are beneficial to the organism? Mutation may be non- Book Reviews 287 random with respect to location (hotspots in the DNA, etc.), but it seems to be always random with respect to adaptive advantage. Lamarckian inheritance is exceptionally exciting if and only if it can actually help to explain the origin and changes in major aspects of morphology, as well as the exquisite machinery of the cell. Although the importance of Darwinian models of evolution is incontrovertible, deep and interesting challenges remain, for this and other theories of life (Behe, 1996; Denton, 1986; Margulis, 1998). References Behe, M. J. (1996). Darwin's black box. New York: The Free Press. Dawkins, R. (1976). The selfish gene. New York: Oxford University Press. Dawkins, R. (1982). The extended phenotype: The gene as the unit of selection. San Francisco: Freeman. Denton, M. (1986). Evolution: A theory in crisis. Bethesda, MD: Adler & Adler. Holland, J. H. (1992). Adaptation in natural and artificial systems. Cambridge, MA: MIT Press. Koza, J. R. (1992). Genetic programming: on the programming of computers by means of natural selection. Cambridge, MA: MIT Press. Margulis, L. (1998). Symbioticplanet. New York: Basic Books. Shapiro, J. A. (1991). Genomes as smart systems. Genetica, 84, 3-4. Smolin, L. (1997). The life of the cosmos. New York: Oxford University Press. Michael Levin Cell Biology Department Haward Medical School Boston, MA 02 115 Authentic Knowing: The Convergence of Science and Spiritual Aspiration by Imants Baruss. Purdue University Press, 1996.228 pp. $14.95 (paper), $32.00 (cloth). ISBN 1-55753-085-8. Many of us believe that the signal task of human existence is to achieve a better understanding of who and why we are. Although the nature of such efforts has varied throughout history, in the past 400 years, they have consistently taken shape along a continuum defined by the polarities of religion and science. This admirably honest book attempts to harmonize these disparate elements into a common theme of man's efforts to define his place within the universe. Baruss, a professor of psychology, contends that both science and spirituality express a need for "authentic knowing," a concept derived from Heidegger's Eigentlichkeit. The author summarizes the concept of authenticity as "the effort to act on the basis of our own understanding." He expands on this idea in a concentrated foray through classical and quantum physics. In so doing, he draws attention to our tendency to compartmentalize scientific and religious beliefs such that they apply little to our everyday actions and 288 Book Reviews thoughts. Such incongruities lead him into a discussion of the philosophy and social psychology of science. Here, he notes the heterogeneous composition of the scientific community, despite its claims to an objective universality (something with which readers of the Journal of Scientific Exploration are no doubt familiar). He quotes an old Sufi story to observe that many scientists seek smaller truths that lie where the light of their methods is brightest, rather than looking in the "darker" places where more important findings may lie concealed. With this introduction, he distinguishes between what he calls authentic and inauthentic science. The latter Baruss refers to as "scientism," and lists some of its characteristics as materialism, an emphasis on the blind accumulation of facts, and a belief in a universal, inflexible scientific method that can guarantee truth. Here, I believe he distinguishes the scientific method as a useful tool for discovery from its worship as a secular icon. In a wide-ranging discussion, Baruss explores theories of the mind ranging from Freud and Jung to Assagioli and Maslow, drawing attention to the similarities between certain types of meditative practice and psychotherapy. I was interested in the analogy he draws between witnessing meditation and psychoanalytic training. He discusses several scientific anomalies and applauds the formation of the Society for Scientific Exploration, although noting that its formation has not solved problems of funding nontraditional areas of research or creating new academic curricula. Baruss stumbles a bit when he attempts to solve the important problem of subjectivity by somewhat obscurely citing "modes of understanding superior to rational thought." These may well exist, but we remain too far from their understanding to employ them as research tools. Baruss goes on to discuss human transcendence (of the bounds of perception and conation) as exemplified in the writings of Franklin Wolff. Wolff, a mathematician by training, after achieving "enlightenment," wrote two books in which he made a sincere effort to describe the ineffable using analogies taken from mathematics. Treading into more dangerous waters, Baruss investigates the theosophical model as a means of describing transcendent states. This topic, too, is given an even-handed and responsible treatment. He is clearly aware of the dangers of blind obedience to any system and notes, as perhaps did John of the Cross, that leaving the security of the conventional entails a certain loss of groundedness, with the temptations of both scientism and blind spiritual allegiance (to regain a footing) being the price of one's authenticity. The author makes many important points. He notes that the normal functioning of the mind may obliterate the transcendent and that the best of science and spiritual practice are performed in a search for authentic knowing. He observes a hidden intelligent dimension in ourselves that occasionally irrupts into our lives with moments of transcendent insight. Here, he reminded me of Mircea Eliade, the great historian of religions, who took greatest delight in works of fiction that attempted to explore the meaning of such moments. To close with a quote from the book, "Knowledge is the currency of science, and when it is freed from the black and white constraints of its inauthentic aspects, Book Reviews 289 science can be opened up to a colorful exploration of consciousness and the ultimate nature of reality." I recommend this book highly. The author does not claim to know the final answers, but he has formulated the questions clearly enough. I suspect Diogenes could rest easily in Professor Baruss' classroom. James A. Scott Massachusetts General Hospital Deptartment of Radiology Harvard Medical School Boston, MA 021 14 Alien Abductions: Creating a Modern Phenomenon by Terry Matheson. Amherst, New York: Prometheus Books, 1998. 317 pp. $26.95, (c). ISBN 157392-244-7. The UFO rumor, craze, myth, or narrative fad has taken a fresh lease on life in the last couple of decades. Perhaps its biggest boost came in 1988 with a vasteyed neotonized grey gazing out from hundreds of thousands of paperback covers of Communion, horror writer Whitley Strieber's purportedly true story of alien abduction. It had already entered mass culture in a big way with Spielberg's 1977 Close Encounters of the Third Kind and is now firmly entrenched via the X-Files and the National Enquirer. What's going on? Are aliens really abducting and anally probing or fetus-harvesting thousands of reluctant people, perhaps for their entire lifetime and even beyond death into endless hybrid rebirths? That is a less interesting question for Terry Matheson, English professor at the University of Saskatchewan, than the puzzle of this narrative's immense popularity in a culture ostensibly shaped by rational bureaucratic and scientific canons. Professor Matheson's enjoyable, plainly written study proposes that the UFO narrative gains its force from its adoption as a myth, an organizing structure that (borrowing from Northrop Frye, Roland Barthes, and other mythologists) blends dream and ritual to "reflect a culture's preoccupations and concerns as well as the things it fears," ordering and structuring (although hardly explaining) "certain aspects of the world that would otherwise remain unintelligible or objects of dread" (p. 285). Some of these elements are age-old-the nature of cosmic, social, and interpersonal forces often allegorized in astrological or sacred doctrines-while others are typical precisely of our technological and allegedly rational epoch. Ancient Greeks figured powerful ambiguities of lust, war, and political power in such mythologized tales as Leda's rape by Zeus and her subsequent triple impregnation with twins, Castor and Pollux, and the Helen fated to ruin two cultures in protracted battle. Today, Matheson argues, we find in the bleak, emotionless, and faceless UFO aliens a strikingly vivid figuration of late twentieth century postindustrial life at its worst; but also at its darkly numinous, I 290 Book Reviews with technology's ceaseless parade of wonders and medical promises, we find its ambiguous offer that we might soon transcend human limitations. Matheson's approach is historical (technically, diachronic), tracing the slow waxing and waning of subordinate elements in this developing mythos (a narrative form often regarded as timeless or synchronic). While he claims to bracket the reality of UFO abductions, attending solely to the narrative or semiotic codes and forces in play, this is disingenuous. The book is published, after all, by Prometheus, notable for its sturdy inquisitorial list of texts by people friendly to CSICOP and other ardent skeptical organizations. His method is known in literary studies as close reading: In a forensic way, he examines the claims and descriptions in several notable books by and about abductees, testing them for consistency, plausibility, sequential influence, and their rhetorical design and devices. A brief introduction to the nature of narrative draws on some technical work from the 1980s and earlier by narratologists and morphologists. Written testimony is not a clear windowpane through which we can see the truth of reported events. Authors help form and reshape the tale they tell. Even in legal narrative, the courtroom "construction of truth" is "primarily a matter of the overall narrative plausibility" (p. 37), the case as a whole rather than the details. Still, he notes, realistic details help, and it is noteworthy how many get inserted later into abduction narratives as authors tidy the blurts and scraps or summarize floods of verbal and pictorial testimony. Avowals of integrity and probity are frequently used to fill the void left by the inevitable absence of hard or compelling evidence, so such books usually begin with "an introduction or preface written by a presumably objective third party who often possesses impressive academic credentials" (p. 38). Such strategies are not invoked to deceive but are part of the protocols and practice of writing and reading, which impose certain "resemblances.. .from account to account [that] may say more about the nature of a realistic narrative's inner logic" (p. 39) than speak to their true content. All this is compelling, yet Matheson's skeptical presumptions can lead him astray. Citing a certain blatant inconsistency in John Fuller's The Interrupted Journey, he remarks: "Because Fuller makes no attempt to resolve this, the Hills' credibility is bound to suffer" (p. 53). Yet it did not do so, on that basis at least, which is a surprising fact that Matheson's methods do not quite resolve. His investigation proper launches from that key 1966 text, which dealt with the celebrated case using hypnosis, in the early 1960s, to unpack (or perhaps instill) the terrifying Ur-abduction reports of Betty and Barney Hill. In the nearly four decades since, its core story has gathered fresh elements, dropped others, strengthened that early reliance on hypnosis while abandoning the stuffy requirement that trained specialists should do the honors, and in general has followed a course convincingly seen as the elaboration of a living myth. Matheson takes us through consecutive versions from Raymond Fowler (Betty Andreasson's abduction and space tour, replete with fundamentalist im- Book Reviews 291 agery and glossolalia appropriate to her Christian faith-and Fowler's own abduction, belatedly realized) to Travis Walton's self-authored testimony, Ann Druffel and D. Scott Rogo's treatment of several gay women who oddly enough were spared the usual phallic probing, the vatic arrival of Budd Hopkins with his menacing extraterrestrial aliens, and Strieber with his even more terrifying occult shapeshifters, Ray Fowler again with four men in a boat whose drawings and stories do not especially resemble each other, despite the extensive interpretative zeal and leading questions of their interrogator, concluding with the arrival of the big guns from academia, historian Dr. David Jacobs and Harvard psychiatrist Professor John Mack. Matheson adopts an annoying tic in these analytic recountings: Often, sentences tell us that "many readers may be inclined to.. ." (p. 86), "some readers may conclude that.. ." (p. 117), "many readers will emerge from this section having concluded.. ." (p. 208), and on and on. This is exactly the kind of narrative bullying Matheson discerns so frequently in the abduction documents. I found myself wondering whether the first draft was a straightforward skeptical demolition of these often woolly, dreamlike, perverse, and inconsistent narratives, lightly rejigged with slightly old-fashioned narratological gadgetry in order to gain a publishing niche. Professor Matheson's account is persuasive, as far as it goes. According to structuralist Claude L6vi-Strauss, myth is an ideational and affective ensemble of stories that articulates schemata of expected behaviors while specifying and defusing its culture's antinomies or internal contradictions. Its aim is the coercive institution of order, regularity, and harmony-even if those ends are met, at times, through controlled ritual passages into frenzy, carnival, and hysteria. This is the type of analysis adverted to in the book, although rarely attempted in any subtlety. Occasional references to Thomas Bullard's 1982 Ph.D. thesis, the only work I know of other than Jacques Vallee's to look closely into the folkloric and mythological components of the emerging mythos, make one hunger for more detailed analytics-a far more exact semiotic unmasking of the codes at work, rather than vague if plausible generalizations about the impact of science fiction iconography on vulnerable people at the ends of their twentieth century tethers. Structuralism, of course, is now out of fashion, replaced in the humanities by variants of critical theory, Lacanian psychoanalysis, poststructuralism, and discourse theory-approaches that emphasize a shimmering uncertainty where earlier models sought a reliable if abstract binary algebra. Even so, like Lkvi-Strauss's myths and Jung's archaic dreams risen from a postulated collective unconscious, discourses are held by contemporary literary theory to possess an eerie autonomy, indeed a preeminence over any thought or intention that one might suppose lies behind the utterances they "enable." Like cosmic radiation, discourses traverse the fragmented subject. Meanwhile, cognitive and experimental neurosciences offer similar accounts, replacing old folk images of a unified self with modular brains and multiple intelligences. Consciousness, individual and collective, is prey to memes, an inward ecology of I 292 I 1 Book Reviews mental viruses that might indeed in their totality constitute the parliament of the self. On this view, a myth might be a kind of commensal package of memes, a roaming mental genome that infests us even as we learn some local patois and hear or speak its latest modish utterances. Can hermeneutic, narratological, semiotic, or discursive analyses take us much farther than Matheson manages in his intriguing but frustrating book? He provides a telling critique of the major texts in the abduction industry, but he does not go very deep or venture far from the books under consideration. There is no discussion of the murky undergrowth that preceded and paralleled the invasion of the grey gynecologists: the delicious semi-occult contortions of John Keel (especially) and Brad Steiger, let alone the subterranean foliations and filiations of mind cults and more reputable marginal belief systems: the Heaven's Gate dupes, Scientology with its space opera theogony, Theosophy and its Ascended Masters, all the charmingly crackpot scholarship that Desmond Leslie assembled in the 1950s in Flying Saucers Have Landed. The vimanas of ancient Atlantis and Mu might not seem direct ancestors of John Mack's aporias, half in this world and half beyond it, but I scent a narrative trail. And there remain endless apertures for an inventive reader of these accounts. Andreasson's friendly ufonauts passed along faux explanations Matheson finds 'unedifying,' cast in "pseudoscientific language" (p. 198). This is so, but consider the following: the hybrid-engineering aliens must "put their 'protoplasma' in the 'nucleus of the fetus and the paragenetic,'" and Betty mentions "balancing 'the oscillating telemeter wheels"' (ibid), although this latter task apparently relates to their propulsion or guidance systems. But Betty simply might have been confused; in the last two years, biologists have found they can significantly extend the lifespan of in vitro human fibroblasts by inserting into the nucleus a genetic package that codes for the recently discovered enzyme telomerase. That prompts the chromosomes to repair their ever-shortening telomeres (key to their replicative capacity), which very recently were found to form a wheel or loop at the ends of DNA strands. Surely that is not what the aliens were telling Betty, but I would not be surprised if the evolving mythos incorporates such a reading with a cry of recognition. This kind of urban myth is so charming, so B-movie sci-fi, that for decades I have gobbled down the revelations of Adamski, Mack, Jacobs, C. D. B. Bryan (that upmarket journalist). I wallowed in Whitley Strieber and his creepy, profitable concoctions or perhaps psychodramas and laughed my head off at Jim Schnabel's splendid travelogue among the beamed-up, Dark White. Somewhere in there, a curious prickle ran down my spine. I started making lists, drawn from these books, of the signs and symptoms of alien abduction. I recalled the primary school near Monash University, where a whole class and the teacher witnessed a close encounter of the third kind, just a kilometer or two from where I was studying in April 1966. I glanced back through my own science fiction novels and, one after another, out popped virtually the entire checklist: the investigation and probing on the floating slab, the wafted transi- Book Reviews 293 tion through a wall in a bubble, the mysterious mutant fetus, the transferred embryo, the creatures suspended in tubes, the occlusions of memory, the greateyed animals with cold voices, the prophecies of doom or transformation, etc. I am not about to spring any unseemly revelations, leap from the UFO closet. But it did focus my amazed attention on the ubiquity of these narrative elements, the odd way in which they seem to have seeped into our dreams and our unconscious (or out of it), long before they were written in fat lurid paperbacks or dramatized for network television and Spielberg movies. I do not know their source, and nobody else does either. Matheson is surely correct: The abduction mythos is a culturally created phenomenon or perhaps a spontaneously emergent one, catching in its slowly shifting narratives the changing pressures, fears, and hopes of turn-of-millennium technological societies. I hope other scholars, drawing on more recent techniques and perspectives, soon open out the trail he has broken. Darnien Broderick Department of English and Cultural Studies University of Melbourne, Australia At the Threshold by Charles F. Emmons. Mill Spring, NC: Wild Flower Press, 1997. xi + 268 pp. Skeptics unfamiliar with the field often dismiss the UFO phenomenon with something like, "If it were true, then surely science would recognize it by now." And even for people sympathetic to ufology, the lack of mainstream acceptance is something of a puzzle. Surely, given the reported size of alien craft and the frequency of sightings, one would expect overwhelming evidence despite institutional resistance. At the Threshold goes a long way toward resolving this puzzle. Author Charles Emmons, a professor of sociology at Gettysburg College, adopts primarily a sociological perspective-but not to explain away the UFO phenomenon as myth and folklore. Instead, At the Threshold examines the social mechanisms through which orthodox opinion has come to reject UFOs and through which ufology has become a deviant field. It also surveys the spectrum of opinion-among scientists, skeptics, and believers-regarding UFOs and explores some of the challenges UFOs pose to science, both as a social institution and as a method of understanding reality. Indeed, there is much explaining to be done. Emmons overviews the entire spectrum of UFO phenomena-from sightings, to physical evidence, to abductions, to channeling and other associated paranormal phenomena, making it clear that there is no facile, obvious means to explain away any category of UFO events. In discussing false-memory syndrome, for instance, Emmons discusses the precautions that hypnotherapists use to test for suggestibility and 294 Book Reviews notes the corroborations between hypnotically induced memories and other evidence. At the Threshold gives the lay reader a good feel for the type of dialogue that exists among debunkers, skeptics, and true believers, Without going into the minutiae of the debates, it overviews each type of UFO phenomenon, the explanations typically used to debunk it, and the severe weaknesses of these explanations. Emmons makes a convincing case that although there may be no incontrovertible evidence for any single UFO event, taken as a whole they defy conventional explanation. The debunkers' position is portrayed as a hodgepodge of ad hoc explanations, ridicule of opponents, and out-of-hand rejection of evidence that "couldn't be true and therefore isn't true." Emmons refrains from impugning anybody's personal integrity, but his point that CSICOP and the debunkers are caught in a narrow "reality tunnel" is hard to resist. At the Threshold draws on an impressive array of secondary sources, representing skeptical positions as well as those of ufologists. He also has gathered some new data (on ufology, not UFOs) that will make this book relevant even to seasoned researchers in the field. The surveys and interviews with mainstream astronomers are particularly illuminating in understanding how ufology became a deviant science. Emmons has found that most astronomers are shockingly uninformed about ufology, despite their majority opinion that the existence of extraterrestrial intelligence is highly likely. As Emmons carefully elucidates through interviews and statistics, astronomers (and other scientists) shun the field mostly because there is no funding for UFO research and because an interest in UFOs could ruin their academic careers. Such institutional mechanisms are abetted by media sensationalism and (official) government insistence that it has no evidence of UFOs, to isolate ufology from the rest of (mainstream) science. The disagreement between skeptics and believers as to what constitutes the realm of scientific possibility is mirrored within ufology itself. At the Threshold devotes an entire chapter to the New Age versus nuts-and-bolts division in ufology. While giving a clear exposition of the factions and controversies among ufologists, At the Threshold laudably avoids delving too deeply into the Byzantine web of conjecture, rebuttal, and conspiracy theory that has grown up around many of ufology's famous cases and phenomena. Just as orthodox science has no justification for summarily excluding ufology from its domain, neither can one justify excluding any UFO phenomena, no matter how New Age or weird (and some are weird indeed!) from ufology itself. Thus, taking UFOs seriously calls into question the very boundaries of science-not just scientific fact, but scientific method as well. Of course, ufology is not the only field to pose such a challenge; Emmons argues that such a challenge may even be raised within science itself by the New Physics. Emmons' final message is thus a call for tolerance and a suspension of judgment-an appropriate attitude for any field of scientific inquiry, but particularly for one that so powerfully challenges so many fundamental assumptions about reality. Book Reviews 295 At the Threshold is written in a clear accessible style, free from jargon and free from any obvious agenda to debunk UFOs or prove they exist. Moreover, because At the Threshold covers most major UFO-related issues, it makes a fine introduction both to the content of the field and to its sociology. Emmons has done a commendable job of placing ufology in a social context that takes its researchers and discoveries at face value. Charles Eisenstein 104 Matthew Circle State College, PA 16801 chuck@statecollege. corn Minds in Many Pieces: Revealing the Spiritual Side of Multiple Personality Disorder by Ralph B. Allison, M.D., with Ted Schwarz, 2nd ed. Los Osos, CA: CIE Publishing, 1998.208 pp. ISBN 0-9668949-0-1. The author of this book is probably the most experienced and honored investigator of Multiple Personality Disorder in the world, and certainly in the United States. His foundational papers on the topic were published in the 1970s, when Dr. Allison started a tradition of giving workshops on the diagnosis and treatment of Dissociative Identity Disorders at meetings of the American Psychiatric Association. Discoveries and conclusions of the last 20 years are featured in the current edition of this work. Certain tentative opinions on some borderline or parapsychological issues followed from Dr. Allison's research. Since the topic is, even today, not widely studied in clinical and therapeutic medicine, Allison's account of how he became interested, and of early vicissitudes and later conflicts, deserves careful attention by borderline investigators. There is a fairly standard aetiology to be discovered in most highly developed cases of multiple personality disorder. Very early trauma and continued conflict within the family would surprise no student of cases. Allison is most impressed by and offers detailed discussion of cases that are not typical. Some of those even suggest spirit possession or successive life histories of the same "essential" person. Dr. Allison, himself, does not endorse either concept. He notes, however, that therapeutic methods suggestive of exorcism have been used, and sometimes were found effective in his own experience. The recommended attitude is pragmatic. If a certain procedure helps the patient recover unity of personality, it may be used before any conclusive interpretation of the forces or mechanisms involved. Some of the methods of exorcism are like those of monotheistic religions while others are more animistic in character. At one point in his career, Allison was investigated by the medical staff of an institution with which he was associated, but any suspicion of unprofessional practice was judged to be unwarranted and totally without merit. The pragmatic viewpoint 296 Book Reviews prevailed, and successful therapy was accepted on its own terms by a majority of the senior staff. To the layman, the number of split-off or dissociated "personalities" in a given case is certainly surprising. Allison finds that once a major division has occurred, new temporary self systems may emerge or be constructed whenever the individual faces new stressful situations. In some cases, what emerges defies all expectations and may be bizarre or even terribly dangerous. For example, a small frail woman may become capable of combat with a few powerful security people; or again, a devoted puritan may become unspeakably vulgar in language and behavior; or a sober person may behave like one deeply intoxicated. Changes in the opposite direction also occur. Allison finds that some dissociated systems may be malevolent or persecutorial, dangerous even to the single organism, while others are benevolent and may play a constructive role. In certain cases, Allison thinks, it is correct to postulate an Inner Self Helper. Some factor in the dissociated personality will emerge to serve as the most effective therapist. This idea came from one of Allison's patients, whose recorded discussions seemed like a theatrical situation involving several actors, one of which was the Inner Self Helper, or prime essence of the patient herself. This has become a pivotal concept in Dr. Allison's theory of the psychotherapeutic process. Robert F Creegan . Philosophy The University at Albany, SUNY Albany, NY 12222 Earth Under Fire: Humanity's Survival of the Apocalypse by Paul LaViolette. Schenectady, NY: Starlane Publications, 1997. 360 pp. $25.95, (c). ISBN 0-9642025-1-4, starecode.aol.com. Breakthrough understandings of reality axiomatically are derived from new perspectives-fresh mental frameworks that like the lens of a pair of eyeglasses allow the wearer to see the external world with a whole new degree of clarity. Most often, these new metaphors come from intellectual structures derived from some discipline other than the one being studied. Machine age imagery, for example, became a common product of the industrial age, providing the architecture for the development of social organizational systems, economic theories, and even extraordinary new areas such as molecular nanotechnology. The principles of quantum physics are now working their way into business theory, and the understandings of the science of complexity are being explored as training metaphors for the military. Although great new understandings have been derived from this approach, Book Reviews 297 at the highest level, there is an intrinsic shortcoming to this mental mechanism: It is unidimensional and essentially linear. It provides only a single intellectual structure for making sense out of a puzzling new situation. It is not multidimensional.. .not a systems approach. This is important, because unlike the structure and process of modem science of the last couple of centuries-one only ventured into the sacred space of other disciplines at the risk of certain professional assault-the future will evolve at the intersection of diverse disciplines. The trend is clear: The indicators abound, but it seems it will be some time before physicists, chemists, and biologists see themselves as colleagues observing the same reality from different perspectives, rather than warriors in the service of defending their own hierarchical scientific fiefdoms. Earth Under Fire is a significant though perhaps flawed attempt to weave together the threads of a number of disciplines into a fabric that provides a picture of a possible near-term future for the earth-one that is laden with disaster. If he is right, Paul LaViolette has done us all a profound service. If he is wrong, well, this is still a beneficial work, for it certainly pushes out our horizons and further provides a very practical example of how difficult it is to know something about a lot of things and make them all work together seamlessly. This is particularly so with disciplines as diverse as astronomy, geology, paleontology, climatology, astrology, and messages from the tarot. This list may be enough for many potential readers to summarily discount the book, particularly the last two sources. But first consider the underlying hypothesis; it might get your attention. The basic notion of this book is that there is evidence from various sources that suggests that every 26,000 years (plus or minus 3,000), a huge blast emanates from the center of our galaxy, sending our way a high intensity "superwave" shower of cosmic rays followed by a giant cloud of cosmic dust. Sometimes this happens on a 13,000-year cycle. The cloud sets up an abrupt change in global weather that generally wipes most things out (i.e., human and animal life). This might lead you to ask when the next cycle is due. Well, any day now, LaViolette suggests. For me at least, that was reason enough to work my way through the book. Think of it as insurance; you never know. I was not disappointed. Not yet convinced, perhaps, but certainly provoked. In my business of professionally thinking about the future, this would be called an "early indicator" of a low probability, high impact "wild card" event. It might not turn into anything, but if it does, it would be big. If the above hasn't piqued your attention, then probably nothing more that I could say about LaViolette's argument and its shortcomings would persuade you to dip into this volume. But if you are one who is now intrigued, then a couple of additional comments are in order. There are as many perspectives through which to view the future as there are scientific disciplines. If you add in nonscientific sources such as astrology and elements of tarot cards, then immediately the rigor that one assumes would accompany a science-based argument quickly begins to erode. In this book, that 298 Book Reviews problem shows up in the first few pages. The zodiac, LaViolette contends, contains a couple of important explicit messages from our ancestors. Starting with Taurus, each of the signs in sequence stands for a specific component of subquantum creation. The key to how everything physical was created is apparently in the night sky. Ah, but there is a problem. A couple of the signs are out of order, we are told, and if you rearrange them "correctly," then they tell the right story. A bit of a reach, it seemed to me. That same kind of logic continues with the assertion that "these six pairs of celestial bodies map out a temperature gradient extending from Leo (the Sun) at the warmest end to Aquarius and Capricorn (Saturn) at the coolest end." But there is another problem. "Although the Moon is cooler than Mercury, it is, by far, much brighter, so this minor exception to the rule may be overlooked," writes the author. Then the zodiacal message from the past also supposedly references the center of our galaxy as the source of regular extraordinary explosions because (among other things) Sagittarius is shooting his arrow straight toward the point that is the center of the Milky Way, our galactic home. Yes, but it really does not point directly at the center-kind of in the general direction of the hub of our spiral neighborhood. It would have been closer 18,200 years ago, but how does one even know that there is any symbolism in the direction that the archer is pointing his bow? But it gets better-really. The geology and climate chapters are quite compelling. Careful assessments of beryllium- 10 deposits in ice corings in Antarctica (presumed to be the product of cosmic ray showers) show that the earth was washed near the end of the last ice age and on several earlier occasionstimes that correlate with other indicators of a superwave event. LaViolette does an admirable job of linking various ancient myths with physical and geological indicators that suggest that the solar system has experienced major conflagrations from solar outbursts throughout its past. He relates these possible events with historical changes in the meltwater discharge from the Mississippi River by looking at cores in the Gulf of Mexico. There are other signs. The rates of disappearance of large and small mammals in North America over the eons line up. Destructive floods called glacier bursts or glacier floods also show up at the same times. The rather famous Siberian mammoths that died and froze so suddenly that they still had grass in their mouths and stomachs were not enclosed within ice, but within frozen silt, "or in other words, within glacial flood sediments." Because grass cannot grow in that part of the world now because of the cold climate, it is fair to posit that the climate changed suddenly. This theory has been offered as the basic evidentiary argument for magnetic pole shifts, when the earth's crust becomes temporarily unclutched from its molten core, but it can also be supported by the notion of glacier floods. It is only in recent years that a great deal about our geological, geographics, Book Reviews 299 ice corings. Rapid change climate, for example, did not seem like a realistic possibility until a decade ago. It is now, therefore, provocative to think about other events signaled by minute deposits that remain after years. I do not pretend to understand, let alone believe in, astrology. I certainly do not know anything about the tarot. But that does not keep me from believing that someday we might learn something that puts these perennial pursuits in a more logical light. In the meantime, setting aside those aspects of the book, if one objectively looks at all of the evidence that LaViolette has collected from the traditional sciences, it is possible to come away saying, "This might be possible." That puts the future in a whole new light. This is a book that is both fantastic and reasonable, soaring and detailed, and theoretical and practical. It promotes one of the biggest ideas in quite some time and therefore merits critical study by others in many disciplines. John L. Petersen is president of The Arlington Institute, a Washington, DCarea research institute, and author of Out of the Blue: Wild Cards and Other Big Future Surprises. John L. Petersen The Arlington Institute 1501 Lee Highway, Suite 204 Arlington, VA 22209-1 109 johnp @ arlingtoninstitute.org SSE News 301 Fifth Biennial SSE European Meeting Announcement and Call for Papers The Fifth Biennial SSE European Meeting will be held from Oct. 20-22, 2000, at the University of Amsterdam in the Netherlands. Dick Bierman and Chris Duif are hosting the meeting and coordinating registration. Ezio Insinna is the chairman of the program committee. Program A number of distinguished scholars have been invited to speak on the theme: Unorthodox Science: Past, Present, and Future. The following speakers have accepted at the time of publication: S. J. Doorman (Dutch Skeptical Society): On the Future of Unorthodox Science. F. H. van Lunteren (University Utrecht, Netherlands): On the History of Unorthodox Science. W. Peschka (Germany): Kinetobaric Efft zts and B ioinformational Transfer. Pyatnitsky (Moscow Institute for High Temp., Russia): Effects of Consciousness on the Structure of Water. S. E. Shnoll (Moscow State University, Russia): Realization of Discrete States during Fluctuations in Macroscopic Processes. R. Sheldrake (UK): Animal Communication. R. van Wijk (University Utrecht, Netherlands): High Dilutions Research. K. Zioutas (University Thessalonike, Greece): Evidence of Dark Matter from Biological Observations. Further information on the program will be announced on the SSE website (www.scientificexploration.org/meetings/euro5 ,html) Call for Papers All interested European SSE members are encouraged to submit abstracts for short contributed presentations to the program committee consisting of Ezio Insinna, Brenda Dunne, and Dick Bierman. Abstracts no longer than 500 words must be sent before August 15, preferably by e-mail to emi2Qworldnet.fr. Address: Ezio Insinna, 18 allee des Freres Lumiere, 77600 Bussy Saint Georges, France. 302 SSE News LOGISTICAL INFORMATION Traveling The International Schiphol Airport can be reached easily from almost everywhere. Train connections from the airport to the city (Central Station) are running every 15 minutes. Travel time from the airport to Central Station is about 15 minutes. If you are traveling by automobile, please note that there is no free parking around the conference location. Check with your hotel or the local organizer Dick Bierman about how to arrange for parking. Hotels Main Hotel (-$90): Lancaster Hotel, Plantage Middenlaan 22, Amsterdam, the Netherlands. Phone: + 31 (0) 20 5356888. Email: [email protected]. Mention group code UVA-SSE when making reservations. Budget hotel, backpackerstyle (-$45): Arena Budget Hotels, Gravezandestraat 51, 1092 AA Amsterdam, the Netherlands. Phone: +31 (0) 20 6947444. Email: [email protected]. Information on other smaller hotels within walking distance will be provided on the website. The hotels are situated in a relatively green area near the Zoo and can be reached by taking either a cab from the Central Station or tram nr. 9 (get off at the Zoo for Lancaster or Tropenmuseum for Arena). Social Activities The University of Amsterdam is situated near the center of the city and the program will be set up in such a way that attendees have opportunities to enjoy the cheerful atmosphere and picturesque canals in Amsterdam. A reception and registration will take place on Thursday evening (October 19) at a nearby Caf6 (to be announced), all within a two-minute walk from the Lancaster Hotel. The Society banquet is planned for Saturday evening at the Zoo restaurant with a splendid view over the animals and gardens. The banquet will be preceded by a trip through the Amsterdam canals. The dress code is informal. Sightseeing Friday evening is free for exploring the city. Sunday afternoon might be used to visit the museums, which have large collections of Rembrandt and van Gogh paintings. The Anne Frank house is open for visitors on Sunday afternoon. For more information on local arrangements contact: Dick Bierman, email: [email protected], fax: +31-20-6391656. Registration Copy this registration form and send it to: Dick Bierman University of Amsterdam Roetersstraat 15 1018 WB Amsterdam The Netherlands Or: Register using the web registration form at our website. SSE News 303 Registration Form European SSE 2000 Conference Contact Information Name: Affiliation for Badge: Street Address: Phone: E-mail: Registration Fee Select Registration Type: Early Fee (before Sept. 1): $85 / Dfl. 190,Late Fee (after Sept. 1): $100 / Dfl. 220,Daily Registration (1 day): $40 / Dfl. 100,Daily Registration (2 days): $75 / Dfl. 170,50% DISCOUNT. Check this if you are an enrolled student. Lunches and Banquet On Site Lunches ($20 for 3 days) Cl Friday Night Banquet ($35) (includes trip through canals) Check this if you are a vegetarian Payment Method Total amount (the sum of the registration fee, lunches, and banquet): Select Payment Method: Credit Card CardNumber: Signature: Check (make check payable to Euro SSE 2000) U $ - - - Expiration Date 1 Bank transfer to PostBank Account 8417584 of Euro SSE2000, the Netherlands Please mail registration form to Dr. Dick J. Bierman, SSE 2000, University of Amsterdam, Roetersstraat 15, 1018 WB Amsterdam, The Netherlands Society for Scientific Exploration Officers Prof. Peter A. Sturrock, President Varian 302 Stanford University Stanford, CA 94305-4060 Ms. Brenda Dunne Executive Vice President for Education C131, School of Engineering &Applied Science, Princeton University, Princeton, NJ 08544-5263 Prof. Robert Jahn, Vice President D334, School of Engineering & Applied Science Princeton University Princeton, NJ 08544-5263 Prof. L. W. Fredrick, Secretary Department of Astronomy P. 0.Box 38 18 University of Virginia Charlottesville, VA 22903-08 18 Prof. Charles R. Tolbert, Treasurer Department of Astronomy P. 0.Box 3818 University of Virginia Charlottesville. VA 22903-08 18 Council Dr. Marsha Adams 1100 Bear Gulch Rd. Woodside, CA 94062 Prof. Henry H. Bauer Chemistry, 234 Lane Hall VPI & su Blacksburg, VA 24061-0247 Dr. Stephan Baumann fMRI Project Director Psychology Software Tools, Inc. 2050 Ardmore Blvd. Pittsburgh, PA 15221 Prof. John Bockris Department of Chemistry Texas A&M University College Station, TX 77843 Dr. John S. Derr Albq. Seismology Center Albuquerque, NM 871 15 Dr. Roger D. Nelson C 131, Engineering Quad. Princeton University Princeton, NJ 08544 Dr. Harold E. Puthoff Institute for Advanced Studies-Austin 4030 W. Braker Ln., Suite 300 Austin, TX 78759-5329 Dr. Beverly Rubik Institute for Frontier Science 6114 LaSalle Avenue, #605 Oakland, CA 946 11 Dr. Robert M. Wood 1727 Candlestick Ln. Newport Beach, CA 92660 Journal of Scientific Exploration, Vol. 14, No. 3, p. 303, 2000 0892-33 10/00 O 2000 Society for Scientific Exploration TRIBUTE The current issue of our Journal, Volume 14, Number 3, is the first to be compiled under the direction of our new Editor, Professor Henry Bauer. The Council is most grateful to Henry for taking on this responsibility and we wish him well. However, it is also appropriate to reflect on the great achievement of the previous Editor-in-Chief, Dr. Bernhard M. Haisch, and the Executive Editor, Ms. Marsha Sims. They served as editors for so long that it is difficult to remember that there was an editor before them, but we are indebted also to Professor Ronald A. Howard who was our first editor, negotiating a contractual relationship with Pergamon Press and bringing the Journal into existence in 1987. Bernie took over the editorship of the Journal in 1988, and Marsha joined him in the editorial office in 1990. The relationship with Pergamon Press ended in 1992, when Allen Press took over. That was a critical juncture for the future of the Journal, and Bernie and Marsha set their sights high: They decided to publish four issues per year rather than two, and they aimed to expand both the length of each issue and the subscription base. They have had great success on all fronts: The last complete volume (Volume 13) totaled 724 pages, and the number of independent subscribers grew to 750 in 1999. These changes came about with no reduction in the quality of the Journal. Bernie instituted an Advisory Board that now has 18 members, and exercised tight control over the scientific standard of papers published in the Journal. It is hard to separate the growth of the Society from the growth of the Journal. This is a symbiotic relationship. In the last 10 years, the Society membership has grown from 303 to 729, comprising 371 associates, 260 full members, 44 emeritus members, and 54 student members. We now have members in over 40 countries scattered over the globe. This growth of the Society is due in no small measure to Bernie's and Marsha's vision and effort. We are also indebted to Bernie and Marsha that they instituted our current smooth and productive relationship with Allen Press, a relationship that is now even more important to the Society, because Allen Press will be handling more of our business, both for the Journal and for the Society. Our efficient and cooperative friends at Allen Press also deserve our thanks. Now that they have handed over the editorship and the management of the Journal, Bernie and Marsha may focus their energies on another exciting new enterprise. Bernie has created, and now directs, the California Institute of Physics and Astrophysics, where Marsha now serves as executive administrator. We wish them well and we hope that some of the results of their activities will be published in due course in the Journal of Scientific Exploration. Peter A. Sturrock Journal of Scientific Exploration, Vol. 14, No. 3, pp. 304-305,2000 0892-3310100 O 2000 Society for Scientific Exploration EDITORIAL I feel honored to have been asked to serve as Editor-in-Chief. The task will be made possible by many people who have agreed to help in various ways and to whom I am enormously grateful: Stephen Braude, Dean Brown, Dean Radin, and Mark Rodeghier as Associate Editors; Michael Epstein, who has taken on book-review management and also serves as Associate Editor; David Moncrief, so assiduous in finding books worthy of review and in finding people to review them. Jill Franklin, our initial Managing Editor at Allen Press, has actually done most of the work for which I will be given credit. Journals often signal changes in editorships with some striking change in design to signal a revivification. I recall that with Chemical & Engineering News, with Science, and more recently, with American Scholar. The Journal of Scientific Exploration, however, stands in no need of revivification. My hope as I become Editor-in-Chief is that no change will be noticed in the substantive, physically attractive periodical that Bernie Haisch and Marsha Sims nurtured for so long; indeed, one might say created. So, in becoming Editor, I call for no new direction but rather recall our intellectual purpose: to provide for intellectually responsible discussion of scientifically anomalous topics not generally treated in mainstream scientific journals. One thinks at once of such subjects as UFOs, psychic phenomena, and cryptozoology. But we have equally provided a welcomed forum for unorthodoxies quite within the mainstream disciplines. The Society's Dinsdale Award has been accepted by Kilmer McCully, Halton Arp, William Corliss, Helmut Schmidt, and Ian Stevenson, so by any reckoning, the Society has honored work within the mainstream at least as often as work supposedly beyond the scientific pale. Mainstream orthodoxy routinely resists novelties that later become accepted. Throughout the 20th century there are examples: Bretz's Spokane flood, McClintock's recognition of "jumping genes", Mitchell's insights into biological energy mechanisms, Woese's Archaea, and McCully's homocysteine. Only late in the 20th century did science reluctantly grant that acupuncture can have some analgesic effect, that ball lightning exists, that the kraken is not myth but the real giant squid, that it is not foolish to look for intelligent life outside the Earth, that 5000-year-old megaliths incorporate substantial knowledge of astronomy, that human beings inhabited the Americas long before the days of the Clovis culture, and that living systems can sense not only electrical but also magnetic fields. Indeed, it may well be that the suppression of unorthodox views in science is on the increase rather than in decline. In Prometheus Bound (1994), John Ziman has outlined how science changed during the 20th century: traditionally (since perhaps the 17th century) a relatively disinterested knowledge-seeking activity, science progressively be- Editorial 305 came handmaiden to industry and government, and its direction of research is increasingly influenced by vested interests and self-interested bureaucracies, including bureaucracies supposedly established to promote good science such as the National Academies, the National Science Foundation, and the National Institutes of Health. Parkinson's Law, it may be, applies to science as to other human activities: no sooner has an organization become successfully established than it is by that token already an obsolescent nuisance. Even recently accepted theories may quickly become too dogmatically held. I had thought that plate tectonics, for example, which just a few decades ago justified Wegener's heresy, remained the generally accepted scientific wisdom, but to my surprise, there exists a group of dissenting geologists who publish a newsletter with discussions of the geological phenomena for which plate tectonics does not provide a satisfactory explanation. David Pratt reviews the situation in this issue of the Journal of Scientific Exploration. Again, it is less than 2 decades since the discovery was hailed of HIV as the cause of AIDS, yet that belief too may be coming apart at the seams after South Africa's President Mbeki challenged the scientific orthodoxy to engage in genuine discourse with the dissidents. Yet another contemporary instance: a book just published makes a compelling case that cold fusion is a real phenomenon, albeit its present name may be misleading. The Society for Scientific Exploration, then, is performing a service to science more immediate and more connected to the mainstream than was envisaged when the society was founded. At the same time, we do not neglect the more heretical subjects. In this issue, Suitbert Ertel reemphasizes how potent is the evidence for the quasi-astrological Mars effect. In a future issue, a major study of the replicability of mind-machine interactions will offer provocative food for thought. So the Journal and the Society fill a unique niche-a forum for disciplined investigative discussion of what may become accepted science later in the 21st century and beyond. Henry Bauer Journal of Scientific Exploration, Vol. 14, No. 3, pp. 307-352,2000 0892-3310100 O 2000 Society for Scientific Exploration Plate Tectonics: A Paradigm Under Threat Dual en Bergselaan 68, 2565 AG The Hague, The Netherlands dp5 @ compuserve.com Abstract-This paper looks at the challenges confronting plate tectonicsthe ruling paradigm in the earth sciences. The classical model of thin lithospheric plates moving over a global asthenosphere is shown to be implausible. Evidence is presented that appears to contradict continental drift, seafloor spreading, and subduction, as well as the claim that the oceanic crust is relatively young. The problems posed by vertical tectonic movements are reviewed, including evidence for large areas of submerged continental crust in today's oceans. It is concluded that the fundamental tenets of plate tectonics might be wrong. Keywords: plate tectonics - continental roots -age of seafloor - vertical tectonics - surge tectonics. Introduction The idea of large-scale continental drift has been around for some 200 years, but the first detailed theory was proposed by Alfred Wegener in 1912. It met with widespread rejection, largely because the mechanism he suggested was inadequate: the continents supposedly plowed slowly through the denser oceanic crust under the influence of gravitational and rotational forces. Interest was revived in the early 1950s with the rise of the new science of paleomagnetism, which seemed to provide strong support for continental drift. In the early 1960s, new data from ocean exploration led to the idea of seafloor spreading. A few years later, these and other concepts were synthesized into the model of plate tectonics, which was originally called "the new global tectonics." According to the orthodox model of plate tectonics, the earth's outer shell, or lithosphere, is divided into a number of large, rigid plates that move over a soft layer of the mantle known as the "asthenosphere" and interact at their boundaries, where they converge, diverge, or slide past one another. Such interactions are believed to be responsible for most of the seismic and volcanic activity of the earth. Plates cause mountains to rise where they push together, and continents to fracture and oceans to form where they rift apart. The continents, sitting passively on the backs of the plates, drift with them, at the rate of a few centimeters per year. At the end of the Permian, some 250 million years ago, all the present continents are said to have been gathered together in a single supercontinent, Pangaea, consisting of two major landmasses: Laurasia in l 307 308 David Pratt the north, and Gondwanaland in the south. Pangaea is widely believed to have started fragmenting in the early Jurassic-although this is sometimes said to have begun earlier, in the Triassic, or even as late as the Cretaceous-resulting in the configuration of oceans and continents observed today. It has been said that "a hypothesis that is appealing for its unity or simplicity acts as a filter, accepting reinforcement with ease but tending to reject evidence that does not seem to fit" (Grad, 197 l , p. 636). Meyerhoff and Meyerhoff (1974b, p. 411) argued that this is "an admirable description of what has happened in the field of earth dynamics, where one hypothesis-the new global tectonics-has been permitted to override and overrule all other hypotheses." Nitecki et al. (1978) reported that in 1961 only 27% of western geologists accepted plate tectonics, but that during the mid-1960s a "chain reaction" took place, and by 1977 it was embraced by as many as 87%. Some proponents of plate tectonics have admitted that a bandwagon atmosphere developed and that data that did not fit into the model were not given sufficient consideration (e.g., Wyllie, 1976), resulting in "a somewhat disturbing dogmatism" (Dott and Batten, 198 1, p. 15 1). McGeary and Plummer (1 998, p. 97) acknowledge that "geologists, like other people, are susceptible to fads." Maxwell (1974) stated that many earth-science papers were concerned with demonstrating that some particular feature or process may be explained by plate tectonics, but that such papers were of limited value in any unbiased assessment of the scientific validity of the hypothesis. Van Andel (1984) conceded that plate tectonics had serious flaws and that the need for a growing number of ad hoc modifications cast doubt on its claim to be the ultimate unifying global theory. Lowman (1 992a) argued that geology has largely become "a bland mixture of descriptive research and interpretive papers in which the interpretation is a facile cookbook application of plate-tectonics concepts... used as confidently as trigonometric functions" (p. 3). Lyttleton and Bondi (1 992) held that the difficulties facing plate tectonics and the lack of study of alternative explanations for seemingly supportive evidence reduced the plausibility of the theory. Saul1 (1986) pointed out that no global tectonic model should ever be considered definitive, because geological and geophysical observations are nearly always open to alternative explanations. He also stated that even if plate tectonics were false, it would be difficult to refute and replace, for the following reasons: the processes supposed to be responsible for plate dynamics are rooted in regions of the earth so poorly known that it is hard to prove or disprove any particular model of them; the hard core of belief in plate tectonics is protected from direct assault by auxiliary hypotheses that are still being generated; and the plate model is so widely believed to be correct that it is difficult to get alternative interpretations published in the scientific literature. When plate tectonics was first elaborated in the 1960s, less than 0.0001% of the deep ocean had been explored and less than 20% of the land area had been mapped in meaningful detail. Even by the mid- 1990s, only about 3%-5% of Plate Tectonics: A Paradigm Under Threat 309 the deep ocean basins had been explored in any kind of detail, and not much more than 25%-30% of the land area could be said to be truly known (Meyerhoff et al., 1996a). Scientific understanding of the earth's surface features is clearly still in its infancy, to say nothing of the earth's interior. Beloussov (1980, 1990) held that plate tectonics was a premature generalization of still very inadequate data on the structure of the ocean floor and had proven to be far removed from geological reality. He wrote: It is ...quite understandable that attempts to employ this conception to explain concrete structural situations in a local rather than a global scale lead to increasingly complicated schemes in which it is suggested that local axes of spreading develop here and there, that they shift their position, die out, and reappear, that the rate of spreading alters repeatedly and often ceases altogether, and that lithospheric plates are broken up into an even greater number of secondary and tertiary plates. All these schemes are characterised by a complete absence of logic, and of patterns of any kind. The impression is given that certain rules of the game have been invented, and that the aim is to fit reality into these rules somehow or other (1 980, p. 303). Criticism of plate tectonics has increased in line with the growing number of observational anomalies. This paper outlines some of the main problems facing the theory. Plates in Motion? According to the classical model of plate tectonics, lithospheric plates creep over a relatively plastic layer of partly molten rock known as the "asthenosphere" (or low-velocity zone). According to a modern geological textbook (McGeary and Plummer, 1998), the lithosphere, which comprises the earth's crust and uppermost mantle, averages about 70 km thick beneath oceans and is at least 125 km thick beneath continents, while the asthenosphere extends to a depth of perhaps 200 km. It points out that some geologists think that the lithosphere beneath continents is at least 250 km thick. Seismic tomography, which produces three-dimensional images of the earth's interior, appears to show that the oldest parts of the continents have deep roots extending to depths of 400-600 km and that the asthenosphere is essentially absent beneath them (Figure 1). McGeary and Plummer (1998) say that these findings cast doubt on the original, simple lithosphere-asthenosphere model of plate behavior. They do not, however, consider any alternatives. Despite the compelling seismotomographic evidence for deep continental roots (Dziewonski and Anderson, 1984; Dziewonski and Woodhouse, 1987; Grand, 1987; Lerner-Lam, 1988; Forte, Dziewonski, and O'Connell, 1995; Gossler and Kind, 1996), some plate tectonicists have suggested that we just happen to live at a time when the continents have drifted over colder mantle (Anderson, Tanimoto, and Zhang, 1992), or that continental roots are really no 310 David Pratt NORTH AMERICAN CRATON [EUROPE W JUAN DE A FUCf RIDGE M - TA T I AL NI D C RIDGE I ICELAND AND E A' Fig. 1. Seismotomographic cross-section showing velocity structure across the North American craton and North Atlantic Ocean. High-velocity (colder) lithosphere, shown in dark tones, underlies the Canadian shield to depths of 250-500 km. (Reprinted with permission from Grand, 1987. Copyright by the American Geophysical Union.) mantle material beneath them, giving the illusion of much deeper roots (Polet and Anderson, 1995). However, evidence from seismic-velocity, heat-flow, and gravity studies has been building up for several decades, showing that ancient continental shields have very deep roots and that the low-velocity asthenosphere is very thin or absent beneath them (e.g., Jordan, 1975, 1978; MacDonald, 1963; Pollack and Chapman, 1977). Seismic tomography has merely reinforced the message that continental cratons, particularly those of Archean and Early Proterozoic age, are "welded" to the underlying mantle, and that the concept of thin (less than 250 km thick) lithospheric plates moving thousands of kilometers over a global asthenosphere is unrealistic. Nevertheless, many textbooks continue to propagate the simplistic lithosphere-asthenosphere model and fail to give the slightest indication that it faces any problems (e.g., McLeish, 1992; Skinner and Porter, 1995; Wicander and Monroe, 1999). Geophysical data show that, far from the asthenosphere being a continuous layer, there are disconnected lenses (asthenolenses), which are observed only in regions of tectonic activation and high heat flow. Although surface-wave observations suggested that the asthenosphere was universally present beneath the oceans, detailed seismic studies show that here, too, there are only asthenospheric lenses. Seismic research has revealed complicated zoning and inhomogeneity in the upper mantle and the alternation of layers with higher and lower velocities and layers of different quality. Individual low-velocity layers are bedded at different depths in different regions and do not compose a single layer. This renders the very concept of the lithosphere ambiguous, at least that of its base. Indeed, the definition of the lithosphere and astheno- Plate Tectonics: A Paradigm Under Threat 311 sphere has become increasingly blurred with time (Pavlenkova, 1990, 1995, 1996). Thus, the lithosphere has a highly complex and irregular structure. Far from being homogeneous, "plates" are actually "a megabreccia, a 'pudding' of inhomogeneities whose nature, size, and properties vary widely" (Chekunov, Gordienko, and Guterman, 1990, p. 404). The crust and uppermost mantle are divided by faults into a mosaic of separate, jostling blocks of different shapes and sizes, generally a few hundred kilometers across, and of varying internal structure and strength. Pavlenkova (1990, p. 78) concludes: "This means that the movement of lithospheric plates over long distances, as single rigid bodies, is hardly possible. Moreover, if we take into account the absence of the asthenosphere as a single continuous zone, then this movement seems utterly impossible." She states that this is further confirmed by the strong evidence that regional geological features, too, are connected with deep (more than 400 km) inhomogeneities and that these connections remain stable during long periods of geologic time; considerable movement between the lithosphere and asthenosphere would detach near-surface structures from their deep mantle roots. Plate tectonicists who accept the evidence for deep continental roots have proposed that plates may extend to and glide along the 400-km, or even 670km, seismic discontinuity (Jordan, 1975, 1978, 1979; Seyfert, 1998). Jordan, for instance, suggested that the oceanic lithosphere moves on the classical low-velocity zone while the continental lithosphere moves along the 400-km discontinuity. However, there is no certainty that a superplastic zone exists at this discontinuity, and no evidence has been found of a shear zone connecting the two decoupling layers along the trailing edge of continents (Lowman, 1985). Moreover, even under the oceans, there appears to be no continuous asthenosphere. Finally, the movement of such thick "plates" poses an even greater problem than that of thin lithosphericic plates. The driving force of plate movements was initially claimed to be mantledeep convection currents welling up beneath midocean ridges, with downwelling occurring beneath ocean trenches. Since the existence of layering in the mantle was considered to render whole-mantle convection unlikely, twolayer convection models were also proposed. Jeffreys (1 974) argued that convection cannot take place because it is a self-damping process, as described by the Lomnitz law. Plate tectonicists expected seismic tomography to provide clear evidence of a well-organized convection-cell pattern, but it has actually provided strong evidence against the existence of large, plate-propelling convection cells in the upper mantle (Anderson, Tanimoto, and Zhang, 1992). Many geologists now think that mantle convection is a result of plate motion rather than its cause and that it is shallow rather than mantle deep (McGeary and Plummer, 1998). The favored plate-driving mechanisms at present are "ridge push" and "slab 312 David Pratt the dominant mechanism and refers to the gravitational subsidence of subducted slabs. However, it will not work for plates that are largely continental or that have leading edges that are continental, because continental crust cannot be bodily subducted due to its low density, and it seems utterly unrealistic to imagine that ridge push from the Mid-Atlantic Ridge alone could move the 120"-wide Eurasian plate (Lowman, 1986). Moreover, evidence for the longterm weakness of large rock masses casts doubt on the idea that edge forces can be transmitted from one margin of a "plate" to its interior or opposite margin (Keith, 1993). Thirteen major plates are currently recognized, ranging in size from about 400 by 2,500 km to 10,000 by 10,000 km, together with a proliferating number of microplates (over 100 so far). Van Andel (1998) writes: Where plate boundaries adjoin continents, matters often become very complex and have demanded an ever denser thicket of ad hoc modifications and amendments to the theory and practice of plate tectonics in the form of microplates, obscure plate boundaries, and exotic terranes. A good example is the Mediterranean, where the collisions between Africa and a swarm of microcontinents have produced a tectonic nightmare that is far from resolved. More disturbingly, some of the present plate boundaries, particularly in the eastern Mediterranean, appear to be so diffuse and so anomalous that they cannot be compared to the three types of plate boundaries of the basic theory. Plate boundaries are identified and defined mainly on the basis of earthquake and volcanic activity. The close correspondence between plate edges and belts of earthquakes and volcanoes is therefore to be expected and can hardly be regarded as one of the "successes" of plate tectonics (McGeary and Plummer, 1998). Moreover, the simple pattern of earthquakes around the Pacific Basin on which plate tectonics models have hitherto been based has been seriously undermined by more recent studies showing a surprisingly large number of earthquakes in deep-sea regions previously thought to be aseismic (Storetvedt, 1997). Another major problem is that several "plate boundaries" are purely theoretical and appear to be nonexistent, including the northwest Pacific boundary of the Pacific, North American, and Eurasian plates, the southern boundary of the Philippine plate, part of the southern boundary of the Pacific plate, and most of the northern and southern boundaries of the South American plate (Stanley, 1989). Continental Drift Geological field mapping provides evidence for horizontal crustal movements of up to several hundred kilometers (Jeffreys, 1976). Plate tectonics, however, claims that continents have moved up to 7,000 km or more since the alleged breakup of Pangaea. Measurements using space-geodetic techniques-very long baseline interferometry, satellite laser-ranging, and the global positioning system-have been hailed by some workers as having Plate Tectonics: A Paradigm Under Threat 313 but do not provide evidence for plate motions of the kind predicted by plate tectonics unless the relative motions predicted among all plates are observed. However, many of the results have shown no definite pattern and have been confusing and contradictory, giving rise to a variety of ad hoc hypotheses (Fallon and Dillinger, 1992; Gordon and Stein, 1992; Smith et al., 1994). Japan and North America appear, as predicted, to be approaching each other, but distances from the Central South American Andes to Japan or Hawaii are more or less constant, whereas plate tectonics predicts significant separation (Storetvedt, 1997). Trans-Atlantic drift has not been demonstrated, because baselines within North America and western Europe have failed to establish that the plates are moving as rigid units; they suggest in fact significant intraplate deformation (James, 1994; Lowman, 1992b). Space-geodetic measurements to date have therefore not confirmed plate tectonics. Moreover, they are open to alternative explanations (e.g., Carey, 1994; Meyerhoff et al., 1996a; Storetvedt, 1997). It is clearly a hazardous exercise to extrapolate present crustal movements tens or hundreds of millions of years into the past or future. Indeed, geodetic surveys across "rift" zones (e.g., in Iceland and East Africa) have failed to detect any consistent and systematic widening as postulated by plate tectonics (Keith, 1993). Fits and Misfits A "compelling" piece of evidence that all the continents were once united in one large landmass is said to be the fact that they can be fitted together like pieces of a jigsaw puzzle. Many reconstructions have been attempted (e.g., Barron, Harrison, and Hay, 1978; Bullard, Everett, and Smith, 1965; Dietz and Holden, 1970; Nafe and Drake, 1969; Scotese, Gahagan, and Larson, 1988; Smith and Hallam, 1970; Smith, Hurley, and Briden, 1981 ;Tarling, 197 I), but none are entirely acceptable (Figures 2 and 3). In the Bullard, Everett, and Smith (1965) computer-generated fit, for example, there are a number of glaring omissions. The whole of Central America and much of southern Mexico are left out, despite the fact that extensive areas of Paleozoic and Precambrian continental rocks occur there. This region of some 2,100,000 km2 overlaps South America in a region consisting of a craton at least 2 billion years old. The entire West Indian archipelago has also been omitted. In fact, much of the Caribbean is underlain by ancient continental crust, and the total area involved (300,000 km 2) overlaps Africa (Meyerhoff and Hatten, 1974). The Cape Verde Islands-Senegal Basin, too, is underlain by ancient continental crust, creating an additional overlap of 800,000 km2. Several major submarine structures that appear to be of continental origin are ignored in the Bullard, Everett, and Smith (1 965) fit, including the FaeroeIceland-Greenland Ridge, Jan Mayen Ridge, Walvis Ridge, Rio Grande Rise, and the Falkland Plateau. However, the Rockall Plateau was included for the sole reason that it could be "slotted in." This fit postulates an east-west shear David Pratt Fig. 2. The Bullard fit. Overlaps and gaps between continents are shown in black. (Reprinted with permission from Bullard, Everett, and Smith, 1965. Copyright by The Royal Society.) field geology does not support either of these suppositions (Meyerhoff and Meyerhoff, 1974a). Even the celebrated fit of South America and Africa is problematic, as it is impossible to match all parts of the coastlines simultaneously; e.g., there is a gap between Guyana and Guinea (Eyles and Eyles, 1993). Like the Bullard, Everett, and Smith (1965) fit, the Smith and Hallam (1970) reconstruction of the Gondwanaland continents is based on the 500fathom depth contour. The South Orkneys and South Georgia are omitted, as is Kerguelen Island in the Indian Ocean, and there is a large gap west of Australia. Fitting India against Australia, as in other fits, leaves a corresponding gap in the western Indian Ocean (Hallam, 1976). Dietz and Holden (1970) based their fit on the 1,000-fathom (2-km) contour, but they still had to omit the Florida-Bahamas platform, ignoring the evidence that it predates the alleged commencement of drift. In many regions, the boundary between continental and oceanic crust appears to occur beneath oceanic depths of 2-4 km or more (Hallam, 1979), and in some places, the ocean-continent transition zone is several hundred kilometers wide (Van der Linden, 1977). This means that any reconstructions based on arbitrarily selected depth contours are flawed. Given the liberties that drifters have had to take to obtain the desired continen- Plate Tectonics: A Paradigm Under Threat 315 Fig. 3. Computer-derived plate tectonic map for Permian time. (Reprinted with permission from Meyerhoff, 1995. Copyright by Elsevier Science.) tal matches, their computer-generated fits may well be a case of "garbage in, garbage out" (Le Grand, 1988). The similarities of rock types and geological structures on coasts that were supposedly once juxtaposed are hailed by drifters as further evidence that the continents were once joined together. However, they rarely mention the many geological dissimilarities. For instance, western Africa and northern Brazil were supposedly once in contact, yet the structural trends of the former run north to south while those of the latter run east to west (Storetvedt, 1997). Some predrift reconstructions show peninsular India against western Antarctica, yet Permian Indian basins do not correspond geographically or in sequence to the western Australian basins (Dickins and Choi, 1997). Gregory (1 929) held that the geological resemblances of opposing Atlantic coastlines are due to the areas having belonged to the same tectonic belt, but that the differences are sufficient to show that the areas were situated in distant parts of the belt. Bucher (1933) showed that the paleontological and geological similarities between the eastern Alps and central Himalayas, 4,000 miles apart, are just as remarkable as those between the Argentine and South Africa, separated by the same distance. The approximate parallelism of the coastlines of the Atlantic Ocean may be due to the boundaries between the continents and oceans having been formed by deep faults, which tend to be grouped into parallel systems (Beloussov, 1980). Moreover, the curvature of continental contours is often so similar that many of them can be joined if they are given the necessary rotation. Lyustikh (1967) gave examples of 15 shorelines that can be fitted together quite well even though they can never have been in juxtaposition. Voisey (1 958) showed that eastern Australia fits well with eastern North America if Cape York is 316 David Pratt placed next to Florida. He pointed out that the geological and paleontological similarities are remarkable, probably due to the similar tectonic backgrounds of the two regions. Paleornagnetic Pitfalls One of the main props of continental drift is paleomagnetism-the study of the magnetism of ancient rocks and sediments. The inclination and declination of fossil magnetism can be used to infer the location of a virtual magnetic pole relative to the location of the sample in question. When virtual poles are determined from progressively older rocks from the same continent, the poles appear to wander with time. Joining the former averaged pole positions generates an apparent polar wander path. Different continents yield different polar wander paths, and from this, it has been concluded that the apparent wandering of the magnetic poles is caused by the actual wandering of the continents over the earth's surface. The possibility that there has been some degree of true polar wander-i.e., a shift of the whole earth relative to the rotation axis (the axial tilt remaining the same)-has not, however, been ruled out. That paleomagnetism can be unreliable is well established (Barron, Harrison, and Hay, 1978; Meyerhoff and Meyerhoff, 1972). For instance, paleomagnetic data imply that during the mid-Cretaceous, Azerbaijan and Japan were in the same place (Meyerhoff, 1970a)! The literature is in fact bursting with inconsistencies (Storetvedt, 1997). Paleomagnetic studies of rocks of different ages suggest a different polar wander path not only for each continent, but also for different parts of each continent. When individual paleomagnetic pole positions, rather than averaged curves, are plotted on world maps, the scatter is huge, often wider than the Atlantic. Furthermore, paleomagnetism can determine only paleolatitude, not paleolongitude. Consequently, it cannot be used to prove continental drift. Paleomagnetism is plagued with uncertainties. Merrill, McElhinny, and McFadden (1996, p. 69) state that "there are numerous pitfalls that await the unwary: first, in sorting out the primary magnetization from secondary magnetizations (acquired subsequent to formation), and second, in extrapolating the properties of the primary magnetization to those of the earth's magnetic field." The interpretation of paleomagnetic data is founded on two basic assumptions: (a) when rocks are formed, they are magnetized in the direction of the geomagnetic field existing at the time and place of their formation, and the acquired magnetization is retained in the rocks at least partially over geologic time; and (b) the geomagnetic field averaged for any period of the order of 10' years (except magnetic-reversal epochs) is a dipole field oriented along the earth's rotation axis. Both these assumptions are questionable. The gradual northward shift of paleopole "scatter ellipses" through time, and the gradual reduction in the diameters of the ellipses suggest that remanent magnetism becomes less stable with time. Rock magnetism is subject to modification by later magnetism, weathering, metamorphism, tectonic defor- Plate Tectonics: A Paradigm Under Threat 317 mation, and chemical changes. Moreover, the geomagnetic field today deviates substantially from that of a geocentric axial dipole. The magnetic axis is tilted by about 11" to the rotation axis, and on some planets, much greater offsets are found: 46.8" in the case of Neptune and 58.6" in the case of Uranus (Merrill, McElhinny, and McFadden, 1996). Nevertheless, because Earth's magnetic field undergoes significant long-term secular variation (e.g., a westward drift), it is thought that the time-averaged field will closely approximate a geocentric axial dipole. However, there is strong evidence that the geomagnetic field had long-term nondipole components in the past, though they have largely been neglected (Kent and Smethurst, 1998; Van der Voo, 1998). To test the axial nature of the geomagnetic field in the past, scientists have to use paleoclimatic data. However, several major paleoclimatic indicators, along with paleontological data, provide powerful evidence against continental-drift models, and therefore against the current interpretation of paleomagnetic data (see below). It is possible that the magnetic poles have wandered considerably with respect to the geographic poles in former times. Also, if in past geological periods, there were stable magnetic anomalies of the same intensity as the presentday East Asian anomaly (or slightly more intensive), this would render the geocentric axial dipole hypothesis invalid (Beloussov, 1990). Regional or semiglobal magnetic fields might be generated by vortexlike cells of thermalmagmatic energy, rising and falling in the earth's mantle (Pratsch, 1990). Another important factor may be magnetostriction-the alteration of the direction of magnetization by directed stress (Jeffreys, 1976; Munk and MacDonald, 1975). Some workers have shown that certain discordant paleomagnetic results that could be explained by large horizontal movements can be explained equally well by vertical block rotations and tilts and by inclination shallowing resulting from sediment compaction (Butler et al., 1989; Dickinson and Butler, 1998; Irving and Archibald, 1990; Hodych and Bijaksana, 1993). Storetvedt (1 992, 1997) has developed a model known as "global wrench tectonics" in which paleomagnetic data are explained by in situ horizontal rotations of continental blocks, together with true polar wander. The possibility that a combination of these factors could be at work simultaneously significantly undermines the use of paleomagnetism to support continental drift. Drift Versus Geology The opening of the Atlantic Ocean allegedly began in the Cretaceous by the rifting apart of the Eurasian and American plates. However, on the other side of the globe, northeastern Eurasia is joined to North America by the BeringChukotsk shelf, which is underlain by Precambrian continental crust that is continuous and unbroken from Alaska to Siberia. Geologically these regions constitute a single unit, and it is unrealistic to suppose that they were formerly divided by an ocean several thousand kilometers wide, which closed to com- 318 David Pratt pensate for the opening of the Atlantic. If a suture is absent there, one ought to be found in Eurasia or North America, but no such suture appears to exist (Beloussov, 1990; Shapiro, 1990). If Baffin Bay and the Labrador Sea had formed by Greenland and North America drifting apart, this would have produced hundreds of kilometers of lateral offset across the Nares Strait between Greenland and Ellesmere Island, but geological field studies reveal no such offset (Grant, 1980, 1992). Greenland is separated from Europe west of Spitsbergen by only 50-75 km at the 1,000-fathom depth contour, and it is joined to Europe by the continental Faeroe-Iceland-Greenland Ridge (Meyerhoff, 1974). All these facts rule out the possibility of east-west drift in the northern hemisphere. Geology indicates that there has been a direct tectonic connection between Europe and Africa across the zones of Gibraltar and Rif on the one hand, and Calabria and Sicily on the other, at least since the end of the Paleozoic, contradicting plate-tectonic claims of significant displacement between Europe and Africa during this period (Beloussov, 1990). Plate tectonicists hold widely varying opinions on the Middle East region. Some advocate the former presence of two or more plates, some postulate several microplates, others support island-arc interpretations, and a majority favor the existence of at least one suture zone that marks the location of a continent-continent collision. Kashfi (1992, p. 119) comments: Nearly all of these hypotheses are mutually exclusive. Most would cease to exist if the field data were honored. These data show that there is nothing in the geologic record to support a past separation of Arabia-Africa from the remainder of the Middle East. India supposedly detached itself from Antarctica sometime during the Mesozoic, and then drifted northeastward up to 9,000 km, over a period of up to 200 million years, until it finally collided with Asia in the mid-Tertiary, pushing up the Himalayas and the Tibetan Plateau. That Asia happened to have an indentation of approximately the correct shape and size and in exactly the right place for India to "dock" into would amount to a remarkable coincidence (Mantura, 1972). There is, however, overwhelming geological and paleontological evidence that India has been an integral part of Asia since Proterozoic or earlier time (Ahmad, 1990; Chatterjee and Hotton, 1986; Meyerhoff et al., 1991; Saxena and Gupta, 1990). There is also abundant evidence that the Tethys Sea in the region of the present Alpine-Himalayan orogenic belt was never a deep, wide ocean but rather a narrow, predominantly shallow, intracontinental seaway (Bhat, 1987; Dickins, 1987, 1994c; McKenzie, 1987; Stocklin, 1989). If the long journey of India had actually occurred, it would have been an isolated island continent for millions of years-sufficient time to have evolved a highly distinct endemic fauna. However, the Mesozoic and Tertiary faunas show no such endemism but indicate instead ~ Plate Tectonics: A Paradigm Under Threat 319 ~ I I and Antarctica (Chatterjee and Hotton, 1986). The stratigraphic, structural, and paleontological continuity of India with Asia and Arabia means that the supposed "flight of India" is no more than a flight of fancy. A striking feature of the oceans and continents today is that they are arranged antipodally: The Arctic Ocean is precisely antipodal to Antarctica; North America is exactly antipodal to the Indian Ocean; Europe and Africa are antipodal to the central area of the Pacific Ocean; Australia is antipodal to the small basin of the North Atlantic; and the South Atlantic correspondsthough less exactly-to the eastern half of Asia (Bucher, 1933; Gregory, 1899, 1901; Steers, 1950). Only 7% of the earth's surface does not obey the antipodal rule. If the continents had slowly drifted thousands of kilometers to their present positions, the antipodal arrangement of land and water would have to be regarded as purely coincidental. Harrison et al. (1 983) calculated that there is one chance in seven that this arrangement is the result of a random process. Paleoclimatology The paleoclimatic record is preserved from Proterozoic time to the present in the geographic distribution of evaporites, carbonate rocks, coals, and tillites. The locations of these paleoclimatic indicators are best explained by stable rather than shifting continents, and by periodic changes in climate, from globally warm or hot to globally cool (Meyerhoff and Meyerhoff, 1974a; Meyerhoff et al., 1996b). For instance, 95% of all evaporites-a dry-climate indicator-from the Proterozoic to the present lie in regions that now receive less than 100 cm of rainfall per year, i.e. in today's dry-wind belts. The evaporite and coal zones show a pronounced northward offset similar to today's northward offset of the thermal equator. Shifting the continents succeeds at best in explaining local or regional paleoclimatic features for a particular period and invariably fails to explain the global climate for the same period. In the Carboniferous and Permian, glaciers covered parts of Antarctica, South Africa, South America, India, and Australia. Drifters claim that this glaciation can be explained in terms of Gondwanaland, which was then situated near the South Pole. However, the Gondwanaland hypothesis defeats itself in this respect because large areas that were glaciated during this period would be removed too far inland for moist ocean-air currents to reach them. Glaciers would have formed only at its margins while the interior would have been a vast, frigid desert (Meyerhoff, 1970a; Meyerhoff and Teichert, 197 1). Shallow epicontinental seas within Pangaea could not have provided the required moisture because they would have been frozen during the winter months. This glaciation is easier to explain in terms of the continents' present positions: nearly all the continental ice centers were adjacent to or near present coastlines, or in high plateaus and/or mountain lands not far from present coasts. Drifters say that the continents have shifted little since the start of the Cenozoic (some 65 million years ago), yet this period has seen significant alterations in climatic conditions. Even since Early Pliocene time the width of the 320 David Pratt temperate zone has changed by more than 15" (1,650 km) in both the northern and southern hemispheres. The uplift of the Rocky Mountains and Tibetan Plateau appears to have been a key factor in the late Cenozoic climatic deterioration (Manabe and Broccoli, 1990; Ruddiman and Kutzbach, 1989). To decide whether past climates are compatible with the present latitudes of the regions concerned, it is clearly essential to take account of vertical crustal movements, which can cause significant changes in atmospheric and oceanic circulation patterns by altering the topography of the continents and ocean floor, and the distribution of land and sea (Brooks, 1949; Dickins, 1994a; Meyerhoff, 1 970b). Biopaleogeography Meyerhoff et al. (1 996b) showed in a detailed study that most major biogeographical boundaries, based on floral and faunal distributions, do not coincide with the partly computer-generated plate boundaries postulated by plate tectonics, Nor do the proposed movements of continents correspond with the known, or necessary, migration routes and directions of biogeographical boundaries. In most cases, the discrepancies are very large, and not even an approximate match can be claimed. The authors comment, "What is puzzling is that such major inconsistencies between plate tectonic postulates and field data, involving as they do boundaries that extend for thousands of kilometers, are permitted to stand unnoticed, unacknowledged, and unstudied" (p. 3). The known distributions of fossil organisms are more consistent with an earth model like that of today than with continental-drift models, and more migration problems are raised by joining the continents in the past than by keeping them separated (Khudoley, 1974; Meyerhoff and Meyerhoff, 1974a; Smiley, 1974, 1976, 1992; Teichert, 1974; Teichert and Meyerhoff, 1972). It is unscientific to select a few faunal identities and ignore the vastly greater number of faunal dissimilarities from different continents that were supposedly once joined. The widespread distribution of the Glossopteris flora in the southern continents is frequently claimed to support the former existence of Gondwanaland, but it is rarely pointed out that this flora has also been found in northeast Asia (Smiley, 1976). Some of the paleontological evidence appears to require the alternate emergence and submergence of land dispersal routes only after the supposed breakup of Pangaea. For example, mammal distribution indicates that there were no direct physical connections between Europe and North America during Late Cretaceous and Paleocene times but suggests a temporary connection with Europe during the Eocene (Meyerhoff and Meyerhoff, 1974a). Continental drift, on the other hand, would have resulted in an initial disconnection with no subsequent reconnection. A few drifters have recognized the need for intermittent land bridges after the supposed separation of the continents (e.g., Briggs, 1987; Tarling, 1982). Various oceanic ridges, rises, and plateaus could have served as land bridges, as many are known to have been partly above Plate Tectonics: A Paradigm Under Threat 321 water at various times in the past. It is also possible that these land bridges formed part of larger former landmasses in the present oceans (see below). Seafloor Spreading and Subduction According to the seafloor-spreading hypothesis, new oceanic lithosphere is generated at midocean ridges ("divergent plate boundaries") by the upwelling of molten material from the earth's mantle, and as the magma cools, it spreads away from the flanks of the ridges. The horizontally moving plates are said to plunge back into the mantle at ocean trenches or "subduction zones" ("convergent plate boundaries"). The melting of the descending slab is believed to give rise to the magmatic-volcanic arcs that lie adjacent to certain trenches. Seafloor Spreading ~ The ocean floor is far from having the uniform characteristics that conveyor-type spreading would imply (Keith, 1993). Although averaged surfacewave data seemed to confirm that the oceanic lithosphere was symmetrical in relation to the ridge axis and increased in thickness with distance from the axial zone, more detailed seismic research has contradicted this simple model. It has shown that the mantle is asymmetrical in relation to the midocean ridges and has a complicated mosaic structure independent of the strike of the ridge. Several low-velocity zones (asthenolenses) occur in the oceanic mantle, but it is difficult to establish any regularity between the depth of the zones and their distance from the midocean ridge (Pavlenkova, 1990). Boreholes drilled in the Atlantic, Indian, and Pacific Oceans have shown the extensive distribution of shallow-water sediments ranging from Triassic to Quaternary. The spatial distribution of shallow-water sediments and their vertical arrangement in some of the sections refute the spreading mechanism for the formation of oceanic lithosphere (Ruditch, 1990). The evidence implies that since the Jurassic, the present oceans have undergone large-amplitude subsidences, and that this occurred mosaically rather than showing a systematic relationship with distance from the ocean ridges. Younger, shallow-water sediments are often located farther from the axial zones of the ridges than older ones-the opposite of what is required by the plate tectonics model, which postulates that as newly formed oceanic lithosphere moves away from the spreading axis and cools, it gradually subsides to greater depths. Furthermore, some areas of the oceans appear to have undergone continuous subsidence, whereas others underwent alternating subsidence and elevation (Figure 4). The height of the ridge along the Romanche fracture zone in the equatorial Atlantic is 1-4 km above that expected by seafloor-spreading models. Large segments of it were close to or above sea level only 5 million years ago, and subsequent subsidence has been one order of magnitude faster than that predicted by plate tectonics (Bonatti and Chermak, 198 1). David Pratt 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 Mln Yrr Fig. 4. Vertical movements of the ocean bed during the last 160 million years: (1) according to the seafloor-spreading model, (2) the real sequence of vertical movements at the corresponding deep-sea drilling sites. The curves in the upper scaleless part of the diagram are tentative. (Reprinted from Ruditch, 1990.) along ocean ridges and fall off steadily with increasing distance from the ridge crests. Actual measurements, however, contradict this simple picture: Ridge crests show a very large scatter in heat-flow magnitudes, and there is generally little difference in thermal flux between the ridge and the rest of the ocean (Keith, 1993; Storetvedt, 1997). All parts of the Indian Ocean display a cold and rather featureless heat-flow picture except the Central Indian Basin. The broad region of intense tectonic deformation in this basin indicates that the basement has a block structure and presents a major puzzle for plate tectonics, especially since it is located in a "midplate" setting. Smoot and Meyerhoff (1995) have shown that nearly all published charts of the world's ocean floors have been drawn deliberately to reflect the predictions of the plate-tectonics hypothesis. For example, the Atlantic Ocean floor is unvaryingly shown to be dominated by a sinuous, north-south midocean ridge, flanked on either side by abyssal plains, cleft at its crest by a rift valley, and offset at more or less regular 40-to-60-km intervals by east-west-striking fracture zones. New, detailed bathymetric surveys indicate that this oversimplified portrayal of the Atlantic Basin is largely wrong, yet the most accurate charts now available are widely ignored because they do not conform to plate tectonic preconceptions. According to plate tectonics, the offset segments of "spreading" oceanic ridges should be connected by "transform fault" plate boundaries. Since the late 1960s, it has been claimed that first-motion studies in ocean fracture Plate Tectonics: A Paradigm Under Threat 323 zones provide overwhelming support for the concept of transform faults. The results of these seismic surveys, however, were never clear cut, and contradictory evidence and alternative explanations have been ignored (Storetvedt, 1997; Meyerhoff and Meyerhoff, 1974a). Instead of being continuous and approximately parallel across the full width of each ridge, ridge-transverse fracture zones tend to be discontinuous, with many unpredicted bends, bifurcations, and changes in strike. In places, the fractures are diagonal rather than perpendicular to the ridge, and several parts of the ridge have no important fracture zones or even traces of them. For instance, they are absent from a 700-km-long portion of the Mid-Atlantic Ridge between the Atlantis and Kane fracture zones. There is a growing recognition that the fracture patterns in the Atlantic "show anomalies that are neither predicted by nor ...y et built into plate tectonic understanding" (Shirley, 1998a, 1998b). Side-scanning radar images show that the midocean ridges are cut by thousands of long, linear, ridge-parallel fissures, fractures, and faults. This strongly suggests that the ridges are underlain at shallow depth by interconnected magma channels, in which semifluid lava moves horizontally and parallel with the ridges rather than at right angles to them. The fault pattern observed is therefore totally different from that predicted by plate tectonics, and it cannot be explained by upwelling mantle diapirs as some plate tectonicists have proposed (Meyerhoff et al, 1992a). A zone of thrust faults, 300-400 km wide, has been discovered flanking the Mid-Atlantic Ridge over a length of 1,000 km (Antipov et al., 1990). Because it was produced under conditions of compression, it contradicts the plate-tectonic hypothesis that midocean ridges are dominated by tension. In Iceland, the largest landmass astride the Mid-Atlantic Ridge, the predominant stresses in the axial zone are likewise compressive rather than extensional (Keith, 1993). Earthquake data compiled by Zoback et al. (1 989) provide further evidence that ocean ridges are characterized by widespread compression, whereas recorded tensional earthquake activity associated with these ridges is rarer. The rough topography and strong tectonic deformation of much of the ocean ridges, particularly in the Atlantic and Indian Oceans, suggest that instead of being "spreading centers," they are a type of foldbelt (Storetvedt, 1997). The continents and oceans are covered with a network of major structures or lineaments, many dating from the Precambrian, along which tectonic and magmatic activity and associated mineralization take place (Anfiloff, 1992; Gay, 1973; Katterfeld and Charushin, 1973; Dickins and Choi, 1997; O'Driscoll, 1980; Wezel, 1992). The oceanic lineaments are not readily compatible with seafloor spreading and subduction, and plate tectonics shows little interest in them. GEOSAT data and SASS multibeam sonar data show that there are NNW-SSE and WSW-ENE megatrends in the Pacific Ocean, composed primarily of fracture zones and linear seamount chains, and these orthogonal lineaments naturally intersect (Smoot, 1997b, 1998a, 1998b, 1999). This is a physical impossibility in plate tectonics, as seamount chains suppos- 324 David Pratt edly indicate the direction of plate movement, and plates would therefore have to move in two directions at once! No satisfactory plate-tectonic explanation of any of these megatrends has been proposed outside the realm of ad hoc "microplates," and they are largely ignored. The orthogonal lineaments in the Atlantic Ocean, Indian Ocean, and Tasmanian Sea are also ignored (Choi, 1997, 1999a, 1 9 9 9 ~ ) . Age of the Seajloor The oldest known rocks from the continents are just under 4 billion years old, whereas-according to plate tectonics-none of the ocean crust is older than 200 million years (Jurassic). This is cited as conclusive evidence that oceanic lithosphere is constantly being created at midocean ridges and consumed in subduction zones. There is in fact abundant evidence against the alleged youth of the ocean floor, though geological textbooks tend to pass over it in silence. The oceanic crust is commonly divided into three main layers: Layer 1 consists of ocean floor sediments and averages 0.5 km in thickness; Layer 2 consists largely of basalt and is 1.0 to 2.5 km thick; and Layer 3 is assumed to consist of gabbro and is about 5 km thick. Scientists involved in the Deep Sea Drilling Project (DSDP) have given the impression that the basalt (Layer 2) found at the base of many deep-sea drill holes is basement, and that there are no further, older sediments below it. However, the DSDP scientists were apparently motivated by a strong desire to confirm seafloor spreading (Storetvedt, 1997). Of the first 429 sites drilled (1 968-1 977), only 165 (38%) reached basalt, and some penetrated more than one basalt. All but 12 of the 165 basalt penetrations were called "basement," including 19 sites where the upper contact of the basalt with the sediments was baked (Meyerhoff et al., 1992a). Baked contacts suggest that the basalt is an intrusive sill, and in some cases this has been confirmed, as the basalts turned out to have radiometric dates younger than the overlying sediments (e.g., Macdougall, 1971). One hundred one sedimentbasalt contacts were never recovered in cores, and therefore never actually seen, yet they were still assumed to be depositional contacts. In 33 cases, depositional contacts were observed, but the basalt sometimes contained sedimentary clasts, suggesting that there might be older sediments below. Indeed, boreholes that have penetrated Layer 2 to some depth have revealed an alternation of basalts and sedimentary rocks (Anderson et al., 1982; Hall and Robinson, 1979). Kamen-Kaye (1970) warned that before drawing conclusions on the youth of the ocean floor, rocks must be penetrated to depths of up to 5 km to see whether there are Triassic, Paleozoic, or Precambrian sediments below the so-called "basement." Plate tectonics predicts that the age of the oceanic crust should increase systematically with distance from the midocean ridge crests. Claims by DSDP scientists to have confirmed this are not supported by a detailed review of the Plate Tectonics: A Paradigm Under Threat 325 drilling results. The dates exhibit a very large scatter, which becomes even larger if dredge hauls are included (Figure 5). On some marine magnetic anomalies, the age scatter is tens of millions of years (Meyerhoff et al., 1992a). On one seamount just west of the crest of the East Pacific Rise, the radiometric dates range from 2.4 to 96 million years. Although a general trend is discernible from younger sediments at ridge crests to older sediments away from them, this is in fact to be expected, because the crest is the highest and most active part of the ridge; older sediments are likely to be buried beneath younger volcanic rocks. The basalt layer in the ocean crust suggests that magma flooding was once oceanwide, but volcanism was subsequently restricted to an increasingly narrow zone centered on the ridge crests. Such magma floods were accompanied by progressive crustal subsidence in large sectors of the present oceans, beginning in the Jurassic (Beloussov, 1980; Keith, 1993). The numerous finds in the Atlantic, Pacific, and Indian Oceans of rocks far older than 200 million years-many of them continental in nature-provide strong evidence against the alleged youth of the underlying crust. In the Atlantic, rock and sediment age should range from Cretaceous (120 million years) adjacent to the continents to very recent at the ridge crest. During legs 37 and 43 of the DSDP, Paleozoic and Proterozoic igneous rocks were recovered in cores on the Mid-Atlantic Ridge and the Bermuda Rise, yet not one of these occurrences of ancient rocks was mentioned in the Cruise Site Reports or Cruise Synthesis Reports (Meyerhoff et al., 1996a). Aumento and Loncarevic (1969) reported that 75% of 84 rock samples dredged from the Bald Mountain region just west of the Mid-Atlantic Ridge crest at 45ON consisted of continental-type rocks and commented that this was a "remarkable phenomenon"; so remarkable, in fact, that they decided to classify these rocks as "glacial erratic~" to give them no further consideration. Another way of dealing with and "anomalous" rock finds is to dismiss them as ship ballast. However, the Bald Mountain locality has an estimated volume of 80 km3, so it is hardly likely to have been rafted out to sea on an iceberg or dumped by a ship! It consists of granitic and silicic metamorphic rocks ranging in age from 1,690 to 1,550 million years and is intruded by 785-million-year mafic rocks (Wanless et al., 1968). Ozima et al. (1976) found basalts of Middle Jurassic age (169 million years) at the junction of the rift valley of the Mid-Atlantic Ridge and the Atlantis fracture zone (30°N), an area where basalt should theoretically be extremely young, and stated that they were unlikely to be ice-rafted rocks. Van Hinte and Ruffman (1995) concluded that Paleozoic limestones dredged from Orphan Knoll in the northwest Atlantic were in situ and not ice rafted. In another attempt to explain away anomalously old rocks and anomalously shallow or emergent crust in certain parts of the ridges, some plate tectonicists have argued that "nonspreading blocks" can be left behind during rifting and that the spreading axis and related transform faults can jump from 326 David Pratt VENDIAN 836 MA 600 Om' 1460 CARADOCIAN TRMOEITES AND ORAPTOLITES L A 380GARB. E L 300- TRILOBITES: PERMIAN OR OLDER 2 150CRET. E JUR. u 3 r r o MA 8138 364 20 813:118 .lo6 8100a3e7 8 387 -1Wsb d l 5 a=4 4: ; a380 e3.O *386 0 adz f14 KMI I 1,wo 1.500 z.doo DSDP DRLLSITE 0--0 LEO 3 WAS USE0 TO CONFmM SEA-FLOOR SPREAOINO da 2.600 3 . ~ 0 MIEOOE HAUL DATES ARE PROVmD ONLY FOR PRE-LATE AIRASBY: LOCALITKS Fig. 5. A plot of rock age versus distance from the crest of the Mid-Atlantic Ridge. The figure shows (to scale) rocks of all ages, whether from drill holes or dredge hauls. (Reprinted with permission from Meyerhoff et al., 1996a, fig. 2.35. Copyright by Kluwer Academic Publishers.) norez, 1971). This hypothesis was invoked by Pilot et al. (1998) to explain the presence of zircons with ages of 330 and 1,600 million years in gabbros beneath the Mid-Atlantic Ridge near the Kane fracture zone. Yet another way of dealing with anomalous rock ages is to reject them as unreliable. For instance, Reynolds and Clay (1977), reporting on a Proterozoic date (635 million years) near the crest of the Mid-Atlantic Ridge, wrote that the age must be wrong because the theoretical age of the site was only about 10 million years. Paleozoic trilobites and graptolites have been dredged from the King's Trough area, on the opposite side of the Mid-Atlantic Ridge to Bald Mountain, and at several localities near the Azores (Furon, 1949; Smoot and Meyerhoff, 1995). Detailed surveys of the equatorial segment of the Mid-Atlantic Ridge have provided a wide variety of data contradicting the seafloor-spreading model, including numerous shallow-water and continental rocks, with ages of up to 3.74 billion years (Timofeyev et al., 1992; Udintsev, 1996; Udintsev et al., 1993). Melson, Hart, and Thompson (1972), studying St. Peter and Paul's Rocks at the crest of the Mid-Atlantic Ridge just north of the equator, found an 835-million-year rock associated with other rocks giving 350-, 450-, and 2,000-million-year ages, whereas according to the seafloor-spreading model, the rock should have been 35 million years old. Numerous igneous and metamorphic rocks giving late Precambrian and Paleozoic radiometric ages have been dredged from the crests of the southern Mid-Atlantic, Mid-Indian, and Carlsberg ridges (Afanas' yev, 1967). Precambrian and Paleozoic granites have been found in several "oceanic" Plate Tectonics: A Paradigm Under Threat 327 plateaus and islands with anomalously thick crusts, including Rockall Plateau, Agulhas Plateau, the Seychelles, the Obruchev Rise, Papua New Guinea, and the Paracel Islands (Ben-Avraham et al., 1981; Sanchez Cela, 1999). In many cases, structural and petrological continuity exists between continents and anomalous "oceanic" crusts-a fact incompatible with seafloor spreading; this applies, for example, in the North Atlantic, where there is a continuous sialic basement, partly of Precambrian age, from North America to Europe. Major Precambrian lineaments in Australia and South America continue into the ocean floors, implying that the "oceanic" crust is at least partly composed of Precambrian rocks, and this has been confirmed by deepsea dredging, drilling, and seismic data, and by evidence for submerged continental crust (ancient paleolands) in the present southeast and northwest Pacific (Choi, 1997, 1998; see below). Marine Magnetic Anomalies Powerful support for seafloor spreading is said to be provided by marine magnetic anomalies-approximately parallel stripes of alternating high and low magnetic intensity that characterize much of the world's midocean ridges. According to the Morley-Vine-Matthews hypothesis, first proposed in 1963, as the fluid basalt welling up along the midocean ridges spreads horizontally and cools, it is magnetized by the earth's magnetic field. Bands of high intensity are believed to have formed during periods of normal magnetic polarity, and bands of low intensity during periods of reversed polarity. They are therefore regarded as time lines or isochrons. As plate tectonics became accepted, attempts to test this hypothesis or to find alternative hypotheses ceased. Correlations have been made between linear magnetic anomalies on either side of a ridge, in different parts of the oceans, and with radiometrically dated magnetic events on land. The results have been used to produce maps showing how the age of the ocean floor increases steadily with increasing distance from the ridge axis (McGeary and Plummer, 1998, fig. 4.19). As shown above, this simple picture can be sustained only by dismissing the possibility of older sediments beneath the basalt "basement" and by ignoring numerous "anomalously" old rock ages. The claimed correlations have been largely qualitative and subjective and are therefore highly suspect; virtually no effort has been made to test them quantitatively by transforming them to the pole (i.e., recalculating each magnetic profile to a common latitude). In one instance where transformation to the pole was carried out, the plate-tectonic interpretation of the magnetic anomalies in the Bay of Biscay was seriously undermined (Storetvedt, 1997). Agocs, Meyerhoff, and Kis (1992) applied the same technique in their detailed, quantitative study of the magnetic anomalies of the Reykjanes Ridge near Iceland and found that the correlations were very poor; the correlation coefficient along strike averaged 0.3 1 and that across the ridge 0.17, with lim- 328 David Pratt Positive magnetic anomaly Negative magnetic anomaly \ lRift crest at ridge *EARTHQUAKE EPICENTERS. NOTE THAT LINEAR ANOMALIESCROSS MIDOCEAN RIDGE AT OFJLlWE ANGLE NORTH OF I C E L A N D Fig. 6. Two views of marine magnetic anomalies. Top: A textbook cartoon. (Reprinted with permission from McGeary and Plummer, 1998. Copyright by The McGraw-Hill Companies.) Bottom: Magnetic anomaly patterns of the North Atlantic. (Reprinted with permission from Meyerhoff and Meyerhoff, 1972. Copyright by the American Geophysical Union.) Linear anomalies are known from only 70% of the seismically active midocean ridges. Moreover, the diagrams of symmetrical, parallel, linear bands of anomalies displayed in many plate-tectonics publications bear little resemblance to reality (Beloussov, 1970; Meyerhoff and Meyerhoff, 1974b) (Figure 6). The anomalies are symmetrical to the ridge axis in less than 50% of the ridge system where they are present, and in about 21% of it, they are oblique to the trend of the ridge. In some areas, linear anomalies are present where a ridge system is completely absent. Magnetic measurements by instruments towed near the sea bottom have indicated that magnetic bands actually consist of many isolated ovals that may be joined together in different ways. The initial, highly simplistic seafloor-spreading model for the origin of magnetic anomalies has been disproven by ocean drilling (Hall and Robinson, 1979; Pratsch, 1986). First, the hypothesis that the anomalies are produced in the upper 500 m of oceanic crust has had to be abandoned. Magnetic intensities, general polarization directions, and often the existence of different polar- Plate Tectonics: A Paradigm Under Threat 329 ity zones at different depths suggest that the source for oceanic magnetic anomalies lies in deeper levels of oceanic crust not yet drilled (or dated). Second, the vertically alternating layers of opposing magnetic polarization directions disprove the theory that the oceanic crust was magnetized entirely as it spread laterally from the magmatic center and strongly indicate that oceanic crustal sequences represent longer geologic times than is now believed. A more likely explanation of marine magnetic anomalies is that they are caused by fault-related bands of rock of different magnetic properties and have nothing to do with seafloor spreading (Choi, Vasil'yev, and Tuezov, 1990; Grant, 1980; Morris 1990; Pratsch, 1986). The fact that not all the charted magnetic anomalies are formed of oceanic crustal materials further undermines the plate-tectonic explanation. In the Labrador Sea, some anomalies occur in an area of continental crust that had previously been defined as oceanic (Grant, 1980). In the northwestern Pacific, some magnetic anomalies are likewise located within an area of continental crust-a submerged paleoland (Choi, Vasil'yev, and Tuezov, 1990; Choi, Vasil'yev, and Bhat, 1992). Magnetic anomaly bands strike into the continents in at least 15 places and "dive" beneath Proterozoic or younger rocks. Furthermore, they are approximately concentric with respect to Archean continental shields (Meyerhoff and Meyerhoff, 1972, 1974b). These facts imply that instead of being a "taped record" of seafloor spreading and geomagnetic field reversals during the past 200 million years, most oceanic magnetic anomalies are the sites of ancient fractures, which partly formed during the Proterozoic and have been rejuvenated since. The evidence also suggests that Archean continental nuclei have held approximately the same positions with respect to one another since their formation-which is utterly at variance with continental drift. Subduction Benioff zones are distinct earthquake zones that begin at an ocean trench and slope landward and downward into the earth. In plate tectonics, these deep-rooted fault zones are interpreted as "subduction zones" where plates descend into the mantle. They are generally depicted as 100-km-thick slabs descending into the earth either at a constant angle, or at a shallow angle near the earth's surface and gradually curving around to an angle of between 60" and 75". Neither representation is correct. Benioff zones often consist of two separate sections: an upper zone with an average dip of 33" extending to a depth of 70-400 km, and a lower zone with an average dip of 60" extending to a depth of up to 700 km (Benioff, 1954; Isacks and Barazangi, 1977). The upper and lower segments are sometimes offset by 100-200 km, and in one case by 350 km (Benioff, 1954, Smoot, 1997a). Furthermore, deep earthquakes are disconnected from shallow ones; very few intermediate earthquakes exist (Smoot, 1997a). Many studies have found transverse as well as vertical discontinuities and segmentation in Benioff zones (e.g., Carr, 1976; 330 David Pratt 500 600 700 0 KILOMETERS I500 1600 KILOMETERS 0 Fig. 7. Cross-sections across the Peru-Chile trench (left) and Bonin-Honshu arc (right), showing hypocenters. (Reprinted with permission from Benioff, 1954. Copyright by the Geological Society of America.) Carr, Stoiber, and Drake, 1973; Ranneft, 1979; Spence, 1977; Swift and Carr, 1974; Teisseyre et al., 1974). The evidence therefore does not favor the notion of a continuous, downgoing slab (Figure 7). Plate tectonicists insist that the volume of crust generated at midocean ridges is equaled by the volume subducted. But whereas 80,000 km of midocean ridges are supposedly producing new crust, only 30,500 km of trenches exist. Even if we add the 9,000 km of "collision zones," the figure is still only half that of the "spreading centers" (Smoot, 1997a). With two minor exceptions (the Scotia and Lesser Antilles trenchlarc systems), Benioff zones are absent from the margins of the Atlantic, Indian, Arctic, and Southern Oceans. Many geological facts demonstrate that subduction is not taking place in the Lesser Antilles arc; if it were, the continental Barbados Ridge should now be 200-400 km beneath the Lesser Antilles (Meyerhoff and Meyerhoff, 1974a). Kiskyras (1990) presented geological, volcanological, petrochemical, and seismological data contradicting the belief that the African plate is being subducted under the Aegean Sea. Africa is allegedly being converged on by plates spreading from the east, south, and west, yet it exhibits no evidence whatsoever for the existence of subduction zones or orogenic belts. Antarctica, too, is almost entirely surrounded by alleged "spreading" ridges without any corresponding subduction zones but fails to show any signs of being crushed. It has been suggested that Africa and Antarctica may remain stationary while the surrounding ridge system migrates away from them, but this would require the ridge marking the "plate boundary" between Africa and Antarctica to move in opposite directions simultaneously (Storetvedt, 1997)! If up to 13,000 km of lithosphere had really been subducted in circum-Pacific deep-sea trenches, vast amounts of oceanic sediments should have been scraped off the ocean floor and piled up against the landward margin of the trenches. However, sediments in the trenches are generally not present in the volumes required, nor do they display the expected degree of deformation Plate Tectonics: A Paradigm Under Threat 331 (Choi, 1999b; Gnibidenko, Krasny, and Popov, 1978; Storetvedt, 1997; Suzuki et al., 1997). Scholl and Marlow (1 974), who support plate tectonics, admitted to being "genuinely perplexed as to why evidence for subduction or offscraping of trench deposits is not glaringly apparent" (p. 268). Plate tectonicists have had to resort to the highly dubious notion that unconsolidated deep-ocean sediments can slide smoothly into a Benioff zone without leaving any significant trace. Moreover, fore-arc sediments, where they have been analyzed, have generally been found to be derived from the volcanic arc and the adjacent continental block, not from the oceanic region (Pratsch, 1990; Wezel, 1986). The very low level of seismicity, the lack of a megathrust, and the existence of flat-lying sediments at the base of oceanic trenches contradict the alleged presence of a downgoing slab (Dickins and Choi, 1998). Attempts by Murdock (1997), who accepts many elements of plate tectonics, to publicize the lack of a megathrust in the Aleutian trench (i.e., a million or more meters of displacement of the Pacific plate as it supposedly underthrusts the North American plate) have met with vigorous resistance and suppression by the plate-tectonics establishment. Subduction along Pacific trenches is also refuted by the fact that the Benioff zone often lies 80 to 150 km landward from the trench, by the evidence that Precambrian continental structures continue into the ocean floor, and by the evidence for submerged continental crust under the northwestern and southeastern Pacific, where there are now deep abyssal plains and trenches (Choi, 1987, 1998, 1999c; Smoot 1998b; Tuezov, 1998). If the "Pacific plate" is colliding with and diving under the "North American plate," there should be a stress buildup along the San Andreas Fault. The deep Cajon Pass drill hole was intended to confirm this but showed instead that no such stress is present (C. W. Hunt, 1992). In the active island-arc complexes of southeast Asia, the arcs bend back on themselves, forming hairpin-like shapes that sometimes involve full 180" changes in direction. This also applies to the postulated subduction zone around India. How plate collisions could produce such a geometry remains a mystery (Meyerhoff, 1995; H. A. Meyerhoff and Meyerhoff, 1977). Rather than being continuous curves, trenches tend to consist of a row of straight segments, which sometimes differ in depth by more than 4 km. Aseismic buoyant features (e.g., seamounts), which are frequently found at the juncture of these segments, are connected with increased deep-earthquake and volcanic activity on the landward side of the trench, whereas theoretically their "arrival" at a subduction zone should reduce or halt such activity (Smoot, 1997a). Plate tectonicists admit that it is hard to see how the subduction of a cold slab could result in the high heat flow or arc volcanism in back-arc regions or how plate convergence could give rise to back-arc spreading (Uyeda, 1986). Evidence suggests that oceanic, continental, and back-arc rifts are actually tensional structures developed to relieve stress in a strong compressional stress system, and therefore have nothing to do with seafloor spreading (Dickins, 1997). 332 David Pratt An alternative view of Benioff zones is that they are very ancient contraction fractures produced by the cooling of the earth (Meyerhoff et al., 1992b, 1996a). The fact that the upper part of the Benioff zones usually dips at less than 45" and the lower part at more than 45' suggests that the lithosphere is under compression and the lower mantle under tension. Furthermore, because a contracting sphere fractures along great circles (Bucher, 1956), this would account for the fact that both the circum-Pacific seismotectonic belt and the Alpine-Himalayan (Tethyan) belt lie on approximate circles. Finally, instead of oceanic crust being absorbed beneath the continents along ocean trenches, continents may actually be overriding adjacent oceanic areas to a limited extent, as is indicated by the historical geology of China, Indonesia, and the western Americas (Krebs, 1975; Pratsch, 1986; Storetvedt, 1997). Uplift and Subsidence Vertical Tectonics Classical plate tectonics seeks to explain all geologic structures primarily in terms of simple lateral movements of lithospheric plates-their rifting, extension, collision, and subduction, But random plate interactions are unable to explain the periodic character of geological processes, i.e., the geotectonic cycle, which sometimes operates on a global scale (Wezel, 1992). Nor can they explain the large-scale uplifts and subsidences that have characterized the evolution of the earth's crust, particularly those occurring far from "plate boundaries" such as in continental interiors, and vertical oscillatory motions involving vast regions (Beloussov, 1980, 1990; Chckunov, Gordienko, and Guterman, 1990; Genshaft and Saltykowski, 1990; Ilich, 1972). The presence of marine strata thousands of meters above sea level (e.g., near the summit of Mount Everest) and the great thicknesses of shallow-water sediment in some old basins indicate that vertical crustal movements of at least 9 km above sea level and 10-1 5 km below sea level have taken place (Spencer, 1977). Major vertical movements have also taken place along continental margins. For example, the Atlantic continental margin of North America has subsided by up to 12 km since the Jurassic (Sheridan, 1974). In Barbados, Tertiary coals representing a shallow-water, tropical environment occur beneath deep-sea oozes, indicating that during the last 12 million years, the crust sank to over 4-5 km depth for the deposition of the ooze and was then raised again. A similar situation occurs in Indonesia, where deep-sea oozes occur above sea level, sandwiched between shallow-water Tertiary sediments (James, 1994). The primary mountain-building mechanism in plate tectonics is lateral compression caused by collisions-of continents, island arcs, oceanic plateaus, seamounts, and ridges. In this model, subduction proceeds without mountain building until collision occurs, whereas in the noncollision model subduction alone is supposed to cause mountain building. As well as being Plate Tectonics: A Paradigm Under Threat 333 plate tectonics have pointed out (e.g., Cebull and Shurbet, 1990, 1992; Van Andel, 1998). The noncollision model fails to explain how continuous subduction can give rise to discontinuous orogeny, while the collision model is challenged by occurrences of mountain building where no continental collision can be assumed, and it fails to explain contemporary mountain-building activity along such chains as the Andes and around much of the rest of the Pacific rim. Asia supposedly collided with Europe in the late Paleozoic, producing the Ural mountains, but abundant geological field data demonstrate that the Siberian and East European (Russian) platforms have formed a single continent since Precambrian times (Meyerhoff and Meyerhoff, 1974a). McGeary and Plummer (1998) state that the plate tectonic reconstruction of the formation of the Appalachians in terms of three successive collisions of North America seems "too implausible even for a science fiction plot" (p. 114) but add that an understanding of plate tectonics makes the theory more palatable. Ollier (1 990), on the other hand, states that fanciful plate-tectonic explanations ignore all the geomorphology and much of the known geological history of the Appalachians. He also says that of all the possible mechanisms that might account for the Alps, the collision of the African and European plates is the most naive. The Himalayas and the Tibetan Plateau were supposedly uplifted by the collision of the Indian plate with the Asian plate. However, this fails to explain why the beds on either side of the supposed collision zone remain comparatively undisturbed and low-dipping, whereas the Himalayas have been uplifted, supposedly as a consequence, some 100 km away, along with the Kunlun Mountains to the north of the Tibetan Plateau. River terraces in various parts of the Himalayas are almost perfectly horizontal and untilted, suggesting that the Himalayas were uplifted vertically, rather than as the result of horizontal compression (Ahmad, 1990). Collision models generally assume that the uplift of the Tibetan Plateau began during or after the early Eocene (post-50 million years), but paleontological, paleoclimatological, paleoecological, and sedimentological data conclusively show that major uplift could not have occurred before earliest Pliocene time (5 million years ago) (Meyerhoff, 1995). There is ample evidence that mantle heat flow and material transport can cause significant changes in crustal thickness, composition, and density, resulting in substantial uplifts and subsidences. This is emphasized in many of the alternative hypotheses to plate tectonics (for an overview, see Yano and Suzuki, 1999), such as the model of endogenous regimes (Beloussov, 1980, 198 1, 1990, 1992; Pavlenkova, 1995, 1998). Plate tectonicists, too, increasingly invoke mantle diapirism as a mechanism for generating or promoting tectogenesis; there is now abundant evidence that shallow magma chambers are ubiquitous beneath active tectonic belts. The popular hypothesis that crustal stretching was the main cause of the for- 334 David Pratt mation of deep sedimentary basins on continental crust has been contradicted by numerous studies; mantle upwelling processes and lithospheric density increases are increasingly being recognized as an alternative mechanism (Anfiloff, 1992; Artyushkov, 1992; Artyushkov and Baer, 1983; Pavlenkova, 1998; Zorin and Lepina, 1989). This may involve gabbro-eclogite phase transformations in the lower crust (Artyushkov, 1992; Haxby, Turcotte, and Bird, 1976; Joyner, 1967), a process that has also been proposed as a possible explanation for the continuing subsidence of the North Sea Basin, where there is likewise no evidence of large-scale stretching (Collette, 1968). Plate tectonics predicts simple heat-flow patterns around the earth. There should be a broad band of high heat flow beneath the full length of the midocean rift system and parallel bands of high and low heat flow along the Benioff zones. Intraplate regions are predicted to have low heat flow. The pattern actually observed is quite different. There are criss-crossing bands of high heat flow covering the entire surface of the earth (Meyerhoff et al., 1996a). Intraplate volcanism is usually attributed to "mantle plumesw-upwellings of hot material from deep in the mantle, presumably the core-mantle boundary. The movement of plates over the plumes is said to give rise to hotspot trails (chains of volcanic islands and seamounts). Such trails should therefore show an age progression from one end to the other, but a large majority show little or no age progression (Baksi, 1999; Keith, 1993). On the basis of geological, geochemical, and geophysical evidence, Sheth (1999) argued that the plume hypothesis is ill-founded, artificial, and invalid, and has led earth scientists up a blind alley. Active tectonic belts are located in bands of high heat flow, which are also characterized by several other phenomena that do not readily fit in with the plate-tectonics hypothesis. These include bands of microearthquakes (including "diffuse plate boundaries") that do not coincide with plate-tectonic-predicted locations; segmented belts of linear faults, fractures, and fissures; segmented belts of mantle upwellings and diapirs; vortical geological structures; linear lenses of anomalous (low-velocity) upper mantle that are commonly overlain by shallower, smaller low-velocity zones; the existence of bisymmetrical deformation in all foldbelts, with coexisting states of compression and tension; strike-slip zones and similar tectonic lines ranging from simple rifts to Verschluckungszonen ("engulfment zones"); eastward-shifting tectonicmagmatic belts; and geothermal zones. Investigation of these phenomena has led to the development of a major new hypothesis of geodynamics, known as surge tectonics, which rejects both seafloor spreading and continental drift (Meyerhoff, 1995; Meyerhoff et al., 1992b, 1996a). Surge tectonics postulates that all the major features of the earth's surface, including rifts, foldbelts, metamorphic belts, and strike-slip zones, are underlain by shallow (less than 80 km) magma chambers and channels (known as "surge channels"). Seismotomographic data suggest that surge channels form an interconnected worldwide network, which has been dubbed "the earth's Plate Tectonics: A Paradigm Under Threat 335 cardiovascular system." Surge channels coincide with the lenses of anomalous mantle and associated low-velocity zones referred to above, and active channels are also characterized by high heat flow and microseismicity. Magma from the asthenosphere flows slowly through active channels at the rate of a few centimeters a year. Horizontal flow is demonstrated by two major surface features: linear, belt-parallel faults, fractures, and fissures; and the division of tectonic belts into fairly uniform segments. The same features characterize all lava flows and tunnels and have also been observed on Mars, Venus, and several moons of the outer planets. Surge tectonics postulates that the main cause of geodynamics is lithosphere compression, generated by the cooling and contraction of the earth. As compression increases during a geotectonic cycle, it causes the magma to move through a channel in pulsed surges and eventually to rupture it so that the contents of the channel surge bilaterally upward and outward to initiate tectogenesis. The asthenosphere (in regions where it is present) alternately contracts during periods of tectonic activity and expands during periods of tectonic quiescence. The earth's rotation, combined with differential lag between the more rigid lithosphere above and the more fluid asthenosphere below, causes the fluid or semifluid materials to move predominantly eastward. This explains the eastward migration through time of many magmatic or volcanic arcs, batholiths, rifts, depocenters, and foldbelts. The Continents It is a striking fact that nearly all the sedimentary rocks composing the continents were laid down under the sea. The continents have suffered repeated marine inundations, but because sediments were mostly deposited in shallow water (less than 250 m), the seas are described as "epicontinental." Marine transgressions and regressions are usually attributed mainly to eustatic changes of sea level caused by alterations in the volume of midocean ridges. Van Andel (1 994) points out that this explanation cannot account for the 100 or so briefer cycles of sea-level changes, particularly because transgressions and regressions are not always simultaneous all over the globe. He proposes that large regions or whole continents must undergo slow vertical, epeirogenic movements, which he attributes to an uneven distribution of temperature and density in the mantle, combined with convective flow. Some workers have linked marine inundations and withdrawals to a global thermal cycle, bringing about continental uplift and subsidence (Rutland, 1982; Sloss and Speed, 1974). Van Andel (1 994) admits that epeirogenic movements "fit poorly into plate tectonics" (p. 170) and are therefore largely ignored (Figures 8 and 9). Van Andel (1994) asserts that "plates" rise or fall by no more than a few hundred meters-this being the maximum depth of most "epicontinental" seas. However, this overlooks an elementary fact: huge thicknesses of sediments were often deposited during marine incursions, often requiring vertical crustal movements of many kilometers. Sediments accumulate in regions of 336 David Pratt Fig. 8. Maximum degree of marine inundation for each Phanerozoic geological period for the former USSR and North America. The older the geological period, the greater the probability of the degree of inundation being underestimated due to the sediments having been eroded or deeply buried beneath younger sediments. (Reprinted with permission from Harrison et al., 1983. Copyright by the American Geophysical Union.) subsidence, and their thickness is usually close to the degree of downwarping. In the unstable, mobile belts bordering stable continental platforms, many geosynclinal troughs and circular depressions have accumulated sedimentary thicknesses of 10-14 km, and in some cases of 20 km. Although the sedimentary cover on the platforms themselves is often less than 1.5 km thick, basins with sedimentary thicknesses of 10 km and even 20 km are not unknown (Beloussov, 198 1; Dillon, 1974; C. B. Hunt, 1992; Pavlenkova, 1998). Subsidence cannot be attributed solely to the weight of the accumulating sediments because the density of sedimentary rocks is much lower than that of the subcrustal material; e.g., the deposition of 1 km of marine sediment will cause only half a kilometer or so of subsidence (Holmes, 1965; Jeffreys, 1976). Moreover, sedimentary basins require not only continual depression of the base of the basin to accommodate more sediments, but also continuous uplift of adjacent land to provide a source for the sediments. In geosynclines, subsidence has commonly been followed by uplift and folding to produce mountain ranges, and this can obviously not be accounted for by changes in surface loading. The complex history of the oscillating uplift and subsidence of the crust appears to require deep-seated changes in lithospheric composition and density, as well as vertical and horizontal movements of mantle material. That density is not the only factor involved is shown by the fact that in regions of tectonic activity vertical movements often intensify gravity anomalies rather than acting to restore isostatic equilibrium. For example, the Greater Caucasus is overloaded, yet it is rising rather than subsiding (Beloussov, 1980; Jeffreys, 1976). In regions where all the sediments were laid down in shallow water, subsidence must somehow have kept pace with sedimentation. In eugeosynclines, on the other hand, subsidence proceeded faster than sedimentation, resulting Plate Tectonics: A Paradigm Under Threat 337 300 OS. AMER. 250 A N. AMER. *EUROPE i +ASIA x x x AFRICA A AUSTRALIA x A *AVERAGE x x Fig. 9. Sea-level changes for six continents. For each time interval, the sea-level elevations for the various continents differ widely, highlighting the importance of vertical tectonic movements on a regional and continental scale. (Reprinted with permission from Hallam, 1977. Copyright by Nature.) in a marine basin several kilometers deep. Examples of eugeosynclines prior to the uplift stage are the Sayans in the early Paleozoic, the eastern slope of the Urals in the early and middle Paleozoic, the Alps in the Jurassic and early Cretaceous, and the Sierra Nevada in the Triassic (Beloussov, 1980). Plate tectonicists often claim that geosynclines are formed solely at plate margins at the boundaries between continents and oceans. However, there are many examples of geosynclines having formed in intracontinental settings (Holmes, 1965), and the belief that the ophiolites found in certain geosynclinal areas are invariably remnants of oceanic crust is contradicted by a large volume of evidence (Beloussov, 198 l ;Bhat, 1987; Luts, 1990; Sheth, 1997). The Oceans In the past, sialic clastic material has been transported to today's continents from the direction of the present-day oceans, where there must have been considerable areas of land that underwent erosion (Beloussov, 1962; Dickins, Choi, and Yeates, 1992). For instance, the Paleozoic geosyncline along the seaboard of eastern North America, an area now occupied by the Appalachian mountains, was fed by sialic clasts from a borderland ("Appalachia") in the adjacent Atlantic. Other submerged borderlands include the North Atlantic Continent or Scandia (west of Spitsbergen and Scotland), Cascadia (west of the Sierra Nevada), and Melanesia (southeast of Asia and east of Australia) (Gilluly, 1955; Holmes, 1965; Umbgrove, 1947). A million cubic kilometers of Devonian micaceous sediments from Bolivia to Argentina imply an extensive continental source to the west where there is now the deep Pacific Ocean (Carey, 1994). During Paleozoic-Mesozoic-Paleogene times, the Japanese 338 David Pratt geosyncline was supplied with sediments from land areas in the Pacific (Choi, 1984, 1987). When trying to explain sediment sources, plate tectonicists sometimes argue that sediments were derived from the existing continents during periods when they were supposedly closer together (Bahlburg, 1993; Dickins, 1994a; Holmes, 1965). Where necessary, they postulate small former land areas (microcontinents or island arcs), which have since been either subducted or accreted against continental margins as "exotic terranes" (Choi, 1984; Kumon et al., 1988; Nur and Ben-Avraham, 1982). However, mounting evidence is being uncovered that favors the foundering of sizable continental landmasses, whose remnants are still present under the ocean floor (see below). Oceanic crust is regarded as much thinner and denser than continental crust: The crust beneath oceans is said to average about 7 km thick and to be composed largely of basalt and gabbro, whereas continental crust averages about 35 km thick and consists chiefly of granitic rock capped by sedimentary rocks. However, ancient continental rocks and crustal types intermediate between standard "continental" and "oceanic" crust are increasingly being discovered in the oceans (Sanchez Cela, 1999), and this is a serious embarrassment for plate tectonics. The traditional picture of the crust beneath oceans being universally thin and graniteless may well be further undermined in the future, as oceanic drilling and seismic research continue. One difficulty is to distinguish the boundary between the lower oceanic crust and upper mantle in areas where high- and low-velocity layers alternate (Choi, Vasil'yev, and Bhat, 1992; Orlenok, 1986). For example, the crust under the Kuril deep-sea basin is 8 km thick if the 7.9 kmls velocity layer is taken as the crust-mantle boundary (Moho), but 20-30 km thick if the 8.2 or 8.4 kmls layer is taken as the Moho (Tuezov, 1 998). Small ocean basins cover an area equal to about 5% of that of the continents and are characterized by transitional types of crust (Menard, 1967). This applies to the Caribbean Sea, the Gulf of Mexico, the Japan Sea, the Okhotsk Sea, the Black Sea, the Caspian Sea, the Mediterranean, the Labrador Sea and Baffin Bay, and the marginal (back-arc) basins along the western side of the Pacific (Beloussov and Ruditch, 196 1 ; Choi, 1984; Grant, 1992; Ross, 1974; Sheridan, 1974). In plate tectonics, the origin of marginal basins, with their complex crustal structure, has remained an enigma, and there is no basis for the assumption that some kind of seafloor spreading must be involved; rather, they appear to have originated by vertical tectonics (Storetvedt, 1997; Wezel, 1986). Some plate tectonicists have tried to explain the transitional crust of the Caribbean in terms of the continentalization of a former deep ocean area, thereby ignoring the stratigraphic evidence that the Caribbean was a land area in the Early Mesozoic (Van Bemmelen, 1972). There are over 100 submarine plateaus and aseismic ridges scattered throughout the oceans, many of which were once subaerially exposed (Dickins, Choi, and Yeates, 1992; Nur and Ben-Avraham, 1982; Storetvedt, 1997) Plate Tectonics: A Paradigm Under Threat 339 Fig. 10. Worldwide distribution of oceanic plateaus (black). (Reprinted with permission from Storetvedt, 1997. Copyright by Fagbokforlaget and K. M. Storetvedt.) (Figure 10). They make up about 10% of the ocean floor. Many appear to be composed of modified continental crust 20-40 km thick-far thicker than "normal" oceanic crust. They often have an upper 10-15-km crust with compressional-wave velocities typical of granitic rocks in continental crust. They have remained obstacles to predrift continental fits and have therefore been interpreted as extinct spreading ridges, anomalously thickened oceanic crust, or subsided continental fragments carried along by the "migrating" seafloor. If seafloor spreading is rejected, they cease to be anomalous and can be interpreted as submerged, in situ continental fragments that have not been completely "oceanized." Shallow-water deposits ranging in age from mid-Jurassic to Miocene, as well as igneous rocks showing evidence of subaerial weathering, were found in 149 of the first 493 boreholes drilled in the Atlantic, Indian, and Pacific Oceans. These shallow-water deposits are now found at depths of 1-7 km, demonstrating that many parts of the present ocean floor were once shallow seas, shallow marshes, or land areas (Orlenok, 1986; Timofeyev and Kholodov, 1984). From a study of 402 oceanic boreholes in which shallowwater or relatively shallow-water sediments were found, Ruditch (1990) concluded that there is no systematic correlation between the age of shallowwater accumulations and their distance from the axes of the midoceanic ridges, thereby disproving the seafloor-spreading model. Some areas of the oceans appear to have undergone continuous subsidence, whereas others experienced alternating episodes of subsidence and elevation. The Pacific 340 David Pratt Ocean appears to have formed mainly from the Late Jurassic to the Miocene, the Atlantic Ocean from the Late Cretaceous to the end of the Eocene, and the Indian Ocean during the Paleocene and Eocene. In the North Atlantic and Arctic Oceans, modified continental crust (mostly 10-20 km thick) underlies not only ridges and plateaus, but most of the ocean floor; only in deep-water depressions is typical oceanic crust found. Because deep-sea drilling has shown that large areas of the North Atlantic were previously covered with shallow seas, it is possible that much of the North Atlantic was continental crust before its rapid subsidence (Pavlenkova, 1995, 1998; Sanchez Cela, 1999). Lower Paleozoic continental rocks with trilobite fossils have been dredged from seamounts scattered over a large area northeast of the Azores. Furon (1 949) concluded that the continental cobbles had not been carried there by icebergs and that the area concerned was a submerged continental zone. Bald Mountain, from which a variety of ancient continental material has been dredged, could certainly be a foundered continental fragment. In the equatorial Atlantic, shallow-water and continental rocks are ubiquitous (Timofeyev et al., 1992; Udintsev, 1996). There is evidence that the midocean ridge system was shallow or partially emergent in Cretaceous to Early Tertiary time. For instance, in the Atlantic, subaerial deposits have been found on the North Brazilian Ridge (Bader et al., 197 I), near the Romanche and Vema fracture zones adjacent to equatorial sectors of the Mid-Atlantic Ridge (Bonatti and Chermak, 198 1 ; Bonatti and Honnorez, 197 I), on the crest of the Reykjanes Ridge, and in the Faeroe-Shetland region (Keith, 1993) (Figure 11). Oceanographic and geological data suggest that a large part of the Indian Ocean, particularly the eastern part, was land ("Lemuria") from the Jurassic until the Miocene. The evidence includes seismic and palynological data and subaerial weathering, which suggest that the Broken and Ninety East Ridges were part of an extensive, now sunken landmass; extensive drilling, seismic, magnetic, and gravity data pointing to the existence an Alpine-Himalayan foldbelt in the northwestern Indian Ocean, associated with a foundered continental basement; data that continental basement underlies the Scott, Exmouth, and Naturaliste plateaus west of Australia; and thick Triassic and Jurassic sedimentation on the western and northwestern shelves of the Australian continent, which shows progradation and current direction indicating a western source (Dickins, 1994a; Udintsev, Illarionov, and Kalinin, 1990; Udintsev and Koreneva, 1982; Wezel, 1988). Geological, geophysical, and dredging data provide strong evidence for the presence of Precambrian and younger continental crust under the deep abyssal plains of the present northwest Pacific (Choi, Vasil'yev, and Bhat, 1992; Choi, Vasil'yev, and Tuezov, 1990). Most of this region was either subaerially exposed or very shallow sea during the Paleozoic to Early Mesozoic and first became deep sea about the end of the Jurassic. Paleolands apparently existed on both sides of the Japanese islands. They were largely emergent during the Paleozoic-Mesozoic-Paleogene but were totally submerged during Paleogene to Plate Tectonics: A Paradigm Under Threat 341 Fig. 1 I . Areas in the Atlantic Ocean for which past subsidence has been established. Subsided areas are shaded. (Reprinted with permission from Dillon, 1974. Copyright by the AAPG, whose permission is required for further use.) Miocene times. Those on the Pacific side included the great Oyashio paleoland and the Kuroshio paleoland. The latter, which was as large as the present Japanese islands and occupied the present Nankai Trough area, subsided in the Miocene, at the same time as the upheaval of the Shimanto geosyncline, to which it had supplied vast amounts of sediments (Choi, 1984, 1987; Harata et al., 1978; Kumon et al., 1988). There is also evidence of paleolands in the southwest Pacific around Australia (Choi, 1997) and in the southeast Pacific during the Paleozoic and Mesozoic (Bahlburg, 1993; Choi, 1998; Isaacson, 1975; Isaacson and Martinez, 1995) (Figure 12). After surveying the extensive evidence for former continental land areas in the present oceans, Dickins, Choi, and Yeates (1 992) concluded, We are surprised and concerned for the objectivity and honesty of science that such data can be overlooked or ignored. There is a vast need for future Ocean Drilling Program initiatives to drill below the base of the basaltic ocean floor crust to confirm the real composition of what is currently designated oceanic crust. (p. 198) Conclusion Plate tectonics-the reigning paradigm in the earth sciences-faces some very severe and apparently fatal problems. Far from being a simple, elegant, 342 David Pratt Fig. 12. Former land areas in the present Pacific and Indian Oceans. Only those areas for which substantial evidence already exists are shown. Their exact outlines and full extent are as yet unknown. G1, Seychelles area; G2, Great Oyashio Paleoland; G3, Obruchev Rise; G4, Lemuria; S1, area of Ontong-Java Plateau, Magellan Sea Mounts, and Mid-Pacific Mountains; S2, Northeast Pacific; S3, Southeast Pacific including Chatham Rise and Campbell Plateau; S4, Southwest Pacific; S5, area including South Tasman Rise; S6, East Tasman Rise and Lord Howe Rise; S7, Northeast Indian Ocean; S8, Northwest Indian Ocean. (Reprinted with permission from Dickins, 1994a, 1994b. Copyright by J. M. Dickins.) all-embracing global theory, it is confronted with a multitude of observational anomalies and has had to be patched up with a complex variety of ad hoc modifications and auxiliary hypotheses. The existence of deep continental roots and the absence of a continuous, global asthenosphere to "lubricate" plate motions have rendered the classical model of plate movements untenable. There is no consensus on the thickness of the "plates" and no certainty as to the forces responsible for their supposed movement. The hypotheses of largescale continental movements, seafloor spreading, and subduction, as well as the relative youth of the oceanic crust are contradicted by a substantial volume of data. Evidence for significant amounts of submerged continental crust in the present-day oceans provides another major challenge to plate tectonics. The fundamental principles of plate tectonics therefore require critical reexamination, revision, or rejection. Acknowledgments I would kike to thank Ismail Bhat, Dong Choi, Mac Dickins, Hetu Sheth, and Chris Smoot for helpful comments and discussions. Plate Tectonics: A Paradigm Under Threat 343 References Afanas'yev, G. D. (1967). New data on relationship between earth's crust and upper mantle. International Geology Review, 9, 1,513-1,536. Agocs, W. B., Meyerhoff, A. A., & Kis, K. (1992). Reykjanes Ridge: Quantitative determinations from magnetic anomalies. In Chatterjee, S., & Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 221-238). Lubbock, TX: Texas Tech University Press. Ahmad, F. (1990). The bearing of paleontological evidence on the origin of the Himalayas. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 129-142). Athens, Greece: Theophrastus Publications, S. A. Anderson, D. L., Tanimoto, T., & Zhang, Y. (1992). Plate tectonics and hotspots: The third dimension. Science, 256, 1,645-1,65 1. Anderson, R. N., Honnorez, J., Becker, K., Adamson, A. C., Alt, J. C., Emmermann, R., Kempton, P. D., Kinoshita, H., Laverne, C., Mottl, M. J., & Newmark, R. L. (1982). DSDP hole 504B, the first reference section over 1 km through Layer 2 of the oceanic crust. Nature, 300, 589-594. Anfiloff, V. (1992). The tectonic framework of Australia. In Chatterjee, S., & Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 75-109). Lubbock, TX: Texas Tech University Press. Antipov, M. P., Zharkov, S. M., Kozhenov, V. Ya., & Pospelov, I. I. (1990). Structure of the MidAtlantic Ridge and adjacent parts of the abyssal plain at lat. 13ON. International Geology Review, 32, 468-478. Artyushkov, E. V. (1992). Role of crustal stretching on subsidence of the continental crust. Tectonophysics, 215, 187-207. Artyushkov, E. V., & Baer, M. A. (1983). Mechanism of continental crust subsidence in fold belts: The Urals, Appalachians and Scandinavian Caledonides. Tectonophysics, 100, 5-42. Aumento, F., & Loncarevic, B. D. (1969). The Mid-Atlantic Ridge near 45"N., 111. Bald Mountain. Canadian Journal of Earth Sciences, 6, 1 1-23. Bader, R. G., Gerard, R. D., Hay, W. W., Benson, W. E., Bolli, H. M., Rothwell, W. T., Ruef, M. H., Riedel, W. R., & Sayles, F. L. (1971). Leg 4 of the Deep Sea Drilling Project. Science, 172, 1,197-1,205. Bahlburg, H. (1993). Hypothetical southeast Pacific continent revisited: New evidence from the middle Paleozoic basins of northern Chile. Geology, 21, 909-912. Baksi, A. K. (1999). Reevaluation of plate motion models based on hotspot tracks in the Atlantic and Indian Oceans. Journal of Geology, 107, 13-26. Barron, E. J., Harrison, C. G. A., & Hay, W. W. (1978). A revised reconstruction of the southern continents. American Geophysical Union Transactions, 59, 436-439. Beloussov, V. V. (1 962). Basic problems in geotectonics. New York: McGraw-Hill. s Beloussov, V. V. (1970). ~ ~ a i nthet hypothesis of ocean-floor spreading. Tectonophysics, 9, 489-5 11. Beloussov, V. V. (1980). Geotectonics. Moscow: Mir. Beloussov, V. V. (1 98 1). Continental Endogenous Regimes. Moscow: Mir. Beloussov, V. V. (1 990). Certain trends in present-day geosciences. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 3-1 5). Athens, Greece: Theophrastus Publications, S. A. Beloussov, V. V. (1992). Endogenic regimes and the evolution of the tectonosphere. In Chatterjee, S., & Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 411-420). Lubbock, TX: Texas Tech University Press. Beloussov, V. V., & Ruditch, E. M. (1961). Island arcs in the development of the earth's structure (especially in the region of Japan and the Sea of Okhotsk). Journal of Geology, 69,647-658. Ben-Avraham, Z., Nur, A., Jones, D., & Cox, A. (1981). Continental accretion: From oceanic plateaus to allochthonous terranes. Science, 213, 47-54. Benioff, H. (1 954). Orogenesis and deep crustal structure-Additional evidence from seismology. Geological Society of America Bulletin, 65, 385-400. Bhat, M. I. (1 987). Spasmodic rift reactivation and its role in pre-orogenic evolution of the Himalayan region. Tectonophysics, 134, 103-1 27. Bonatti, E. (1990). Subcontinental mantle exposed in the Atlantic Ocean on St Peter-Paul islets. Nature, 345, 800-802. 344 David Pratt Bonatti, E., & Chermak, A. (1981). Formerly emerging crustal blocks in the Equatorial Atlantic. Tectonophysics, 72, 165-180. Bonatti, E., & Crane, K. (1982). Oscillatory spreading explanation of anomalously old uplifted crust near oceanic transforms. Nature, 300, 343-345. Bonatti, E., & Honnorez, J. (1971). Nonspreading crustal blocks at the Mid-Atlantic Ridge. Science, 174, 1,329-1,33 1. Briggs, J. C. (1 987). Biogeography andplate tectonics. Amsterdam: Elsevier. Brooks, C. E. P. (1 949). Climate through the ages. London: Ernest Benn. Bucher, W. H. (1933). The deformation of the earth k crust. Princeton, NJ: Princeton University Press. Bucher, W. H. (1956). Role of gravity in orogenesis. Geological Society ofAmerica Bulletin, 67, 1,295-1,318. Bullard, E. C., Everett, J. E., & Smith, A. G. (1965). The fit of the continents around the Atlantic. In A symposium on continental drift (Series A, 258, pp. 41-51). London: Royal Society of London Philosophical Transactions. Butler, R. F., Gehrels, G. E., McClelland, W. C., May, S. R., & Klepacki, D. (1989). Discordant paleomagnetic data from the Canadian Coast Plutonic Complex: Regional tilt rather than large-scale displacement? Geology, 17, 69 1-694. Carey, S. W. (1994). Creeds of physics. In Barone, M., & Selleri, F. (Eds.), Frontiers offundamentalphysics (pp. 241-255). New York: Plenum. Carr, M. J . (1 976). Underthrusting and Quaternary faulting in northern Central America. Geological Society ofAmerica Bulletin, 87, 825-829. Carr, M . J., Stoiber, R. E., & Drake, C. L. (1973). Discontinuities in the deep seismic zones under the Japanese arcs. Geological Society ofAmerica Bulletin, 84, 2,9 17-2,930. Cebull, S. E., & Shurbet, D. H. (1990). Fundamental problems with the plate-tectonic explanation of orogeny. In Barto-Kyriakidis, A. (Ed.), Critical aspects of theplate tectonics theory (Vol. 2, pp. 435-444). Athens, Greece: S. A. Theophrastus Publications. Cebull, S. E., & Shurbet, D. H. (1992). Conventional plate tectonics and orogenic models. In Chatterjee, S., & Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 111-1 17). Lubbock, TX: Texas Tech University Press. Chatterjee, S., & Hotton, N., I11 (1986). The paleoposition of India. Journal of Southeast Asian Earth Sciences, 1, 145-189. Chekunov, A. V., Gordienko, V. V., & Guterman, V. G. (1 990). Difficulties of plate tectonics and possible alternative mechanisms. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 397-433). Athens, Greece: Theophrastus Publications, S. A. Choi, D. R. (1984). Late Permian-early Triassic paleogeography of northern Japan: Did microplates accrete to Japan? Geology, 12, 728-73 l . Choi, D. R. (1987). Continental crust under the NW Pacific Ocean. Journal ofPetroleum Geology, 10,425-440. Choi, D. R. (1997). Geology of the oceans around Australia, parts 1-111. New Concepts in Global Tectonics Newsletter, 3, 8-1 3. Choi, D. R. (1998). Geology of the southeast Pacific, parts 1-3. New Concepts in Global Tectonics Newsletter, 7, 11-15; 8, 8-13; 9, 12-14. Choi, D. R. (1999a). Oceanic lineaments and major structures in Central America. New Concepts in Global Tectonics Newsletter, 11, 2 1-22. Choi, D. R. (1999b). Geology of East Pacific: Middle America trench. New Concepts in Global Tectonics Newsletter, 12, 10-1 6. . Choi, D. R. ( 1 9 9 9 ~ )Precambrian structures in South America: Their connection to the Pacific and Atlantic Oceans. New Concepts in Global Tectonics Newsletter, 13, 5-7. Choi, D. R., Vasil'yev, B. I., & Bhat, M. I. (1992). Paleoland, crustal structure, and composition under the northwestern Pacific Ocean. In Chatterjee, S., & Hotton, N. I11 (Eds.), New concepts in global tectonics (pp. 179-191). Lubbock, TX: Texas Tech University Press. Choi, D. R., Vasil'yev, B. I., & Tuezov, I. K. (1990). The great Oyashio paleoland: A PaleozoicMesozoic landmass in the northwestern Pacific. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 197-213). Athens, Greece: Theophrastus Publications, S. A. Collette, B. J. (1968). On the subsidence of the North Sea area. In Donovan, D.T. (Ed.), Geology ofsheIfseas (pp. 15-30). Edinburgh, Scotland: Oliver & Boyd. Dickins, J. M. (1987). Tethys-A geosyncline formed on continental crust? In McKenzie, K. G . Plate Tectonics: A Paradigm Under Threat (Ed.), Shallow Tethys 2: International Symposium, Wagga Wagga, 1986 (pp. 149-158). Rotterdam, The Netherlands: A. A. Balkema. Dickins, J. M. (1994a). What is Pangaea? In Embry, A. F., Beauchamp, B., & Glass, D. G. (Eds.), Pangaea: Global environments and resources (Memoir 17, pp. 67-80). Calgary, Alberta: Canadian Society of Petroleum Geologists. Dickins, J. M. (1994b). The nature of the oceans or Gondwanaland, fact and fiction. In Gondwana nine, Ninth International Gondwana Symposium, Hyderabad, India, 1994 (pp. 387-396). Rotterdam, The Netherlands: A. A. Balkema. The southern margin of Tethys. In Gondwana nine, Ninth International Dickins, J. M. (1994~). Gondwana Symposium, Hyderabad, India, 1994 (pp. 1,125-1,134). Rotterdam, The Netherlands: A. A. Balkema. Dickins, J. M. (1997). Rift, rifting. New Concepts in Global Tectonics Newsletter, 4, 3-4. Dickins, J. M., & Choi, D. R. (1997). Editorial. New Concepts in Global Tectonics Newsletter, 5, 1-2. I Dickins, J. M., & Choi, D. R. (1998). Fatal flaw-Who are the culprits? New Concepts in Global Tectonics Newsletter, 8, 1-2. Dickins, J. M., Choi, D. R., & Yeates, A. N. (1992). Past distribution of oceans and continents. In Chatterjee, S., & Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 193-199). Lubbock, TX: Texas Tech University Press. Dickinson, W. R., & Butler, R. F. (1998). Coastal and Baja California paleomagnetism reconsidered. Geological Society of America Bulletin, 110, 1,268-1,280. Dietz, R. S., & Holden, J. C. (1970). The breakup of Pangaea. Scientific American, 223, 3 0 4 1 . Dillon, L. S. (1974). Neovolcanism: A proposed replacement for the concepts of plate tectonics and continental drift. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 167-239). Tulsa, OK: American Association of Petroleum Geologists. Dott, R. H., Jr., & Batten, R. L. (1981). Evolution of the earth (3rd ed.). New York: McGraw-Hill. Dziewonski, A. M., & Anderson, D. L. (1984). Seismic tomography of the earth's interior. American Scientist, 72,483-494. Dziewonski, A. M., & Woodhouse, J. H. (1987). Global images of the earth's interior. Science, 236,37-48. Eyles, N., & Eyles, C. H. (1993). Glacial geologic confirmation of an intraplate boundary in the Parana basin of Brazil. Geology, 21, 459-462. Fallon, F. W., & Dillinger, W. H. (1992). Crustal velocities from geodetic very long baseline interferometry. Journal of Geophysical Research, 97, 7,129-7,136. Forte, A. M., Dziewonski, A. M., & O'Connell, R. J. (1995). Continent-ocean chemical heterogeneity in the mantle based on seismic tomography. Science, 268, 386-388. Furon, R. (1949). Sur les trilobites draguks a 4225 m de profondeur par le Talisman (1 883). Paris, AcadCmie des Sciences, Comptes Rendus, 228, 1,509-1,5 10. (For translation, see Schneck, 1974.) Gay, S. Parker, Jr. (1973). Pervasive orthogonal fracturing in earth k continental crust. Salt Lake City, UT: American Stereo Map Co. Genshaft, Yu. S., & Saltykowsky, A. Ya. (1990). Continental volcanism, xenoliths, and "plate tectonics." In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 2, pp. 267-280). Athens, Greece: Theophrastus Publications, S. A. Gilluly, J. (1955). Geologic contrasts between continents and ocean basins. In Crust of the Earth (Special paper 62, pp. 7-1 8). Boulder, CO: Geological Society of America. Gnibidenko, H. S., Krasny, M. L., & Popov, A. A. (1978). Tectonics of the Kuril-Kamchatka deep-sea trench. Eos, 59, 1,184. Gordon, R. G., & Stein, S. (1992). Global tectonics and space geodesy. Science, 256, 333-342. Gossler, J., & Kind, R. (1996). Seismic evidence for very deep roots of continents. Earth and Planetary Science Letters, 138, 1-13. Grad, H. (1971). Magnetic properties of a contained plasma. New York Academy of Science Annals, 182(17), 636-650. Grand, S. P. (1987). Tomographic inversion for shear velocity beneath the North American plate. Journal of Geophysical Research, 92, 14,065-14,090. Grant, A. C. (1980). Problems with plate tectonics: The Labrador Sea. Bulletin of Canadian Petroleum Geology, 28, 252-278. Grant, A. C. (1992). Intracratonic tectonism: Key to the mechanism of diastrophism. In Chatter- 346 David Pratt jee, S., & Hotton, N., 111 (Eds.), New concepts in global tectonics (pp. 65-73). Lubbock, TX: Texas Tech University Press. Gregory, J. W. (1899). The plan of the earth and its causes. The Geographical Journal, 13, 225-250. Gregory, J. W. (1 90 1). The plan of the earth and its causes. American Geologist, 27, 100-119, 137-147. Gregory, J. W. (1929). The geological history of the Atlantic Ocean. Quarterly Journal of Geological Society, 85. 68-122. Hall, J. M., & Robinson, P. T. (1979). Deep crustal drilling in the North Atlantic Ocean. Science, 204, 573-586. Hallam, A. (1 976). How closely did the continents fit together? Nature, 262, 94-95. Hallam, A. (1977). Secular changes in marine inundation of USSR and North America through the Phanerozoic. Nature, 269, 769-772. Hallam, A. (1979). A decade of plate tectonics. Nature, 279, 478. Harata, T., Hisatomi, K., Kumon, F., Nakazawa, K., Tateishi, M., Suzuki, H., & Tokuoka, T. (1978). Shimanto geosyncline and Kuroshio paleoland. Journal of Physics of Earth, 26(S~ppl.), 357-366. Harrison, C. G. A., Miskell, K. J., Brass, G. W., Saltzman, E. S., & Sloan, J. L. (1983). Continental hypsography. Tectonics, 2, 357-377. Haxby, W. F., Turcotte, D. L., & Bird, J. M. (1976). Thermal and mechanical evolution of the Michigan Basin. Tectonophysics, 36, 57-75. Hodych, J. P., & Bijaksana, S. (1993). Can remanence anisotropy detect paleomagnetic inclination shallowing due to compaction? A case study using Cretaceous deep-sea limestones. Journal of Geophysical Research, 98, 22,429-22,44 1. Holmes, A. (1965). Principles ofphysical geology (2nd ed.). London: Thomas Nelson and Sons. Hunt, C. B. (1992). Geochronology. Encyclopaedia Britannica (1 5th ed., Vol. 19, pp. 824-826). Hunt, C. W. (Ed.). (1992). Expandinggeospheres. Calgary, Alberta: Polar Publishing. Ilich, M. (1972). New global tectonics: Pros and cons. American Association of Petroleum Geologists Bulletin, 56, 360-363. Irving, E., & Archibald, D. A. (1990). Bathozonal tilt corrections to paleomagnetic data from mid-Cretaceous plutonic rocks: Examples from the Omineca belt, British Columbia. Journal of Geophysical Research, 95, 4,579-4,585. Isaacson, P. E. (1975). Evidence for a western extracontinental land source during the Devonian period in the Central Andes. Geological Society of America Bulletin, 86, 39-46. Isaacson, P. E., & Martinez, E. D. (1995). Evidence for a middle-late Paleozoic foreland basin and significant paleolatitudinal shift, central Andes. In Tankard, A. J., Soruco, R. S., & Welsink, H. J. (Eds.), Petroleum basins ofSouth America (Memoir 62, pp. 231-249). Tulsa, OK: American Association of Petroleum Geologists. Isacks, B. L., & Barazangi, M. (1977). Geometry of Benioff zones: Lateral segmentation and downwards bending of the subducted lithosphere. In Talwani, M., & Pitman, W. C., I11 (Eds.), Island arcs, deep sea trenches and back-arc basins (Maurice Ewing Series 1, pp. 99-1 14). Washington, DC: American Geophysical Union. James, P. (1994). The tectonics of Geoid changes. Calgary, Alberta: Polar Publishing. Jeffreys, H. (1974). Theoretical aspects of continental drift. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 395-405). Tulsa, OK: American Association of Petroleum Geologists. Jeffreys, H. (1976). The earth: its origin, history andphysical constitution (6th ed.). Cambridge, UK: Cambridge University Press. Jordan, T. H. (1975). The continental tectosphere. Reviews of Geophysics and Space Physics, 13, 1-12. Jordan, T. H. (1978). Composition and development of the continental tectosphere. Nature, 274, 544-548. Jordan, T. H. (1 979). The deep structure of the continents. Scientific American, 240, 70-82. Joyner, W. B. (1967). Basalt-eclogite transition as a cause for subsidence and uplift. Journal of Geophysical Research, 72, 4,9774,998. Kamen-Kaye, M. (1970). Age of the basins. Geotimes, 115, 6-8. Kashfi, M. S. (1992). Geological evidence for a simple horizontal compression of the crust in the Zagros Crush Zone. In Chatterjee, S., & Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 119-130). Lubbock, TX: Texas Tech University Press. Plate Tectonics: A Paradigm Under Threat 347 Katterfeld, G. H., & Charushin, G. V. (1973). General grid systems of planets. Modern Geology, 4,243-287. Keith, M. L. (1993). Geodynamics and mantle flow: An alternative earth model. Earth-Science Reviews, 33, 153-337. Kent, D. V., & Smethurst, M. A. (1998). Shallow bias of paleomagnetic inclinations in the Paleozoic and Precambrian. Earth and Planetary Science Letters, 160, 391-402. Khudoley, K. M. (1974). Circum-Pacific Mesozoic ammonoid distribution: Relation to hypotheses of continental drift, polar wandering, and earth expansion. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 295-330). Tulsa, OK: American Association of Petroleum Geologists. Kiskyras, D. A. (1990). Some remarks on the plate tectonics concept with respect to geological and geophysical problems in the Greek area. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 2 15-244). Athens, Greece: Theophrastus Publications, S. A. Krebs, W. (1 975). Formation of southwest Pacific island arc-trench and mountain systems: Plate or global-vertical tectonics? American Association of Petroleum Geologists Bulletin, 59, 1,639-1,666. Kumon, F., Suzuki, H., Nakazawa, K., Tokuoka, T., Harata, T., Kimura, K., Nakaya, S., Ishigami, T., & Nakamura, K. (1988). Shimanto belt in the Kii peninsula, southwestern Japan. Modern Geology, 12, 71-96. Le Grand, H. E. (1988). Drifting continents and shifting theories. Cambridge, UK: Cambridge University Press. Lerner-Lam, A. L. (1988). Seismological studies of the lithosphere. Lamont-Doherty Geological Observatory, Yearbook, pp. 50-55. Lowman, P. D., Jr. (1985). Plate tectonics with fixed continents: A testable hypothesis-I. Journal of Petroleum Geology, 8, 373-388. Lowman, P. D., Jr. (1986). Plate tectonics with fixed continents: A testable hypothesis-11. Journal ofPetroleum Geology, 9, 71-87. Lowman, P. D., Jr. (1992a). Plate tectonics and continental drift in geologic education. In Chatterjee, s . , & Hotton, N., 111 (Eds.), New concepts in global tectonics (pp. 3-9). Lubbock, TX: Texas Tech University Press. Lowman, P. D., Jr. (1992b). Geophysics from orbit: The unexpected surprise. Endeavour, 16, 50-58. Luts, B. G. (1990). Types of ophiolitic formations (are they remnants of oceanic crust?). In BartoKyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 2, pp. 281-305). Athens, Greece: Theophrastus Publications, S. A. Lyttleton, R. A., & Bondi, H. (1992). How plate tectonics may appear to a physicist. Journal of the British Astronomical Association, 102, 194-195. Lyustikh, E. N. (1967). Criticism of hypotheses of convection and continental drift. Geophysical Journal of Royal Astronomical Society, 14, 347-352. MacDonald, G. J. F. (1963). The deep structure of continents. Reviews of Geophysics, 1, 587-665. Macdougall, D. (1971). Deep sea drilling: Age and composition of an Atlantic basaltic intrusion. Science, 171, 1,244-1,245. Manabe, S., & Broccoli, A. J. (1990). Mountains and arid climates of middle latitudes. Science, 247, 192-195. Mantura, A. J. (1972). New global tectonics and "the new geometry." American Association of Petroleum Geologists Bulletin, 56, 2,451-2,455. Maxwell, J. C. (1974). The new global tectonics-An assessment. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 24-42). Tulsa, OK: American Association of Petroleum Geologists. McGeary, D., & Plummer, C. C. (1998). Physicalgeology: Earth revealed (3rd ed.). Boston, MA: WCB, McGraw-Hill. McKenzie, K. G. (1987). Tethys and her progeny. In McKenzie, K. G. (Ed.), Shallow Tethys 2, International Symposium, Wagga Wagga, 1986 (pp. 501-523). Rotterdam, The Netherlands: A. A. Balkema. McLeish, A. (1992). Geological science. Walton-on-Thames, UK: Thomas Nelson and Sons. Melson, W. G., Hart, S. R., & Thompson, G. (1972). St. Paul's Rocks, equatorial Atlantic: Petrogenesis, radiometric ages, and implications on sea-floor spreading. In Shagam, R., Hargraves, R. B., Morgan, W. J., Van Houten, F. B., Burk, C. A., Holland, H. D., & Hollister, L. C. (Eds.), 348 David Pratt Studies in earth and space sciences (Memoir 132, pp. 241-272). Boulder, CO: Geological Society of America. Menard, H. W. (1967). Transitional types of crust under small ocean basins. Journal of Geophysical Research, 72, 3,061-3,073. Merrill, R. T., McElhinny, M. W., & McFadden, P. L. (1996). The magnetic$eld of the earth. San Diego, CA: Academic Press. Meyerhoff, A. A. (1 970a). Continental drift: Implications of paleomagnetic studies, meteorology, physical oceanography, and climatology. Journal of Geology, 78, 1-5 1. Meyerhoff, A. A. (1970b). Continental drift, 11: High latitude evaporite deposits and geologic history of Arctic and North Atlantic oceans. Journal of Geology, 78,406-444. Meyerhoff, A. A. (1974). Crustal structure of northern North Atlantic Ocean-A review. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 423-433). Tulsa, OK: American Association of Petroleum Geologists. Meyerhoff, A. A. (1995). Surge-tectonic evolution of southeastern Asia: A geohydrodynamics approach. Journal of Southeast Asian Earth Sciences, 12, 143-247. Meyerhoff, A. A., & Hatten, C. W. (1 974). Bahamas salient of North America. In Burk, C. A., & Drake, C. L. (Eds.), The geology of continental margins (pp. 429-446). Berlin, Germany: Springer-Verlag. Meyerhoff, A. A., & Meyerhoff, H. A. (1972). The new global tectonics: Age of linear magnetic anomalies of ocean basins. American Association of Petroleum Geologists Bulletin, 56, 337-359. Meyerhoff, A. A., & Meyerhoff, H. A. (1974a). Tests of plate tectonics. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 43-145). Tulsa, OK: American Association of Petroleum Geologists. Meyerhoff, A. A., & Meyerhoff, H. A. (1974b). Ocean magnetic anomalies and their relations to continents. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 41 1-422). Tulsa, OK: American Association of Petroleum Geologists. Meyerhoff, A. A., & Teichert, C. (1971). Continental drift, 111: Late Paleozoic glacial centers and Devonian-Eocene coal distribution. Journal of Geology, 79, 285-321. Meyerhoff, A. A., Kamen-Kaye, M., Chen, C., & Taner, 1. (1991). China-Stratigraphy, paleogeography and tectonics. Dordrecht: Kluwer. Meyerhoff, A. A., Agocs, W. B., Taner, I., Morris, A. E. L., & Martin, B. D. (1992a). Origin of midocean ridges. In Chatterjee, S., & Hotton, N., 111(Eds.), New concepts in global tectonics (pp. 151-178). Lubbock, TX: Texas Tech University Press. Meyerhoff, A. A., Taner, I., Morris, A. E. L., Martin, B. D., Agocs, W. B., & Meyerhoff, H. A. (1992b). Surge tectonics: A new hypothesis of earth dynamics. In Chatterjee, S., & Hotton, N., 111(Eds.), New concepts in global tectonics (pp. 309-409). Lubbock, TX: Texas Tech University Press. Meyerhoff, A. A., Taner, I., Morris, A. E. L., Agocs, W. B., Kaymen-Kaye, M., Bhat, M. I., Smoot, N. C., & Choi, D. R. (1996a). Surge tectonics: A new hypothesis of global geodynamics (Meyerhoff Hull, D., Ed.). Dordrecht: Kluwer. Meyerhoff, A. A., Boucot, A. J., Meyerhoff Hull, D., & Dickins, J. M. (1996b). Phanerozoic faunal &floral realms of the earth: The intercalary relations of the Malvinokaffric and Gondwana faunal realms with the Tethyan faunal realm (Memoir 189). Geological Society of America. Meyerhoff, H. A., & Meyerhoff, A. A. (1977). Genesis of island arcs. In Internationalsymposium on geodynamics in Southwest Pacific Noumea, Novelle Cal Edonie (pp. 357-370). Paris: ~ d i tions Technip. Morris, A. E. L., Taner, I., Meyerhoff, H. A., & Meyerhoff, A. A. (1990). Tectonic evolution of the Caribbean region: Alternative hypothesis. In Dengo, G., & Case, J. E. (Eds.), The Caribbean Region (pp. 433-457). Boulder, CO: Geological Society of America. Munk, W. H., & MacDonald, G. J. F. (1975). The rotation of the earth. Cambridge, UK: Cambridge University Press. Murdock, J. N. (1997). Overview of the history of one man's challenges to strict plate tectonics. New Concepts in Global Tectonics Newsletter, 4, 23-25. Nafe, J. E., & Drake, C. L. (1969). Floor of the North Atlantic-Summary of geophysical data. In Kay, M. (Ed.), North Atlantic-Geology and continental drift (Memoir 12, pp. 59-87). Tulsa, OK: American Association of Petroleum Geologists. Plate Tectonics: A Paradigm Under Threat 349 I Nitecki, M. H., Lemke, J. L., Pullman, H. W., & Johnson, M. E. (1978). Acceptance of plate tectonic theory by geologists. Geology, 6, 661-664. Nur, A., & Ben-Avraham, Z. (1982). Displaced terranes and mountain building. In Hsii, K. J. (Ed.), Mountain buildingprocesses (pp. 73-84). London: Academic Press. O'Driscoll, E. S. T. (1 980). The double helix in global tectonics. Tectonophysics, 63, 3 9 7 4 17. Ollier, C. D. (1990). Mountains. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 2, pp. 211-236). Athens, Greece: Theophrastus Publications, S. A. Orlenok, V. V. (1986). The evolution of ocean basins during Cenozoic time. Journal of Petroleum Geology, 9, 207-216. Ozima, M., Saito, K., Matsuda, J., Zashu, S., Aramaki, S., & Shido, F. (1976). Additional evidence of existence of ancient rocks in the Mid-Atlantic Ridge and the age of the opening of the Atlantic. Tectonophysics, 31, 59-71. Pavlenkova, N. 1. (1 990). Crustal and upper mantle structure and plate tectonics. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 73-86). Athens, Greece: Theophrastus Publications, S. A. Pavlenkova, N. I. (1995). Structural regularities in the lithosphere of continents and plate tectonics. Tectonophysics, 243, 223-229. Pavlenkova, N. I. (1 996). General features of the uppermost mantle stratification from long-range seismic profiles. Tectonophysics, 264, 261-278. Pavlenkova, N. I. (1998). Endogenous regimes and plate tectonics in northern Eurasia. Physics and Chemistry of the Earth, 23, 799-810. Pilot, J., Werner, C.-D., Haubrich, F., & Baumann, D. (1 998). Paleozoic and Proterozoic zircons from the Mid-Atlantic Ridge. Nature, 393, 676-679. Polet, J., & Anderson, D. L. (1995). Depth extent of cratons as inferred from tomographic studies. Geology, 23,205-208. Pollack, H. N., & Chapman, D. S. (1977). On the regional variation of heat flow, geotherms, and lithospheric thickness. Tectonophysics, 38, 279-296. Pratsch, J. C. (1986, July 14). Petroleum geologist's view of oceanic crust age. Oil and Gas Journal, 112-116. Pratsch, J. C. (1990). Relative motions in geology: Some philosophical differences. Journal of Petroleum Geology, 13, 229-234. Ranneft, T. S. M. (1979). Segmentation of island arcs and application to petroleum geology. Journal of Petroleum Geology, 1, 35-53. ~r Reynolds, P. H., & Clay, W. (1977). Leg 37 basalts and gabbro: K-Ar and 4 0 ~ r - 3 9 dating. In Aumento, F., Melson, W. G., et al. (Eds), Initial reports of the Deep Sea Drilling Project (Vol. 37, pp. 629-630). Washington, DC: U.S. Government Printing Office. Ross, D. A. (1974). The Black Sea. In Burk, C. A., & Drake, C. L. (Eds.), The geology of continental margins (pp. 669-682). Berlin, Germany: Springer-Verlag. Ruddiman, W. F., & Kutzbach, J. E. (1989). Effects of plateau uplift on late Cenozoic climate. Eos, 70, 294. Ruditch, E. M. (1990). The world ocean without spreading. In Barto-Kyriakidis, A. (Ed.), Critical aspects of theplate tectonics theory (Vol. 2, pp. 343-395). Athens, Greece: Theophrastus Publications, S. A. Rutland, R. W. R. (1982). On the growth and evolution of continental crust: A comparative tectonic approach. Journal and Proceedings of the Royal Society uf New South Wales, 115, 33-60. Sanchez Cela, V. (1999). Formation of Ma6c-ultramafic rocks in the crust: need for a new upper mantle. Zaragoza: Universidad de Zaragoza. Saull, V. A. (1986). Wanted: Alternatives to plate tectonics. Geology, 14, 536. Saxena, M. N., & Gupta, V. J. (1990). Role of foredeep tectonics, geological and palaeontological data, gravity tectonics in the orogeny and uplift of the Himalayas, vis-a-vis continental drift and plate tectonics concepts. In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 105-128). Athens, Greece: Theophrastus Publications, S. A. Schneck, M. C. (1974). Mid-Atlantic trilobites. Geotimes, 19, 16. Scholl, D. W., & Marlow, M. S. (1974). Global tectonics and the sediments of modern and ancient trenches: some different interpretations. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 255-272). American Association of Petroleum Geologists. 350 David Pratt Scotese, C. R., Gahagan, L. M., & Larson, R. L. (1988). Plate tectonic reconstructions of the Cretaceous and Cenozoic ocean basins. Tectonophysics, 155, 27-48. Seyfert, C. K. (1998). The earth: Its properties, composition, and structure. Encyclopaedia Brit a n n i c ~CD-ROM, 1994-1998. , Shapiro, M. N. (1990). Is the opening of the Atlantic manifested in the structure of the northern framing of the Pacific Ocean? In Barto-Kyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 1, pp. 179-196). Athens, Greece: S. A. Theophrastus Publications. Sheridan, R. E. (1 974). Atlantic continental margin of North America. In Burk, C. A., & Drake, C. L. (Eds.), The geology of continental margins (pp. 39 1-407). Berlin, Germany: Springer-Verlag. Sheth, H. C. (1997). Ophiolites: Another paradox. New Concepts in Global Tectonics Newsletter, 5, 5-6. Sheth, H. C. (1999). Flood basalts and large igneous provinces from deep mantle plumes: Fact, fiction, and fallacy. Tectonophysics, 311, 1-29. Shirley, K. (1998a). Ocean floor mapped from space. AAPG Explorer, October, 20-23. Shirley, K. (1998b). New maps show continent ties. AAPG Explorer, November, 20-2 1. Skinner, B. J., & Porter, S. C. (1995). The dynamic earth: An introduction to physical geology (3rd ed.). New York: John Wiley & Sons. Sloss, L. L., & Speed, R. C. (1974). Relationships of cratonic and continental-margin tectonic episodes. In Dickinson, W. R. (Ed.), Tectonics and sedimentation (Special Publication 22, pp. 98-119). Tulsa, OK: Society of Economic Paleontologists and Mineralogists. Smiley, C. J. (1 974). Analysis of crustal relative stability from some late Paleozoic and Mesozoic floral records. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 33 1-360). Tulsa, OK: American Association of Petroleum Geologists. Smiley, C. J. (1976). Pre-Tertiary phytogeography and continental drift-Some apparent discrepancies. In Gray, J., & Boucot, A. J. (Eds.), Historical biogeography, plate tectonics, and the changing environment (pp. 3 1 1 319). Corvallis, OR: Oregon State University Press. Smiley, C. J. (1992). Paleofloras, faunas, and continental drift: Some problem areas. In Chatterjee, s., & Hotton, N., 111 (Eds.), New concepts in global tectonics (pp. 241-257). Lubbock, TX: Texas Tech University Press. Smith, A. G., & Hallam, A. (1970). The fit of the southern continents. Nature, 225, 139-144. Smith, A. G., Hurley, A. M., & Briden, J. C. (198 1). Phanerozoic paleocontinental world maps. Cambridge, UK: Cambridge University Press. Smith, 1). E., Kolenkiewicz, R., Nerem, R. S., Dunn, P. J., Torrence, M. H., Robbins, J. W., Klosko, S. M., Williamson, R. G., & Pavlis, E. C. (1994). Contemporary global horizontal crustal motion. Geophysics Journal International, 11 9, 5 11-520. Smoot, N. C. (1997a). Aligned buoyant highs, across-trench deformation, clustered volcanoes, and deep earthquakes are not aligned with plate-tectonic theory. Geomorphology, 18, 199-222. Smoot, N. C. (1997b). Magma floods, microplates, and orthogonal intersections. New Concepts in Global Tectonics Newsletter, 5, 8-1 3. Smoot, N. C. (1998a). The trans-Pacific Chinook Trough megatrend. Geomorphology, 24, 333-351. Smoot, N. C. (1998b). WNW-ESE Pacific lineations. New Concepts in Global Tectonics Newsletter, 9, 7-11. Smoot, N. C. (1999). Orthogonal intersections of megatrends in the Western Pacific ocean basin: A case study of the Mid-Pacific mountains. Geomorphology, 30, 323-356. Smoot, N. C., & Meyerhoff, A. A. (1995). Tectonic fabric of the Atlantic Ocean floor: Speculation vs. reality. Journal of Petroleum Geology, 18, 207-222. Spence, W. (1977). The Aleutian arc: Tectonic blocks, episodic subduction, strain diffusion, and magmatic generation. Journal of Geophysical Research, 82, 2 13-230. Spencer, E. W. (1 977). Introduction to the structure of the earth (2nd ed.). New York: McGrawHill. Stanley, S. M. (1989). Earth and life through time (2nd ed.). New York: W.H. Freeman and Company. Steers, J. A. (1950). The unstable earth (5th ed.). London: Methuen. Stocklin, J. (1989). Tethys evolution in the Afghanistan-Pamir-Pakistan region. In Sengor, A. M. C. (Ed.), Tectonic evolution of the Tethyan Region (pp. 241-264). Dordrecht: Kluwer. Storetvedt, K. M. (1992). Rotating plates: New concept of global tectonics. In Chatterjee, S., & Plate Tectonics: A Paradigm Under Threat 35 1 Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 203-220). Lubbock, TX: Texas Tech University Press. Storetvedt, K. M. (1997). Our evolving planet: Earth history in new perspective. Bergen, Norway: Alma Mater. Suzuki, Y., Harada, I., Iikawa, K., Kobayashi, K., Nomura, N., Oda, K., Ogawa, Y., Watanabe, F., & Yamazaki, K. (1997). Geological structure of northeast Honshu, Japan in contradiction to the plate tectonics. New Concepts in Global Tectonics Newsletter, 5, 17-1 9. Swift, S. A., & Carr, M. J. (1974). The segmented nature of the Chilean seismic zone. Physics of the Earth and Planetary Interiors, 9, 183-19 1. Tarling, D. H. (1971). Gondwanaland, palaeomagnetism and continental drift. Nature, 229, 17-2 1. Tarling, D. H. (1982). Land bridges and plate tectonics. Geobios, MCmoire SpCcial, 6, 361-374. Teichert, C. (1974). Marine sedimentary environments and their faunas in Gondwana area. In Kahle, C. F. (Ed.), Plate tectonics-Assessments and reassessments (Memoir 23, pp. 361-394). Tulsa, OK: American Association of Petroleum Geologists. Teichert, C., & Meyerhoff, A. A. (1972). Continental drift and the marine environment. Montreal: 24th International Geology Conference, pp. 339-349. Teisseyre, R., Wojtczak-Gadomska, B., Vesanen, E., & Maki, M-L. (1974). Focus distribution in South American deep-earthquake regions and their relation to geodynamic development. Physics of the Earth and Planetary Interiors, 9, 290-305. Timofeyev, P. P., & Kholodov, V. N. (1984). The problem of existence of oceans in geologic history. Doklady Akademii Nauk. SSSR, Earth Science Sections, 276, 61-63. Timofeyev, P. P., Udintsev, G. B., Agapova, G. V., Antipov, M. P., Boyko, N. I., Yeremeyev, V. V., Yefimov, V. N., Kurentsova, N. A., & Lyubimov, V. V. (1992). Equatorial segment of the MidAtlantic Ridge as a possible structural barrier between the North and South Atlantic. USSR Academy of Sciences Transactions (Doklady), Earth Science Sections, 312, 133-135. Tuezov, I. K. (1 998). Tectonics, structure, geodynamics and geological nature of the west Pacific active margin, parts I & 2. New Concepts in Global Tectonics Newslettev, 7, 16-23; 8, 20-24. Udintsev, G. B. (Ed.). (1996). Equatorial segment of the Mid-Atlantic Ridge. IOC Technical Series No. 46, UNESCO. Udintsev, G. B., & Koreneva, E. V. (1982). The origin of aseismic ridges of the eastern Indian Ocean. In Scrutton, R. A., & Talwani, M. (Eds.), The oceanfloor (pp. 204-209). Chichester: John Wiley & Sons. Udintsev, G. B., Illarionov, V. K., & Kalinin, A. V. (1990). The West Australian ridge. In BartoKyriakidis, A. (Ed.), Critical aspects of the plate tectonics theory (Vol. 2, pp. 307-341). Athens, Greece: Theophrastus Publications, S. A. Udintsev, G. B., Kurentsova, N. A., Pronina, N. V., Smirnova, S. B., & Ushakova, M. G . (1993). Finds of continental rocks and sediments of anomalous age in the equatorial segment of the Mid-Atlantic Ridge. USSR Academy of Sciences Transactions (Doklady), Earth Science Sections, 312, 111-114. Umbgrove, J. H. F. (1 947). Thepulse of the earth (2nd ed.). The Hague: Martinus Nijhoff. Uyeda, S. (1986). Facts, ideas and open problems on trench-arc-backarc systems. In Wezel, F. C. (Ed.), The origin of arcs (pp. 435-460). Amsterdam: Elsevier. Van Andel, T. H. (1984). Plate tectonics at the threshold of middle age. Geologie en Mijnbouw, 63, 337-341. Van Andel, T. H. (1994). New views on an oldplanet: A history ofglobal change (2nd ed.). Cambridge, UK: Cambridge University Press. Van Andel, T. H. (1998). Plate Tectonics. Encyclopaedia Britannica, CD-ROM, 1994-1998. Van Bemmelen, R. W. (1972). Geodynamic models: An evaluation and a synthesis. Amsterdam: Elsevier. Van der Linden, W. J. M. (1977). How much continent under the ocean? Marine Geophysical Research, 3,209-224. Van der Voo, R. (1998). A complex field. Science, 281, 791-792. Van Hinte, J. E., & Ruffman, A. (1995). Palaeozoic microfossils from Orphan Knoll, NW Atlantic Ocean. Scripta Geologica, 109, 1-63. Voisey, A. H. (1958). Some comments on the hypothesis of continental drift. In Continental drif-A symposium (pp. 162-17 1). Hobart: University of Tasmania. Wanless, R. K., Stevens, R. D., Lachance, G. R., & Edmonds, C. M. (1968). Age determinations ' 352 David Pratt and geological studies. K-Ar isotopic ages, report 8. Geological Survey of Canada, Paper 672, pt. A, pp. 140-141. Wezel, F.-C. (1986). The Pacific island arcs: Produced by post-orogenic vertical tectonics? In Wezel, F.-C. (Ed.), The origin of arcs (pp. 529-567). Amsterdam: Elsevier. Wezel, F.-C. (1988). A young Jura-type fold belt within the central Indian Ocean. Bollettino di Oceanologia Teorica ed Applicata, 6, 75-90. Wezel, F.-C. (1992). Global change: shear-dominated geotectonics modulated by rhythmic earth pulsations. In Chatterjee, S., & Hotton, N., I11 (Eds.), New concepts in global tectonics (pp. 421-439). Lubbock, TX: Texas Tech University Press. Wicander, R., & Monroe, J. S. (1999). Essentials of geology (2nd ed.). Belmont, CA: Wadsworth Publishing. Wyllie, P. J. (1 976). The way the earth works. New York: John Wiley & Sons. Yano, T., & Suzuki, Y. (1999). Current hypotheses on global tectonics. New Concepts in Global Tectonics Newsletter, 10, 4. Zoback, M. L., Zoback, M. D., Adams, J., Assumpq50, M., Bell, S., Bergman, E. A., Bliimling, P., Brereton, N. R., Denham, D., Ding, J., Fuchs, K., Gay, N., Gregersen, S., Gupta, H. K., Gvishiani, A., Jacob, K., Klein, R., Knoll, P., Magee, M., Mercier, J. L., Miiller, B. C., Paquin, C., Rajendran, K., Stephansson, O., Suarez, G., Suter, M., Udias, A., Xu, Z. H., & Zhizhin, M. (1989). Global patterns of tectonic stress. Nature, 341, 291-298. Zorin, Yu. A., & Lepina, S. V. (1989). On the formation mechanism of postrift intracontinental sedimentary basins and the thermal conditions of oil and gas generation. Journal of Geodynamics, 11, 13 1-142. Journal of Scientific Exploration, Vol. 14, No. 3, pp. 353-364, 2000 0892-33 10100 O 2000 Society for Scientific Exploration The Effect of the "Laying On of Hands" on Transplanted Breast Cancer in Mice WILLIAM F. BENGSTON St. Joseph S College, Patchogue, NY 11772 e-mail: wbengston @sjcny.edu 2216 NE Douglas Street Newport, Oregon 97365 Abstract- After witnessing numerous cases of cancer remission associated with a healer who used "laying on of hands" in New York, one of us (W.B.) "apprenticed" in techniques alleged to reproduce the healing effect. We obtained five experimental mice with mammary adenocarcinoma (code: H2712; host strain: C3HlHeJ; strain of origin: C3H/HeHu), which had a predicted 100% fatality between 14 and 27 days subsequent to injection. Bengston treated these mice for 1 hour per day for 1 month. The tumors developed a "blackened area," then they ulcerated, imploded, and closed, and the mice lived their normal life spans. Control mice sent to another city died within the predicted time frame. Three replications using skeptical volunteers (including D.K.) and laboratories at Queens College and St. Joseph's College produced an overall cure rate of 87.9% in 33 experimental mice. An additional informal test by Krinsley at Arizona State resulted in the same patterns. Histological studies indicated viable cancer cells through all stages of remission. Reinjections of cancer into the mice in remission in Arizona and New York did not take, suggesting a stimulated immunological response to the treatment. Our tentative conclusions: Belief in laying on of hands is not necessary in order to produce the effect; there is a stimulated immune response to treatment, which is reproducible and predictable; and the mice retain an immunity to the same cancer after remission. Future work should involve testing on various diseases and conventional immunological studies of treatment effects on experimental animals. Keywords: mammary cancer remission - cancer remission - healing laying on of hands -healer -alternative medicine. Introduction Researchers who have studied psi phenomena in general and healing in particular have been plagued by the apparent unreliability of the phenomena. Added to this problem are unresolved questions of the role and necessity of belief, and the question of whether subjects can be taught to produce significant effects. Our research on healing addresses these issues and appears to present a reliable and potentially efficacious production of healing taught to, and produced by, nonbelieving subjects. 354 William F. Bengston & David Krinsley Previous Research There is a growing body of research into what has been variously termed "anomalous" or "paranormal healing," "healing with intent," "spiritual healing," "Therapeutic Touch," and "laying on of hands," to name but a few. There are by now so many terms used interchangeably that some researchers do not even distinguish among them (Bunnell, 1999). Compilations of controlled studies (Benor, 1992; Murphy, 1992) often cluster previous work by the subject of the intended healing. Benor (1992), for example, discusses healing action on enzymes, cells in the laboratory, fungilyeasts, bacteria, plants, singlecell organisms, animals, electrodermal activity, and human physical problems among the areas that have been submitted to controlled scientific study. And although some studies in each of these areas have produced significant results, they are nevertheless dogged by unresolved questions of reliability. Benor (1992) reports that psi phenomena in general tend to be demonstrated only in the first series of experiments but not in attempted replications. In healing research, for example, Snel (1980) reported significantly inhibited growth of mouse leukemia cells in tissue culture, but not on attempted replication. Even thoroughly studied and practiced healer Oskar Estebany has shown an inability to reproduce significant effects of trypsin in vitro when he was personally not at ease (Smith, 1972). Healers are reported to be unable to produce sufficiently consistent results, frustrating some researchers into claiming that it is virtually impossible to establish a repeatable experiment in which healing occurs in the same combination more than once (Benor, 1990). And because healing does not conform even statistically in regularly repeatable fashion, skeptics can argue that claims for healing efficacy probably present chance variations, rather than responses to healing treatment (Benor, 1990). The question of the role of belief in producing psi phenomena has also been debated. The well-known sheep-goat effect (Schmeidler, 1945; Schmeidler and Murphy, 1946; Palmer, 197 1) does not seem to carry over consistently into healing. Many presume that some sort of faith is a requisite for positive results in healing, and some even argue that skeptical observers may inhibit the effectiveness of the treatment (Benor, 1990). But there is not agreement on the issue. Grad (1961) found that skeptical medical students treating cages of mice produce slower healing than the untreated control group. On the other hand, Krieger's study of Therapeutic Touch (1979) found that belief in the effectiveness of healing does not affect its success. There is also disagreement about whether healing can be taught or whether it is entirely an innate ability. Reviewing some biographies of well-known healers, Benor (1992) notes that the Russian healer Yefin Shubentsov believes it is a physiological process that can be taught to anyone. Dolores Krieger's Therapeutic Touch method claims that to become a healer, a person must have clear intentions, motivation to help, and an ability to understand personal motivation for wanting to heal. On the other hand, both Oskar Estebany and Olga Worrell felt that healing could not be developed by study. Experimentally, Effect of the "Laying on of Hands" 355 though, Nash (1980, 1984) found that subjects not known to be healers could be taught to significantly affect the growth of bacteria in cultures. Our research addresses these issues of reliability, belief, and "teachability" in healing experimental mice, and by extension, the question of the efficacy of healing is raised. Certainly all healing work in experimental animals must acknowledge the pioneering work of Grad (1961, 1965, 1976), who set the stage and the standards for work in this area. In the first controlled experiments, Grad studied the Hungarian healer Oskar Estebany's ability to accelerate the healing rate of mice with one-half by one-inch wounds. Estebany held the cages of mice twice daily for 15 minutes. The treated group healed significantly more rapidly than the untreated group (Grad, 196 1). Grad also induced goiters in mice by feeding them an iodine-deficient diet (Grad, 1976). The thyroid glands of mice treated by a healer twice daily for 15 months grew significantly more slowly than those of the control mice. This effect was also obtained when Estebany did not treat the mice directly with his hands, but instead held in his hands cotton cuttings, which were placed in contact with the mice in the cages. There has been little work done in the area of cancer in live animals. Onetto and Elguin (1966) experimented in inhibiting tumor growth in mice that had been injected subcutaneously with a tumoral suspension. They found the area, weight, and volume of tumor growth in one group of 30 tumorigenic mice was significantly less than that of 30 untreated control mice. Interestingly, a second group of 30 mice was treated in an attempt to increase tumor growth, but these mice did not differ from the control mice. Null (198 1) gave 50 healers two mice each to screen for their ability to prolong the lives of mice injected with cancer cells. Only one healer produced total tumor regression in one of his mice, and the other survived longer than predicted. The one successful healer was then asked to twice replicate the healing on 10 mice. The healer was able to extend the average survival of the treated mice to a statistically significant number of days beyond that of the control group (Grad, 1976). The Present Research Our research grew out of an attempt to empirically test a New York-based healer. This individual claimed that without the benefit of study or training, he was naturally able to perform psychometry, or token object reading, as well as healings by laying on of hands. Over the course of several years, Bengston watched hundreds of people being treated for conditions covering a wide range of afflictions. Some conditions such as long-term diabetes seemed to respond slowly while others such as cancer appeared to respond almost immediately. Among the most interesting observations was that the entire process did not involve belief of any sort. The person being healed was not asked to believe in anything, and the healer himself did not espouse belief. Truly, the healings could be considered "faithless" on the part of all concerned. Anecdotally, it 356 William F. Bengston & David Krinsley appeared that nonbelievers responded more dramatically than believers. Conventional medical tests determined the success or failure of treatment. Over the course of months of questioning by Bengston about the process by which the healer was able to treat others, techniques were developed wherein the healer claimed that complete skeptics could be trained to reproduce healing effects. The techniques did not involve belief of any sort, nor did they include meditation, focused visualization, spiritual discipline, or lifestyle changes. The initial techniques involved a series of routine mental tasks that were not directly intended to produce healing. Subsequent to mastery, these would be followed by laying on of hands. The mental techniques required several weeks of practice to achieve sufficient mastery to move to the laying on of hands techniques. We wanted to test the authenticity of the healings in a rigorously controlled setting that would allow for completely unambiguous results. For obvious reasons, "clinical" work with humans involves less than ideal conditions, so we asked the healer if he would come into a laboratory and attempt to treat laboratory animals. He initially agreed, but just before the start of our experiment, he refused to participate. At that point, Bengston, who had apprenticed the longest and spent the most time learning the techniques, reluctantly became the first experimental healer. Methods and Data The First Experiment Krinsley was a professor at Queens College of the City University of New York. He had arranged for a disinterested professor of biology who was doing conventional cancer research to prepare experimental animals. Her area of expertise was mammary cancer, so she was familiar with mammary adenocarcinoma and obtained from The Jackson Laboratory a "standard" mammary adenocarcinoma (code H2712; host strain C3JlHeJ; strain of origin C3HIHeHu). The normal progression after the mouse is injected is the development of a nonmetastatic palpable and visible tumor that grows so large that it crushes the internal organs of the host. Host survival in the conventional literature was 100% fatality between 14 and 27 days after injection. The experimental procedure was planned as follows: Bengston was to place his hands around the outside of a standard laboratory plastic cage containing six mice for 1 hour per day while applying the healing technique, beginning 3 days after injection. At no time were the mice to be directly touched. Six control mice were kept in a separate laboratory in the same building. One experimental mouse died of natural causes before treatment began, so only five mice were actually treated. Our initial hope was that we might get a significant difference in survival between the experimental animals and their controls. Remission was not seriously considered. Our results were totally beyond expectation. About 10 days into the procedure, the experimental mice began to develop a "blackened area" on their tu- Effect o f the "Laying on of Hands" Fig. I . Typical mouse 14 days after rnlcctlon. rnors (Figures 1 and 2). At this point, Beng\ton presumed that the experiment was failing and wantcd to call it off. Krinsley convinced him to continue, reasoning that there was nothing to lose. Approximately I week later, the blackened areas "ulcerated" as if they had bccn split open (Figures 3 and 4). In some Fig. 2. Twenty-two days after 111jectlor1 358 William F. Bengston & David Krinsley Fig. 3. Twenty-eight days after injection. cases, the ulceration grew extremely large (Figure 5 ) then appeared to implode (not shown), and the wound closed. The mice then lived their normal life span of approximately 2 years. T the figures, the index card notation "A-3" identin fies the mouse, and the day number indicates elapsed time since injection. In Fig. 4. Thirty-five days after injection. Effect of the "Laying on of Hands" 359 Figure I (Day 14), the tumor is visible o n the left posterior dorsal aspect of the mouse. On Day 22 (Figure 2 ) , the tumor is clearly larger but has developed an encrusted area on i t \ surt'ace (nlost posterior aspect of the tumor). This is the earregression. liest indication of lu111o1Days 28, 35, and 38 (Figures 3 through 5 ) illustrate the next significant stage. The tumor appears to be resorbed internally and remains clear of infection. From this stage on, the tumor regresses completely (not shown), and the mouse lives its norsilal life span. The control mice presented us with sorne llniyue challenges. In the initial stages o f developing the experimental procedure, the healer warned that he could not be near or see the control mice, or they, too, would go into remission. Although skeptical, we agreed to keep the control mice in another laboratory. When Rengston became the substitute healer, we relaxed this protocol. After two control mice had died "on schedule '-- that is, between 14 and 17 days after ir~jectior~-Bengsttrr? went to see the remaining four. They exhibited normal tumor progression patterns and were obviously in the last stages of the disease. However, after Hengston observeci the four control mice in their cage, several days later, they too developed the blackened area, the tumor ulcerated, and the mice went into full remission, although they lagged behind the regularly treated experimental mice in remission rate. 7 The Set-ond Experitucrlt The results of our first experiment clearly amazed and confounded us, and we immediately set out to replicate the procedure. Krinsley offered to try the tech- 360 William F. Bengston & David Krinsley nique, and he solicited a skeptical faculty volunteer from Queens College who had neither belief nor experience with any sort of paranormal phenomena. Bengston approached a half dozen students at St. Joseph's College to act as volunteers and selected the two most skeptical students to serve as healers. The two students also had no previous experience with anomalous healing phenomena, did not believe in the legitimacy of healing, and reported afterward that they believed Bengston was actually conducting a study on student gullibility. Bengston trained the four volunteers for several hours once a week for 6 weeks. Between training sessions, the volunteers were assigned practice mental tasks. Each volunteer was then given one cage with two mice. One experimental mouse died of natural causes 2 days after treatment began, reducing the actual number of experimental mice to seven. There were six control mice in an adjacent laboratory in the same building. All seven experimental mice developed the remission pattern and lived their normal life span. Without our knowledge, and despite warnings to not do so, after two control mice had died, the faculty volunteer at Queens College began daily observations of the remaining four. All four of the remaining control mice then went into remission. The Third Experiment This experiment produced the most puzzling results. We moved the study to the Brooklyn campus of St. Joseph's College, where we convinced Carol Hayes, the extremely skeptical chairperson of the biology department, to perform the procedure. She agreed as long as she could pick some volunteers. She selected three undergraduate biology majors; Bengston selected two volunteer healers: one undergraduate sociology major and one child study major. As in the second experiment, all participants had no previous paranormal experiences and were nonbelievers in the legitimacy of laying on of hands. They were trained by Bengston in an identical manner to the second experiment. In this run, we attempted to solve the problem of control remissions and to find out if every volunteer could individually produce remissions. Thus, each volunteer was given one mouse to treat in the laboratory and one mouse to treat at home. We reasoned that because exposure to a trained individual appeared to produce remissions in previous experiments, in order to test whether each individual was effective, he or she had to remit their "private" mouse at home. It followed from the previous results that if any student could produce remissions, then all five experimental laboratory mice should also remit. There were two control groups. The first was a cage with six mice in an adjacent laboratory in the same building. The second control group was a cage with four untreated mice sent to a laboratory in another city known only to the experimental biologist. The results of this run have frustrated attempts to discern a pattern. All five experimental mice taken home by the students remitted. But in the laboratory, all three of the experimental mice treated by the biology majors died within the Effect of the "Laying on of Hands" 361 expected time frame. Only the sociology and child study majors were able to remit their mouse in the laboratory. Even as the biology majors' experimental mice died, the biology students began to look in on the control mice in the adjacent lab after three had died, and the remaining three control mice remitted. All four of the control mice sent to a distant city died well within the expected 27-day maximum. The Fourth Experiment The fourth experiment took place entirely in a laboratory at the Brooklyn campus of St. Joseph's College. As in the third experiment, the mice were prepared and monitored by Carol Hayes, the chairperson of the biology department. In the previous experiments, any mouse that developed a "blackened area" followed our now "classic pattern" to full remission. None of the mice that died developed this blackened area. As such, we were now confident that about 3 weeks after injection, we could spot with certainty which mice would be completely cured and which would die. Thus, we decided to repeat the experiment; this time sacrificing all the mice 38 days after injection. We selected 38 days because at this point, some of the mice would still have large ulcerations and some would have already closed wounds and be fully remitted. Six student volunteers were given a cage with two mice each. One of the students was a biology major who had failed to remit his mouse in the laboratory in the third experiment and wanted to try again; two more students had been volunteers in a previous run and wanted to do it again out of sheer disbelief at the results they had obtained. Bengston chose three new and skeptical volunteers who were not biology majors. Unknown to us, the experimental biologist elected to not inject one mouse to observe any behavioral changes (there were none); one mouse was given two separate injections. In all, there were 11 viable experimental mice with cancer. Eight mice on site served as the first control group. And four mice were sent to a laboratory in another city to serve as the second control group. Ten of eleven of the experimental mice were in various stages of the remission process when they were sacrificed on Day 38 after injection. One experimental mouse never developed the blackened area and died on Day 30 after injection. Although 30 days is technically beyond the predicted 100% fatality, within 27 days, we considered this mouse to have not responded to treatment. Seven of eight on-site control mice, which were regularly looked in on by the student volunteers, remitted. All four of the control mice sent to a laboratory in another city died within the expected time frame. After the mice were sacrificed, we sent tissue samples out to an independent laboratory for histological analysis. Viable mammary adenocarcinoma cells were present at all stages of remission. Only those mice whose ulcerations were completely closed were free of cancer. A summary of the four experi- William F. Bengston & David Krinsley TABLE Summary of Remission Patterns Experiment Experiment 1 Experimental mice Control mice on site Experiment 2 Experimental mice Control mice on site Experiment 3 Experimental mice Control mice on site Control mice off site Experiment 4 Experimental mice Control mice on site Control mice off site Overall results Experimental Mice Control mice on site Control mice off site N No. of remissions Remissions (%) Discussion and Conclusions We offer several preliminary conclusions. First, the treatment was successful in curing mammary adenocarcinoma. Second, it is apparent that stated belief in anomalous phenomena in general, or healing through laying on of hands in particular, is not necessary to produce healing of mammary adenocarcinoma in laboratory mice. None of the experimental healers were believers, though as the experiments progressed, they clearly hoped that their mice would live. Despite the attachment they felt toward their mice, most could be considered at least fairly strong skeptics. We cannot generalize that anyone can produce these remissions, since our sample of volunteers was clearly a nonrepresentative convenience sample. Actually, we have never tested whether believers or people who have allegedly experienced or produced previous anomalous phenomena are also able to produce remissions of this type of tumor. Third, given the peculiar situation that biology majors were unable to produce remissions in the laboratory but were able to do so at home (the third experiment), it is also possible that systematic intellectual activity (these students kept scientific logs) is antagonistic to the production of healing effects. Fourth, it seems likely that there is a stimulated immune response to treatment. There are several reasons to draw this conclusion. The discovery in the fourth experiment that there were viable mammary adenocarcinoma cells during tumor remission is consistent with an immune response (though cutting off of the blood supply to the tumor might produce similar effects). Effect of the "Laying on of Hands" Fig. 6. Example ol'large ulceration Some of' the ulcerations grew extre~nely large (Figure 6), yet no mozise ever developed any gros.s signs c?finfection,P V C J ~ ~ t + l ~ t ~ xli ~ e the ulceration alone Z ltzt' t of should huve been enough to kill it. Finally, and perhaps most persuasively, we were unaware that the experimental biologist at St. Joseph's College had reinjected several remitted mice months after the experiments were over. Without ,further treatment, tlzese mice were immune to thr. nlatnmury a&nocarcinoma. Finally, we may conclucie that we are apparently able to cure mammary adenocarcinoma in experimental mice on demand. The reliability of the procedure has been established (we have also produced remissions at Arizona State University, though the results are not reported here). Future Reseclrch We find ourselves in a somewhat unique position, because the bane ofprevio ~ l research into anomalous phenomena ha\ often been the lack of predictabils ity o f results. Several interesting possibilities emerge. In terms of mammary adenocarcinorna, conventional immunological studies of mice in future experiments might yield data suggestive of the process by which remissions occur. Are there any new antibodies or an increased production of antibodies occurririg during tumor regression? If so, long-term goals might include attempting to stimulate the identified immunological reaction without the laying on of hands. In short, the results of these healing techniques might provide useful information about how to reproduce the remissions using more conventional therapies. Other types of cancer need to be studied to see whether they also respond to 364 William F. Bengston & David Krinsley the healing techniques used here. To date, experimentally we have only treated mammary adenocarcinoma. Finally, because we have established such a degree of reliability, the healing techniques can be used to try to discern what is happening between the healer and the animal. Experiments can be designed to shed light on the mechanisms by which anomalous healing actually occurs. For example, in producing remissions in other species of mammals, we have anecdotally observed that the speed of remission is a function of the size of the animal. Larger animals remit more slowly than smaller ones. Are the different response rates due to metabolism rate, the mass of the animal, or the spreading of a healing energy over a wider area? Would 50 mice simultaneously treated remit at a slower rate than 25? The possibilities for research are almost endless. Acknowledgments We gratefully acknowledge the input of Bernard Grad, Eugene Carpenter, Ted and Diane Mann, Joann LaScala, and Daniel J. Benor. References Benor, D. (1990). A psychiatrist examines fears of healing. Journal of the Society for Psychical Research, 56, 287-299. Benor, D. (1992). Healing research-Research in healing (Vol. 1). Munich, Germany: Helix Editions. Bunnell, T. (1999). The effect of "healing with intent" on pepsin enzyme activity. Journal of Scientific Exploration, 13(2), 139-148. Grad, B., Cadoret, R. J., & Paul, G. I. (1961). The influence of an unorthodox method of treatment on wound healing in mice. International Journal of Parapsychology, 3, 5-24. Grad. B. (1965). Some biological effects of laying on of hands: A review of experiments with animals and plants. Journal of the American Society for Psychical Research, 59, 95-1 27. Grad, B. (1976). The biological effects of "laying on of hands" on animals and plants: Implications for biology. In Schmeidler, Gertrude (Ed.), Parapsychology: Its relation to physics, biology, psychology and psychiatry (pp. 76-89). Metuchen, NJ: Scarecrow Press. Krieger, D. ( I 979). The therapeutic touch: How to use your hands to help or heal. Englewood Cliffs, NJ: Prentice-Hall. Murphy, M. (1992). The future of the body: Explorations into the further evolution of human nature-Chapter 13. Los Angeles: Jeremy Tarcher. Nash, C. (1982). Psychokinetic control of bacterial growth. Journal of the Society for Psychical Research, 51, 2 17-221. Nash, C. (1984). Test of psychokinetic control of bacterial mutation. Journal of the American Society for Psychical Research, 78, 145-1 52. Null, G. ( 198 1). Healers or hustlers? Part 1V. SelfHelp Update, Spring. Onetto, B., & Elguin, G. (1966). Psychokinesis in experimental tumorigenesis. Journal of Parapsychology, 30, 220. Palmer, J. (197 1). Scoring in ESP tests as a function of belief: The sheep-goat effect. Journal of the American Society for Psychical Research, 65, 373-408. Schmeidler, G. (1945). Separating the sheep from the goats. Journal of the American Society for Psychical Research, 39, 47-50. Schmeidler, G., & Murphy, G . (1946). The influence of belief and disbelief in ESP upon ESP scoring level. Journal of Experimental Psychology, 36, 271-276. Smith, J. (1972). Paranormal effects on enzyme activity. Human Dimensions, 1, 15-19. Snel, F. (1980). PK influences on malignant cell growth. Research Letter No. 10, Parapsychology Laboratory, University of Utrecht. Journal of Scientific Exploration, Vol. 14, No. 3, pp. 365-382,2000 0892-3310100 0 2000 Society for Scientific Exploration The Stability of Assessments of Paranormal Connections in Reincarnation-QpeCases Division of Personality Studies, Department of Psychiatric Medicine University of Virginia, Charlottesville, VA 22908 e-mail: [email protected] Department of Psychology University of Tasmania, G.PO. Box 252-30, Hobart, Tasmania, Australia Abstract-The phrase "case of the reincarnation type" refers to any child who has information and/or other characteristics that suggest a paranormal connection between himself and a particular person who has died before he was born. By a "paranormal connection," we mean the communication of information without the recognized sensory channels. Initially unintentionally, several cases previously studied by one of us (I.S.) were investigated by the other (J.K.) about 20 years after I.S.'s first investigation. Additional cases were then intentionally reinvestigated to test the stability of the paranormality assessments of 15 cases. All but one of the cases investigated 20 years later received the same or a lower paranormality rating. These results support the view that such case studies, even if carried out 2 decades after the relevant events occurred, do not generate inflated paranormality assessments. Keywords: paranormal phenomena -reincarnation -memories Introduction For the last 40 years, investigators have been examining cases in which a child is said to remember a previous life. For most of the cases that were investigated by one of the authors (I.S.) (Stevenson, 1974) and later by Mills (1989), Pasricha (1990), Haraldsson (199 I), and the other author (J.K.) (Keil, 199 I), interviews were conducted several years after the most important events had taken place. The "most important events" are nearly always ones that suggest the transmission of information without the usual sensory channels, i.e., paranormal processes. Some cases that seem important examples suggesting paranormal processes did not come to the attention of the above investigators until more than a decade had passed. The question must be asked whether in spite of such long intervals, relevant information remains sufficiently stable, or whether these case studies become significantly distorted, and in particular, whether informants, with the passage of time, tend to emphasize more heavily aspects of a case suggesting paranormal processes. 366 Ian Stevenson & Jiirgen Keil In an attempt to answer this question, 4 cases in Thailand, 10 in Turkey, and 1 case in Myanmar were investigated after (on average) more than 20 years. The study developed initially from the accidental reinvestigation of some cases by J.K. that had already been studied by I.S. about 2 decades earlier. J.K. only became aware of this after he had more or less completed his investigations. He found that the informants did not seem to be influenced by the earlier investigations, and some of them mentioned only toward the end of the interviews that an American professor had conducted a similar study years earlier. We then decided to reinvestigate some further cases to compare the results of 15 cases. It could be argued that we should be primarily concerned in this paper with the study of memory over time. However, many aspects of memory, although relevant to the comparison of paranormality assessments, cannot be readily evaluated. A full appraisal of the memories of informants for these cases would require a larger series of cases than we report here. Nevertheless, we will cite, in a later section of this paper, some of the relevant literature on long-term memory. For cases of the reincarnation type, the paranormality components in the connection between a person who has previously died (the previous personality) and a child (the subject) is the most important aspect. Although for most case studies, the strength of these components is not directly assessed; it is almost always assumed that such components may be present. Otherwise, the case report would not really be relevant to the reincarnation hypothesis. As will be outlined in the Discussion section, an objective rating scale for paranormality in these cases should be developed, but we have not done this yet. The rating scores reported here are based on our more subjective evaluations. By presenting in some detail half of the cases that were compared, we hope that readers can make their own assessments. Memory Aspects Although the main feature of this study is the stability of judgments about paranormal components of a case, we cannot isolate this question from the aspects of memory that are inevitably involved when informants report events that happened many years earlier. It must be acknowledged that memories generally diminish over time. Bartlett (1932) also showed that informants may add new material without being aware of these changes. On the other hand, Prince (1918, 1919) reported two instances in which accounts of experiences written 15 and 20 years apart differed in only one detail. Similarly, Parsons (1962) found that a description of a building written 6 years after it was seen was correct in 18 out of 21 details. Two studies of reports of experiences near death are directly pertinent to the question of the embellishment of reports with the passage of time. Greyson (1983) developed a scale for such experiences that indicates by increasing Paranormality in Reincarnation Cases 367 scores the complexity-i.e., the amount of detail-in a report. He found no significant correlation between the score on the scale, indicating the complexity of the report, and the elapsed time between the experience and its report. Using data from a different series of near-death experiences, Alvarado and Zingrone (1 997198) also found no significant correlation between complexity of the report of an experience and the elapsed time between the experience and its report. Vehicular accidents are often regarded as examples that show that the memories of several witnesses may differ substantially. However, vehicular accidents contain sequences of events that happen quickly and then end. Although our case studies may also include some dramatic sequences, such as when someone died in an accident or was killed intentionally, the events that were reported to us were already relatively settled and fixed in the minds of our informants before the first investigations were carried out. The memory aspects of our case studies are similar to what are termed "autobiographical memories." Baddeley (1998) reviewed the evidence for the accuracy of autobiographical memories and concluded that normal people can recall earlier events in their lives reasonably accurately over a long interval. Distortions and omissions can occur due to emotional factors and the perceived unpleasantness of the events. Disagreements about the recollections of accidentsapart from distortions due to personal involvement-are probably often due to the relatively large number of relevant details, which are perceived during a short period. Cases of the reincarnation type are not always regarded by the families involved as definite manifestations of rebirths of previous personalities. In this sense, they differ from the happenings associated with such definite events as marriages and funerals. Nevertheless, events associated with the assumed rebirth of a child are, in many cases, discussed by family members, friends, and more distant relatives quite frequently. An investigation of events that took place, say, 15 years ago does not usually involve obtaining information from people who have not talked about these events since they happened. Discussions of the events among family members will tend to enhance retention and accessibility of the informants' memories for them. Method For the first cases to be compared, J.K. had not seen I.S.'s notes before he conducted his own investigation. In the second phase of the project, I.S. furnished J.K. with the names and addresses of subjects whose cases he had investigated many years earlier. J.K. then went to the informants for these cases and, conducting new interviews, made records of what the informants then remembered about the cases. J.K. did not read I.S.'s notes before he made his investigations. He had perhaps read reports of some of I.S.'s cases that had been published, but he had retained in his conscious memory no details of these that influenced his investigations. When J.K. had completed his inquiries, we 368 Ian Stevenson & Jiirgen Keil compared the two accounts for the following differences: loss of essential details (in the J.K. reports compared to the I.S. ones), distorted accounts to J.K. of what were obviously the same events described earlier to I.S., and mention to J.K. of important details not told to I.S. We also noted reports of events to J.K. that differed in no essential detail from reports of the same events to I.S. Because we were primarily interested in the stability of information relevant to the assessment of paranormal processes, J.K. concentrated on interviews. He did not extend his inquiries, as I.S. often did, to the study of relevant documents, such as hospital records and postmortem reports. As J.K. had learned in connection with other cases in Turkey, such records are nearly always discarded within a decade. In Thailand, termites frequently produce the same outcome. Because of the relatively small number of cases and relatively large number of uncontrolled variables, we did not attempt to compare specific memory aspects. We tried to keep them in mind, however, when for each investigation we rated the strength of the paranormality component on a scale of 1 to 10. For these ratings, we assigned the descriptive values listed below to coordinate I.S.'s and J.K.'s assessments and to provide the reader with an illustration of how we judged the strength of each case. (However, comparisons between the earlier and later investigations could also be carried out with a different set of descriptive values.) A rating of 1 means no suggestion of paranormal processes. A rating of 2 to 3 means there are indications that a paranormal component may be involved, but the probability is equally high that the apparently paranormal features could be due to chance or other normal processes. A rating of 4 to 7 indicates that paranormal processes increasingly outweigh alternative explanations without reaching the equivalent of the .05 significance level. A rating of 8 means the presence of a paranormal component is suggested as being equivalent to the .05 significance level. A rating of 9 is similar to a rating of 8, but at the .O1 significance level. A rating of 10 is similar to 9, but at the .OO1 significance level. In mentioning figures of probability, we do not mean to minimize the subjectivity of our judgments about the element of paranormality in the cases. Moreover, it is not possible to list, in a kind of hierarchical order, conditions that would help to exclude normal communications. Nevertheless, the following conditions (taken from Keil, 1991) are likely to strengthen a claim for the presence of paranormal aspects. In this list and elsewhere, the letter S stands for the subject, PP for the previous personality, and PL for the (claimed) previous life. 1. The S and the S's family do not know and have no contact with the PP's family until the S has made definite and potentially verifiable verbal statements. Paranormality in Reincarnation Cases 369 2. The S makes numerous unambiguous statements, which are relatively independent of each other, about a PP and a PL that can be verified. 3. The S provides information about something not known to anybodyexcept, in the past, to the PP-that can be verified, such as some item hidden by the PP and recovered by the S. 4. Statements made by the S-preferably before the PP's family is involved-are noted by more than one member of the S's family and preferably by other people not belonging to the family. 5. Without any opportunity to learn or imitate, the S is able to do something that corresponds to some activity the PP was able to perform. For example, the S is able to speak a language or dialect that is not spoken by people with whom the S had contact. but 6. Similar to (9, the S has some opportunity to learn and imitate. Nevertheless, the S seems far more proficient than would normally be expected. 7. The S has birthmarks and/or malformations that correspond to injuries or other peculiarities of the PP. (Several birthmarks on the S that correspond to specific injuries sustained by the PP are more impressive than a single birthmark that corresponds imprecisely to a large injury on the PP.) 8. Unusual behavior, such as phobias or preferences of the S manifested at an early age, that do not make sense in terms of the S's experiences in his or her life, but that correspond to some important event in the PP's life or feature of the PP's character. 9. Although more difficult to evaluate, the intensity and spontaneity with which statements are made and emotions are expressed also have a bearing on the assessment. This list could be extended. We can also integrate its items, to some extent, if we remember that there are two main criteria in operation on the basis of which we can assess paranormality. First, we need to appraise the number and the complexity of relatively independent statements and other characteristics that correspond to verifiable statements, features, and facts associated with the PP; a small number of statements based on fantasies could, by chance, agree with some aspects of the PP's life. In addition, we need to evaluate the barriers-geographical and social-that may have made it impossible or at least unlikely that seemingly paranormal connections between the PP and the S occurred by normal means. For the relatively subjective evaluation of the reports, we had to take account of the following difficulties: 1. For several cases, J.K. could not interview the same informants who had provided information to I.S. Some of I.S.'s informants had died, some had moved and could not be contacted, and some could no longer remember anything related to the relevant events. 370 Ian Stevenson & Jiirgen Keil 2. Additional information recorded by J.K. (compared with statements recorded by I.S.) may be due to (a) embellishments; (b) differences in the time available for an interview (the informants had the information but did not refer to it); (c) different informants; and (d) I.S. investigated the case when some children were still quite young. A child may have made additional statements after I.S.'s visit, and these may have included paranormal information. A comparison of the two versions of the normal factual aspects of a case may provide some reassurance when similar sets of data were obtained. However, minor discrepancies and/or omissions among these data are generally relatively unimportant unless they have a bearing on the paranormality assessment. For instance, if during J.K.'s interview, the informants could not remember the names of some relatives that were recorded by I.S., we did not regard this as a serious sign of instability unless the names were relevant to statements or events that suggested some paranormal process. For most cases, we agreed on the paranormality ratings; if we disagreedusually by 1 point on the scale-we discussed the reports until we reached agreement. Principal Features of 15 Cases Selected for This Comparison Of the 15 cases we compared, 10 were from Turkey, 4 from Thailand, and 1 from Myanmar (formerly Burma). Five of the subjects were female and 10 were male. There were no cases of the sex-change type in the group, i.e., cases in which the subject claimed to remember the life of a person of the opposite sex. The I.S. cases were first investigated between 1966 and 1984 (median: 1971). The J.K. cases were first investigated between 1990 and 1996 (median: 1995). In both groups, the "first investigation" was typically followed by further interviews in later years when we wished to interview additional informants or talk again with an earlier informant. The mean interval between the first investigations of I.S. and J.K. was 22 years, and the median interval was 23 years. Particularly with I.S. and sometimes with J.K., the investigations continued for several years or even longer after the initial inquiries began an investigation. For example, I.S. began investigating the case of Bongkuch Promsin in 1966, but he had further interviews in the 1960s and 1970s, and he last met Bongkuch and his family in 1980. J.K. began his investigation of this case in 1995, only 15 years after I.S.'s last contact with the family. There was thus a much shorter interval for many cases between I.S.'s last contact with the case and J.K.'s first one. It remains true, however, that in most cases, I.S. obtained the substantial information about the case when he first met the families concerned. Paranormality in Reincarnation Cases TABLE Paranormality Scale Ratings Rating of I.S. investigation Niirsel Karaali (Turkey) Cengiz Elma (Turkey) Zeynep Emel Celik (Turkey) Dellil Beyaz (Turkey) Semihe Atasoy (Turkey) Necati Caylak (Turkey) Nasir Toksoz (Turkey) Cemal Kurt (Turkey) Faris Yuyucuer (Turkey) Cemil Fahrici (Turkey) Ratana Wongsombat (Thailand) Bongkuch Promsin (Thailand) Anurak Sithipan (Thailand) Chanai Choomalaiwong (Thailand) Myint Myint Zaw (Myanmar) Mean: Note: I.S. = Ian Stevenson; J.K. = Jiirgen Keil. Rating of J.K. investigation 371 Results Summary The Table lists the 15 cases we compared and gives the ratings on the paranormality scale for the data of the two investigations. The mean ratings (rounded) on the scale were 6.5 for the data of the I.S. investigation and 5.5 for those of the J.K. investigation. In only one case did we assign a higher paranormality rating for the data of the J.K. investigation than for those of the I.S. investigation. This was the case of Anurak Sithipan. The informants mentioned to J.K. an incident in which Anurak showed paranormal knowledge of an object that had belonged to the PP in the case, his older deceased brother. They had not mentioned this item to I.S. Eight Case ~ e ~ o r t s ' The Niirsel Karaali Case (Turkey). The subject, Niirsel, was regarded as the rebirth of the previous personality, Vesile Gorur. In 1970, I.S. obtained some preliminary information about this case from the PP's younger brother Cemil Ian Stevenson has published detailed reports of his investigations of most of the cases reviewed here. For the benefit of readers wishing details of what he learned about these cases, we have given the references to his reports. 372 Ian Stevenson & Jurgen Keil Goriir. In 1973, one of I.S.'s collaborators interviewed the PP's father, Cabir Goriir, and in 1975, the S 's father, Siileyman Karaali. In 1975, I.S. interviewed a musician, Hasan Eyup Demirel, as well as the S's mother, Vesile Karaali, and Nurse1 herself. In 1977, I.S. interviewed the PP's father, Cabir Gorur. The PP's mother was also present during this interview and volunteered some information. The PP died at the age of 16 in 1962. The S was born in a different village in 1963. The two villages were 18 kilometers apart. The two families were not related and the parents of the S and the PP had no contacts. The S's grandfather, however, did know the PP's father from their military service. When the S became able to speak, she told her father that he was not her father and that her father was in another village, which she named. She stated some names of other members of this family and threatened to run away if she were not taken to them. In 1990, J.K. interviewed the S, her husband, and her mother, as well as two siblings of the PP. I.S. and his Turkish collaborator obtained more details than J.K., but major events were presented in a very similar way in the 1970s and in 1990. Once contact was established between the families from the information the S provided, she continued to visit the PP's relatives. In 1990, she remembered that as a young child, she knew the names of the PP's relatives. (This had been included in I.S.'s notes.) In the meantime, the S had seen the PP's relatives so often that she knew their names because of these contacts. Both I.S. and J.K. were told that at birth, the S's fingers were colored in agreement with the henna color that was applied to the PP's hands after the PP had died. I.S. was told that in about 1967 (at the age of 4%), the S recognized (in her village) some musicians who had played at the PP's funeral. This was not mentioned during the later study of J.K. In 1973 as well as 1990, I.S. and J.K. learned that the S had recognized the PP's old house. In 1977, I.S. was told that the S could correctly sort out the PP's clothes when mixed with other clothes. J.K. learned of this without substantial change in 1990. In 1990, J.K. was told that the S distinguished six of the PP's clothes from 40 other clothes not belonging to the PP but mixed with those that did. I.S. in 1977 had not been given figures such as these but was told only that the "S could distinguish perfectly the clothes of the PP from those of the other members of the family." Was the addition of numbers to the report an embellishment? We cannot say, but the addition, if such it is, does not alter the rating on the paranormality scale. Nursel's correct statements as a young child before contact was established with the PP's relatives cannot be easily explained as normal information and fairly strongly suggest a paranormal connection between her and the PP. The assessments in this respect are similar for the earlier and later investigations. The earlier investigations managed to find more people and details that confirmed the suggested paranormal connections. In 1990, some details were Paranormality in Reincarnation Cases 373 added; these may include some embellishments but may also be factual additions. One detail-the S's recognitions of the musicians from the PP's village-was lost. Generally, these differences did not have a bearing on the assessment of the paranormality component. We rated I.S.'s report as 8 and J.K.'s report as 7. The Delliil Beyaz Case (Turkey). The subject, Delliil, was regarded as the rebirth of the previous personality, Zehide Kose (Stevenson, 1997). The S was born in 1970, within 1 month of the PP's death. She had a birthmark on the top of her head. When the S was born, there was a suggestion, based on a dream, that she might be a rebirth case, but there were no specific indications of this. The S's mother told a collaborator of I.S. in 1975 that she had a dream connecting the S with the PP's village. However, during a second interview later in the same year, the S's mother denied that this dream referred to the PP's village. There was no other information in the dream that could link the S to the PP. In 1975, I.S. interviewed the S, the S's mother, the S's grandmother, as well as the PP's husband, the PP's oldest daughter and her husband, and a nephew of the PP's husband. In 1983, a collaborator of I.S. returned once more to the S with some questions prepared by I.S. In 1991 and 1992, J.K. interviewed the S and the S's mother. The PP's brother Mithat was also present in 199 1. During a further visit in 1994, J.K. interviewed the PP's daughter Nuriye Sen. Most of the information was collected in 1975 by I.S. and in 1991, by J.K. I.S. did not receive any information that the S's relatives had any connections with the PP's relatives. J.K. heard later that the S's relatives had not met the PP, but they did know some of the PP's relatives. There was no contact between the two immediate families when the S was born. Both I.S. and J.K. were told that the S had a birthmark on her head. They also both learned that the S had told her parents how she (the PP) died by falling through an opening from a flat roof while hanging out some washing. (The PP had died of "head injury," a fact relevant to the S's birthmark.) Similarly, informants told both of them that the S had correctly mentioned the name of the PP and the names of two or three of the PP's relatives and the name of the PP's village. The S's father probably knew the PP's village, but the S's mother claimed that she did not know the name of this village, which is probably correct. The S conveyed most of the information related to the PP when she was less than 2 years old. She had started to talk about a previous life as soon as she had learned to talk. The S's statements referring to the PP were correct, and a nephew of the PP, who was visiting a neighbor of the S's family, heard about her statements. This nephew, I.S. was told, arranged the first meeting between the two families when the S was about 2 years old. J.K. did not hear these details, but he was told that the PP's relatives heard that the S was talking about a previous life in 374 Ian Stevenson & Jiirgen Keil a way that suggested that she was referring to the PP. Both I.S. and J.K. were told that the S recognized various people and objects when she visited the PP's home for the first time. The details vary slightly, but J.K. could not interview the same relatives of the PP who had met I.S. There were no signs of embellishment in the more recent accounts of this case. I.S. obtained medical records, which confirmed the death of the PP that informants had indicated to us, but the informants did not clearly link the S's birthmark with the PP's fatal fall. There was, however, agreement that the PP died from head injuries. Although some connections existed prior to the S's birth, it seems unlikely that the information conveyed by the S prior to the first meeting of the two families could have come about by normal means. The various recognitions at a later stage carry some weight but are more doubtful indicators of paranormal processes. We rated the paranormality component as 8 for I.S.'s accounts and 7 for J.K.'s. The Semihe Atasoy Case (Turkey). The subject, Semihe, who was born in 1963, was regarded as the rebirth of the PP, Nesime Dogruel, who died in 1960 at the age of 35 (Stevenson, 1997). The two families are not related, but they were acquainted. They did not meet on a regular basis but occasionally exchanged visits. The PP's husband had shot and killed her. I.S. was told that the S's parents did not know this, and that they only became aware of the circumstances of the PP's death when the S started to talk at the age of 2. This means that for a period of approximately 5 years, the S's parents apparently were not aware that the PP had been killed by the PP's husband. Even if the S's parents met some other members of the PP's family only once or twice a year, it is surprising that they were not aware of the circumstances of the PP's death. Both I.S. and J.K. heard that the S had two birthmarks, which agree with gunshot wounds of entry and exit on the PP. In fact, the PP was shot four times and a further birthmark on the S's elbow (only mentioned to I.S.) may correspond to the wound caused by a bullet. With the help of an autopsy report (only available in 1977), I.S. carefully reconstructed the entry and exit marks on the PP's body. The entry position of two bullets and the exit position of one bullet seem to be in fairly good agreement with the S's birthmarks, but because the PP was hit by four bullets, the agreement may be due to chance. If one shot fired at the PP is regarded as a link that accounts for the S's birthmarks, the question must also be asked why only one bullet and not four resulted in corresponding birthmarks. Based on similar inconsistencies in connection with other cases, I.S. (1997) suggested that events prior to death, such as loss of consciousness by the PP after the first wounding, could explain why only some of the PP's wounds correspond to birthmarks on the S. Nevertheless, the limitations in the correspondence between the PP's wounds and the S's birthmarks must be regarded as a negative Paranormality in Reincarnation Cases 375 feature when the overall probability of a paranormal component for this case is estimated. I.S. visited this case for the first time in 1977. He interviewed the S and the S's parents. He also interviewed the PP's brother Salih. In 1994, J.K. interviewed the S and her father. In 1996, he interviewed the S's mother, who was not present in 1994. The PP's brother Salih had died in 1993. The S still had contact with one of the PP's daughters, but this relationship then seemed fairly independent of the S's connection with a previous life. Nevertheless, the S told J.K. that she still had direct memories of being the PP. J.K. could not find any relatives or friends of the PP who might have been able to provide relevant information about the PP's life. Both I.S. and J.K. heard that the PP was shot when the PP's husband returned home drunk. Only J.K. was told that the PP questioned the PP's husband about his late arrival home. I.S. was told that the PP was shot twice. Only one shot was mentioned to J.K. Both I.S. and J.K. heard that the S had two birthmarks in agreement with the way the PP was shot. When I.S. visited the S in 1977, the birthmarks were still faintly visible. They were not visible in 1994. However, there was no doubt in 1994 (in the minds of the informants) that the S had had two birthmarks (corresponding to the entry and exit wounds of a bullet) when the S was young. I.S. heard of another birthmark on the S's arm, which was not mentioned to J.K. Both I.S. and J.K. were told that the S did not visit any of the PP's relatives until after the S had started to talk about a previous life. I.S. was told that when the S was 10 she met the PP's relatives for the first time. The account given to J.K. in 1994 was less definite and indirectly suggested that the S might have. met the PP's relatives at a younger age. In 1996, the S's mother also said that the S was 10 when this first meeting took place. (J.K. simply asked whether she could remember how old the S was when the S met the PP's relatives for the first time. Ten as a possible age was never mentioned by J.K.) Both I.S. and J.K. were told that the two families are unrelated, that they met only occasionally, and that the connection between the PP and the S was only discovered when the S started to talk about the PP. The reports, mainly obtained 17 years apart, are in good agreement. For this case, it is difficult to estimate a paranormality rating on our scale. The birthmarks do not exactly correspond to only one bullet entry and exit. Although the S's parents did not regard the S as a rebirth case until the S started to talk about a previous life, the two families had had some contacts. The news of the PP's murder may have reached the S even if the S's parents were not aware of it. On the other hand, the S had referred to these events at a very young age. We estimate that a rating of 5 is appropriate for both reports. The Necati Caylak Case (Turkey). The subject, Necati, was regarded as the rebirth of the PP, Abdiilkerim Hadduroglu (Stevenson, 1980). The two families, who lived 8 kilometers apart in different villages, are not related and did 376 Ian Stevenson & Jiirgen Keil not know each other until the S started to talk about a previous life. The PP died in 1963 in a car accident. The S was born about 1 month later according to the information that I.S. obtained. In 1995, J.K. was told that the S was born in 1965. In 1995, when J.K. studied the case, the S was away, working in Saudi Arabia. His relatives (J.K.'s informants) probably only estimated his year of birth. In Turkey, many families do not attach much importance to birthdays and the age of a person is not regarded as important information. I.S. met the S for the first time in 1967, and this meeting confirmed that the S was born around 1963, and not in 1965. I.S. visited relevant informants in 1967, 1971, and twice in 1973. He interviewed the S, the S's parents, and two of the S's older brothers, as well as the PP's wife, the PP's father, the PP's stepmother, two of the PP's sons, five friends and neighbors of the PP (some of them distantly related to the PP), and the driver of the car in which the PP died. In 1995, J.K. interviewed the S's mother and two brothers of the S, Ekrem and Mehmet. Ekrem, who provided most of the information, had also been interviewed by I.S. As mentioned, the S himself was working in Saudi Arabia. In addition, J.K. interviewed the PP's youngest son (also an informant during I.S.'s visits), and the PP's sister Kadife. Both I.S. and J.K. were told that the PP was killed in a car accident at a bridge and that the S talked about this accident, giving correct details, when he was less than 3 years old. Both heard that the S was afraid of the bridge and refused to go near it. When the S came to this bridge for the first time, he also mentioned a number of names including the PP's first name. I.S. heard more details than J.K., which generally strengthened the paranormality hypothesis of this case. However, I.S. also became aware of several inconsistencies that could only be partly resolved through further investigations. Apparently, the S mentioned a wrong family name for the PP. The wrong name may have had some relevance but could not directly be associated with the PP. J.K. was only told that the S mentioned the names of the PP's father, mother, and wife. The two accounts agree with each other, but I.S. obtained many more details supporting a paranormality component. The 1995 report contained no embellishments. The informants failed to mention to J.K. that the S had got the family name of the PP wrong, but most Turks would be satisfied with the S's correct statement of the names of members of the PP's family. In 1995, J.K. recorded a different name for the driver of the car in which the PP died. It is not clear whether J.K.'s informants referred to an alternative name (a distinct possibility in Turkey), whether the same person was remembered as the driver but referred to by a wrong name, or whether in 1995, a different person was regarded as the driver. This difference between the two reports has no bearing on the assessment of the paranormality component. The evidence for this component is fairly strong in I.S.'s report because of the many details and because the two families did not know each other. It is likely, though, that some of the details related to the accident were generally Paranormality in Reincarnation Cases 377 known in both villages. Nevertheless, a paranormality rating of 7.5 for I.S.'s report and 5 for J.K.'s report seems appropriate. The Cemil Fahrici Case (Turkey). The subject, Cemil Fahrici, was regarded as the rebirth of the PP, Cemil Hayik (Stevenson, 1997). He was a well-known outlaw in the Samande, Hatay, region of Turkey. The PP died in 1935. The S was born about 2 or 3 days after the PP's death. The S's verbal account of the PP's life and death is interesting and may contain paranormal elements, but because these events were publicly known, the S may have heard about the PP's life by normal means. The late investigation of this case makes a more detailed assessment of verbal statements with respect to the paranormality hypothesis virtually impossible. When in 1966 I.S. saw the S for the first time, the S was already more than 30 years old. J.K. saw the S 26 years later, in 1994. Nevertheless, the case has some noteworthy features that suggest some degree of paranormality. The S has two birthmarks on his head that correspond to an entry and exit wound of a bullet. These marks appear to be in agreement with what is known about the death of the PP. (The PP shot himself by firing a bullet through his head when he could not escape from the police.) As a child, the S rejected his own name and indicated that he wanted to be called Cemil. The S's parents eventually agreed. The S was afraid of soldiers and policemen and was hostile toward them. As a young child, he pretended to shoot them with a stick. This behavior could also have been encouraged because of some general resentment toward soldiers and policemen, which was not uncommon at that time in the Hatay region of Turkey. Of the 10 informants interviewed by I.S., J.K. could only contact three. Most and perhaps all other informants had died in the meantime. One of these three, the S's sister, had recently lost her son and was too disturbed by this to answer questions. J.K. was able to interview the S and the PP's younger sister. (I.S. had also met this sister of the PP.) She mentioned a few details to J.K. that I.S. had not recorded. Generally, J.K. did not hear as many details as I.S., but J.K. had only two informants. None of the additional details recorded by I.S. or J.K. had any bearing on the paranormality hypothesis. The S's birthmarks, the S's desire as a very young child to be called "Cemil," and the S's reaction to policemen and soldiers were still remembered when J.K. visited the S and the PP's sister in 1994. It seems fair to conclude that the basis for the paranormality assessment did not change after 17 years. The S's birthmarks had faded to some extent-which frequently happensand were less clearly visible when J.K. investigated the case. For this reason, it could be argued that J.K.'s rating should be marginally lower. We agree that if any change is contemplated, it should be in the direction of a slightly lower rating for the more recent report. On the paranormality scale, we rated this case-mainly on the basis of the S's birthmarks-as 7 for the records of I.S. and 6 for those of J.K.. 378 Ian Stevenson & Jiirgen Keil The Ratana Wongsombat Case (Thailand). The subject, Ratana, was regarded as the reincarnation of the PP, Kim Lan (Stevenson, 1983). Kim Lan died in September 1962, at the age of 68. The S was born in May 1964. She started to talk at a very early age, and her apparently accurate memories about some aspects of the PP's life were reported in a Bangkok newspaper in 1967 when the S was 3 years old. I.S. mainly studied this case in 1969, but he visited Ratana again during six further visits to Thailand, the last one occurring in 1980. J.K. studied this case in 1995. The S was raised (from soon after her birth) by her grandmother and stepgrandfather. The latter was particularly interested in Ratana's statements about a previous life. The S's family did not know the PP at all, but the S's step-grandfather knew a few people who previously had some contact with the PP when the PP visited a particular Wat (Buddhist temple) in Bangkok. However, the S's step-grandfather initially was not aware of this connection. I.S. interviewed 13 people including the S, the PP's daughter, Anan, and a nun from the Wat in Bangkok. J.K. also met another nun and a monk who had some limited contact with the PP but who had not been interviewed by I.S. The S's grandmother and the S's step-grandfather had died by 1995. It is likely that some of the other informants interviewed by I.S. had also died by then. Except for the S, J.K.'s other four informants were between 76 and 84 years old. The information provided by the S-when she was a young child-supports the paranormality hypothesis mainly because her family did not know the PP or the PP's relatives, and because the S gave many correct details, including the PP's name when the S was only 2 years old. In 1995, the S herself had no direct memories of the life of the PP, but remembered various details she had later heard from others, including some details that she had mentioned at a very young age. This is not at all unusual in these cases. There are no significant disagreements between the data collected by I.S. and J.K., but between one third and one half of the details was lost when J.K. studied this case. J.K.'s notes do not suggest any additional details that could be regarded as embellishments. Although a reduction of the reported details might be expected because of the advanced age of four of J.K.'s five informants, the paranormality assessment based on the more recent interviews must be rated somewhat lower. On the scale, we rated I.S.'s account as 8 and J.K.'s account as 7. The Anurak Sithipan Case (Thailand). In some areas of Thailand, as well as in Myanmar and parts of India, bodies are sometimes marked, usually shortly after death, but occasionally also before a person is pronounced dead, in order to enable the relatives to recognize a child with a similar birthmark as being the deceased reborn. Although this custom is generally known in areas where Paranormality in Reincarnation Cases 379 I such "experimental birthmarks" (Stevenson, 1997) agree by chance with the original marks. The present case provides an example of this practice. The subject, Anurak, was regarded as the reincarnation of his older brother Chachewan, who had drowned about 3 years before Anurak was born (Stevenson, 1997). Chachewan's body had been marked on the right elbow with charcoal. One informant said the mark was made with ink. There is good agreement between the notes obtained by I.S. in 1977 and those obtained by J.K. in 1994. The two reports agree that the PP's arm was marked near his elbow and that the S was born with a birthmark on the same elbow close to the site where the PP had been marked. On both occasions, the S's relatives claimed that other members of the S's family had no birthmarks like that of the S. According to I.S.'s examination, this is not quite correct, and we therefore have an example of increased emphasis on the birthmarks of the S and a tendency to ignore smaller birthmarks on other siblings. However, these errors were similar on both occasions. The informants for I.S. spontaneously mentioned that the S's birthmark was not exactly in the same position as the mark made on the PP. J.K. was not told of this discrepancy and this could be seen as an unconscious tendency to support the interpretation of reincarnation. On the other hand, after 17 years, the S's parents may have forgotten the small difference in location or may have thought that it was of no further importance. The two reports agree that the S spontaneously searched for the PP's Boy Scout uniform. The more recent statement about this detail provides a somewhat more elaborate framework but does not really add any information that would increase the strength of the paranormality component. There is also agreement about the S's familiarity with the PP's friend who came for a date with the S's sister. The earlier report included more details that could easily have been forgotten after 17 years. The more recent report includes one more incident reported by the S's parents: The S found a special spoon, which the PP had kept on a high shelf in a generally inaccessible place. This incident supports the paranormality component, and it could be argued that in some way, this incident was later invented to provide more support for the reincarnation hypothesis. However, if this incident is considered, together with other statements made by our informants, it seems more likely that the S really behaved in the way his parents reported, and that for some reason, this incident was forgotten or not mentioned when I.S. visited 17 years earlier. Although there are some differences in the records with respect to those events that have a bearing on the paranormality component, we do not regard them as significant. The strength of the paranormality component is broadly similar for the 1977 and 1994 investigations. We agreed that for this case, the paranormality rating for the more recent report should be at least as high as that for the earlier one and possibly higher. We agreed on a rating of 5 for I.S.'s report and a rating of 5.5 for J.K.'s report. 380 Ian Stevenson & Jiirgen Keil The Bongkuch Promsin Case (Thailand). Bongkuch, the subject, was regarded as the rebirth of the PP, Chamrat, who died in 1954, at the age of 18 (Stevenson, 1983). The PP had been attacked from the back, stabbed, and killed. The two families were not related and did not know each other. The PP's relatives had never visited the S's village. The S's father had traveled to the small town where the PP lived. He knew some people there, but not the PP or the PP's relatives, and after the PP was killed, the S's father did not hear anything about the murder. Contact between the two families was established only after the S had started to talk and had made a number of statements about the PP. The PP's relatives then learned about these statements. I.S. carried out a very detailed investigation (mainly in 1966) and interviewed 17 people including the S, the S's parents, 2 sisters of the S, the PP's father, the PP's sister, the PP's stepbrother, the PP's girlfriend, as well as 3 police officers associated with the investigation of the PP's murder. When J.K. investigated this case in 1995, the PP's parents and the S's father had died. J.K. interviewed the PP's half-brother Muan, the younger brother of the PP's mother, the S, the S's mother, and the S's sister Chorthip. I.S. had obtained a much more detailed record of the events, but both he and J.K. found that the first contact phase between the two families suggested a paranormal component. The S had made a number of statements. He mentioned the PP's name, possessions of the PP, such as a bicycle and a gold chain, and how PP had been killed. The PP's relatives confirmed these statements during their first meeting with S. They conducted a "test" by presenting the S with-according to I.S.'s notes-two bicycles or-according to J.K.'s notes-three bicycles, from which the S selected the PP's bicycle correctly. The increase from 2 bicycles to 3 most likely is an embellishment and not a simple error. Both I.S. and J.K. heard that the S used Laotian words not used in the S's family and preferred sticky rice (a distinctly different rice preparation) in agreement with the PP's background, which was Laotian and not Thai. I.S. recorded additional details relevant to the paranormality assessment of this case. For example, his informants told him that the S remembered the names of eight people from the previous life and that he had referred to the PP's watch, gold ring, and knife, and to the way the PP was dressed when he was killed. There was some disagreement among the three police officers about whether the S's statement that the PP wore shorts was correct. There is also less certainty about whether such statements as the names of eight people were made before the S could have heard some of them. Nevertheless, some of the information obtained only by I.S. strengthens the paranormality assessment. J.K. was told that the S had a birthmark on his neck apparently in agreement with the way the PP was stabbed. It is unlikely that this is a later invention without any factual basis. On the other hand, I.S. had inquired about birthmarks and had not received any information that the S had any birthmarks. It is possible that the S's birthmark had faded-assuming that the S indeed had a birthmark-and that this was conveyed to I.S. in such a way that I.S. assumed Paranormality in Reincarnation Cases 381 that the S never had one. Nevertheless, the later statement about the birthmark strengthens the paranormality assessment of J.K.'s report. However, based on the larger number of details recorded by I.S. that support the paranormality component, we rated the I.S. report as 8 and the J.K. report as 7.5. Discussion The results show no evidence that information obtained about these cases produce higher paranormality assessments over time. Some new details were added, but generally, more details were lost. Embellishments may be recognized if the paranormal components of a particular event are more strongly emphasized in later versions. Only sporadic additions or possible embellishments occurred, and most of these additions were probably of actual events that had not been mentioned during the earlier investigations or occurred afterward. The Anurak Sithipan case, with a slightly higher rating for the J.K. report, is probably such an example. Both parents of Anurak mentioned to J.K., but not to I.S., that as a young child, Anurak found a special spoon that the PP had placed on a high shelf and that Anurak apparently found without prior information or access to the place in which it was kept. It seems unlikely that this incident was invented. On the other hand, it apparently occurred before I.S. investigated the case. We must also acknowledge that the lower ratings for the J.K. reports may be partly due to J.K.'s more limited investigations. For instance, in one case, a birthmark, which had already faded when I.S. investigated this case, was not even mentioned to J.K. J.K.'s informants might still have remembered the details about this birthmark if J.K. had asked specific questions. Nevertheless, the results present a practical reassurance that if relevant informants can be found, events in connection with reincarnation-type cases are remembered with reasonable accuracy after many years and the paranormality assessments are only rarely inflated over time. We think this conclusion warranted for cases in Turkey. Our series contains too few cases from Thailand and Myanmar to justify any generalization about cases in those countries. If investigators of these cases in the future wish to adopt something like a "paranormality rating scale," we recommend that they aim at greater objectivity in making the ratings than we were able to achieve in the present comparison. We think it is possible to provide a scale with specific points awarded for some of the features we described in the Method section of this paper (Keil, 1991). Acknowledgments The research of the Division of Personality Studies is supported by The Bernstein Brothers Foundation, The Nagamasa Azuma Fund, The Japan-U.S. Fund for Health Sciences, The Lifebridge Foundation, and Richard Adams. We are grateful to J. Tucker and Patricia Estes for reading and offering com- 382 Ian Stevenson & Jiirgen Keil ments on earlier versions of this paper. Thanks are also due to Dawn Hunt for other assistance. The comments of an anonymous referee have helped us to improve the clarity of our presentation in several places. We also wish to extend our thanks to our informants. Many of them responded to two sets of inquiries about the same events; we are grateful for their patient cooperation. References Alvarado, C. S., & Zingrone, N. L. (1997198). Factors related to the depth of near-death experiences: Testing the "embellishment over time" hypothesis. Imagination, Cognition and Personality, 17, 339-344. Baddeley, A. D. (1998). Human memory: Theory andpractice. London: Allyn and Bacon. Bartlett, F. C. (1 932). Remembering: A study of experimental and socialpsychology. Cambridge, UK: Cambridge University Press. Greyson, B. (1983). The near-death experience scale: Construction, reliability, and validity. Journal ofNewous and Mental Disease, 171,369-375. Haraldsson, E. (1991). Children claiming past-life memories: Four cases in Sri Lanka. Journal of Scientzjlc Exploration, 5, 233-261. Keil, J. (1991). New cases in Burma, Thailand, and Turkey: A limited field study replication of some aspects of Ian Stevenson's research. Journal of Scientijc Exploration, 5,27-59. Mills, A. (1989). A replication study: Three cases of children in Northern India who are said to remember a previous life. Journal of Scientijc Exploration, 3, 133-1 84. Parsons, D. (1962). A nonexistent building located. Journal of the Society for Psychical Research, 41,292. Pasricha, S. (1990). Claims of reincarnation: An empirical study of cases in India. N e w Delhi: Harman Publishing House. Prince, W. F. (1918). Experiences of Augustine Jones. Journal of the American Society for Psychical Research, 12,718. Prince, W. F. (1919). Some coincidental dreams. Journal of the American Society for Psychical Research, 13, 6 1. Stevenson, I. (1974). Twenty cases suggestive of reincarnation (Rev. ed.). Charlottesville: University Press of Virginia. Stevenson, I. (1980). Cases of the reincarnation type. Vol. 3, twelve cases in Lebanon and Turkey. Charlottesville: University Press of Virginia. Stevenson, I. (1983). Cases of the reincarnation type. Vol. 4, Twelve cases in Thailand and Burma. Charlottesville: University Press of Virginia. Stevenson, I. (1997). Reincarnation and biology: A contribution to the etiology of birthmarks and birth defects (Vols. 1-2). Westport, CT: Praeger Publishers. Journal of Scientific Exploration, Vol. 14, No. 3, pp. 383-409,2000 0892-3310lOO O 2000 Society for Scientific Exploration ArtREG: A Random Event Experiment Utilizing Picture-Preference Feedback Princeton Engineering Anomalies Research Princeton University, Princeton, NJ 08544 Abstract- An experiment addressing anomalous humanlmachine interactions utilizing a feedback display of two competing pictures, the relative dominance of which is controlled by a microelectronic random generator, has yielded a number of equivocal results. On the one hand, an ingoing hypothesis that such a visually engaging mode of feedback might facilitate larger anomalous effects has not been supported by the composite results of 49 operators performing some 390,000 experimental trials. Likewise, a smaller a d hoc study of the relative efficacy of a subset of target pictures having religious or spiritual themes, although yielding effect sizes comparable with earlier random event generator (REG) data, has insufficient statistical power to resolve the question. Also, an attempt to assess the relative importance of the pictorial feedback, vis-a-vis the output of the REG, p e r se, in facilitating operator performance has not been definitive. Yet, certain secondary anomalies in these databases, such as gender disparities, individual operator performances, and serial position effects, show several characteristics akin to those previously found in other humanlmachine experiments in this laboratory. Whether these indicators can be used to develop more effective experiments of this class or to achieve a more fundamental understanding of the basic phenomena is the focus of ongoing research. Keywords: random event generator (REG) - random event experiments humanlmachine anomalies -visually engaging feedback Introduction Two decades of rigorous empirical study of the interaction of human operators with microelectronic random event generators (REGS) in this laboratory have consistently yielded small but statistically significant anomalous departures from chance behavior that correlate with the prestated intentions of the operators (Jahn et al., 1997). These effects, however, have proven remarkably insensitive to any of the attendant physical parameters so far explored, such as data set sizes and acquisition rates, nature of the random physical sources, and spatial and temporal separations of the operators from the functioning equipment (Nelson et al., 1999). What secondary correlations have been observed tend to be of a more subjective nature, relating to the individual operator genders, preferences, strategies, reactions to the feedback, and feelings of resonance with the devices. In an effort to exploit these intangible correlates to 384 R. G. Jahn et al. yield larger anomalous effect sizes, a group of related experiments have been designed and implemented wherein the feedback provided the operators, and in some cases the random physical processes themselves, have more aesthetically engaging visual or auditory properties than the standard REG experiments. Examples include a large crystal pendulum, a bubbling water fountain, a randomly impacted Native American drum, a mobile robot, and, in the study reported here, a display of two competing pictures superimposed on a CRT screen. It was hypothesized that experiments that offered such aesthetic appeal to the operators would enhance their sense of resonance with them, and thereby enable larger deviations from chance behaviors. Experimental Design, Protocol, and Analysis In this article, we shall review the results of several years of experimentation with a target device termed "ArtREG," wherein two attractive pictures compete for dominance on a CRT screen. To achieve this presentation, the screen is illuminated by a field of 640 x 480 = 307,200 pixels, each of which may correspond to either of the competing illustrations, which are drawn from a library of 24 arbitrarily selected paintings, photographs, and designs previously scanned and digitized (Appendix A). The relative fraction of pixels corresponding to each of the two pictures is controlled by a small REG unit based on microelectronic Johnson noise (Nelson, Bradish, and Dobyns, 1992), in accordance with a software recipe that uses the cumulative deviation of the REG digital output from the chance mean. More specifically, the REG output is collected in trials of 200 sample bits, and the compounding difference of the trialmean values from 100 is used to determine the fraction of pixels illuminated by each picture, as described more fully below. The formal experimental protocol, which evolved over a period of informal preliminary experimentation (Appendix B), calls for the operator to examine the library of available pictures and select two for the competition. At the start of each experimental run, these appear superimposed on the screen in equal prominence, after which the balance evolves in accordance with the progress of the REG output. The goal of the operator is to bring one of the pictures into full or at least partial dominance, in accordance with a prerecorded intention. (In one popular variant, the operator may choose only one picture, to compete with a multicolored random pixel illumination, so that the chosen image appears to be sharpening from, or diffusing into, a random noise background.) When saturation by either picture occurs, or if 250 trials have been compounded without saturation, the run is terminated, the results are recorded, and a subsequent run is initiated using the same or the opposite picture preference. Runs are generated until 2,000 trials (400,000 bits) have been accumulated, comprising one experimental series, or session. Operators are constrained to use the same pair of pictures throughout a given session, but the picture preference is optional for each run. In an effort to focus the operators' attention solely on the visual display, ArtREG: A Random Event Experiment 385 rather than on the REG output per se, a pseudorandom algorithm determines whether the emergence of the preferred picture on any given run will be associated with higher or lower deviations of the REG output from its chance expectancy of 100. Because this assignment is not revealed to the operators, they remain blind to the directional criterion that determines the success or failure of a desired run outcome. In this sense, ArtREG protocol differs substantially from that of standard REG experiments, where the feedback presented to the operator directly reflects success or failure in achieving the desired intention to produce higher or lower electronic counts. By keeping the assignment blind in the ArtREG experiment, it was hoped to determine whether the visual feedback or a more direct coupling of the operator's intention to the REG process itself was the more important factor in enabling operator performance. Data thus are processed under two complementary criteria: success on chosen pictures and success on high or low deviations of the REG trial means. Because the algorithm determining high or low assignment ensures that each series contains equal numbers of trials (1,000) associated with each direction, the highllow analysis in ArtREG is directly comparable to that employed in the benchmark REG experiments, even though in this case the data are produced under "double-blind" conditions. Several other possible correlations also are explored in the data assessments, e.g., the effectiveness of particular pictures or picture combinations, run-level versus series-level success, and operator gender disparities. Results The formal ArtREG database consists of 195 series, or sessions, comprising 390,000 trials, generated by 49 volunteer operators: 21 males (98 series) and 28 females (97 series). Their results are summarized in Tables 1 through 4, which show the number of completed series (Ns); the number of "successful" series (Ss) where the average results were consistent with intention; the total the number of runs (NR); number of successful runs (SR);and the overall mean shift (p),standard error (o), and Z score for each operator. As displayed in the first column of Table 1, of the total of 3,597 runs, 1,798 were partially or completely successful and 1,799 were not. Using a theoretical standard deviation for the binary expectation of success, o = (.25lV)ll2= 29.987, yields a totally insignificant bottom-line Z score of -0.0167. The next four columns display the correlations of successes with the high and low REG drivers of the preferred pictures and with the gender of the operators. The last four columns further subdivide the results into high-driven and low-driven runs by males and females, respectively. Clearly, none of these sets or subsets displays any anomalous statistical character at this level, with the possible exception of the marginal disparity between male and female results on the low-driven pictures (Zdiff= 1.687), which probably can be discounted on multiple-test grounds. Operator-specific representations of the results are presented in Tables 2 and 3 in the form of individual male and female achievements on complete ex- 386 R. G. Jahn et al. TABLE 1 Overall Run Score Summaries Subset All Male Female High Low HighM LowM HighF LowF NR SR NR- SR Z -score* MRL+ (Suc.lTot.) NR= number of runs SR= number of successful runs NR- SR= number of unsuccessful runs *Z = (2SR- N ~ ) / [ ( N ~ ) " ~ ] +MRL= mean run lengths (0-250 trials). perimental series, again segregated in terms of high-driven versus low-driven trials. At the bottom of each table are listed the sums of values over the operators and the corresponding probabilities that the distributions are chance. Also included are the aggregate totals of series performance by the operators. Table 4 compounds the x2 and series performance results over all operators. Again, there appears to be little of statistical interest in any of these composite indicators, but despite the overall low yield of the total database, more detailed examination reveals a few potentially instructive features. For example, 11% of the database consists of single series generated by 45% of the operators. Seventeen of these single-series databases, or 74%, yield positive results in the combined high-low data (2= 2.558). As an alternative representation of the individual operator data, Figure 1 displays the fraction of successful runs achieved by each as a function of database size, in units of ( N ~ ) "on, which are superimposed the loci of one stan~ dard error confidence limits. The use of different symbols for female and male operators helps to illustrate the gender differences appearing in these data. Figure 2 is a similar representation in terms of individual operator effect sizes, in units of Z score per series. Correlation of operator success with particular pictures is somewhat complicated by their use in pairs; that is, while one might endeavor to search for a relative picture-effectiveness distribution that compounds uniformly over all competing pictures, because these are selected at will by the operators, the full competition matrix is far from uniformly populated. Nonetheless, some indications can be extracted. Table 5 lists the overall success ratios for each of the 24 illustrations (displayed in Appendix A) when used as the preferred target, ArtREG: A Random Event Experiment 387 regardless of its competitors. Listed are the ratios of successful runs (SIN) achieved using the picture as a preferred target, along with the corresponding Z scores, for all operators, men alone, and women alone. Also listed in the composite portion of the table are the number of successful runs that saturated compared to the total number that saturated (Sat.), and the average number of trials per run for the successful runs, compared to the average number for all runs (tsltt). Once again, the bottom-line results are indistinguishable from chance, as are the X 2 goodness-of-fit tests. Figure 3 illustrates the relative successes of the pictures in graphical form. Table 6 and Figure 4 display lists of the operator achievements under the modified protocol that uses the random noise pattern, rather than a second illustration, as the competitor to the chosen picture. Clearly, there is no indication that this modality is any more effective than the two-picture version in enabling the intended anomalous achievements. Ad Hoc Experiments In an admittedly a posteriori effort to explore possible causes of the apparent failure of this experimental concept, it was noted that a particular subset of the target pictures seemed to facilitate disproportionately good performances. These images could crudely be subsumed within a category of religious, mystical, or symbolic patterns, labeled as follows: (1) Anubis; (2) Apache; (3) Wave; (7) Mask; (8) Bear; (9) India; and (1 1) Egypt. Specifically, this subset yielded 521 successful runs out of 961 (54.2%), with a corresponding Z score of 2.6 13. A similar correlation was found within the single picture versus random background efforts within this group: 134 successes out of 244 (54.9%), Z = 1.536. A subsequent ad hoc experiment was undertaken using only these seven pictures plus the random background option, within a short-form protocol that permitted more rapid accumulation of data. Specifically, all runs were forced to have equal numbers of trials (loo), with only four runs comprising a session, and full saturation was precluded by a progressive difficulty algorithm in the software (i.e., by using a difficulty criterion, d, up to M pixel saturation, a difficulty criterion of 2d over the saturation interval M to %, 4d over the saturation interval % to 76 etc.). Thus, highly successful runs could be sustained near saturation without premature termination of the run. To conserve laboratory and operator time, only 100 series (400 runs) were planned, the minimum deemed necessary to distinguish this subset from the main database. Despite all these changes in experimental design, the composite results of this phase, as summarized in Tables 7-9 and Figure 5, again were statistically insignificant. Specifically, only 200 runs out of 404 were successful (Z = -0.1990); the all-operator performance compounded only to ZA= 0.5305; four of the eight selected pictures now scored below the chance mean; and the X 2 on both the operator and picture distributions were within chance. On the other hand, it is worth noting that the ZAseems to correspond to an absolute ef- R. G. Jahn et al. TABLE 2 Male Operator Results 14 3 High Low A High Low A High L OW A High Low A High LOW A High Low A High Low A High L OW A High Low A High Low A High Low A High Low A High LOW A High LOW A High Low A 1 2 2 1 3 1 1 2 1 2 2 2 1 2 3 1 0 1 0 1 1 9 4 8 7 9 9 0 1 1 3 1 3 1 1 1 1 0 1 1 0 0 2 2 3 27 22 49 36 39 75 37 38 75 25 26 51 31 45 76 47 48 95 10 7 17 143 135 278 122 140 262 9 7 16 75 80 155 9 6 15 11 10 21 9 7 16 47 47 94 13 12 25 -. 1000 .0483 21 4 14 20 34 16 20 36 14 14 28 18 21 39 21 19 40 5 4 9 75 59 134 59 74 133 4 5 9 36 29 65 8 4 12 8 5 13 5 2 7 24 24 48 -. 1530 .0120 41 4 -.1065 -.0063 .0833 .06 17 .0423 -.0400 -.0962 .I118 .I118 .I291 .I291 .I118 .I118 .lo00 .lo00 .2236 .2236 .0577 .0577 .0598 .0598 .2236 .2236 .0791 .0791 .2236 .2236 .2236 .2236 .2236 .2236 .lo00 .lo00 -0.9526 -0.0559 -0.7131 0.6455 0.4777 0.7942 0.3779 -0.3578 0.0142 -0.9620 -1.8920(*) -2.0181(*) -0.1789 0.1834 0.0032 0.7829 -1.0196 -0.1 674 -0.2295 0.6598 0.3043 -0.6887 1.4356 0.528 1 -0.3573 -2.3685(*) -1.9275(*) 2.0214* 0.6977 1.9227* 1.2075 -0.1431 0.7526 0.3801 -0.8944 -0.3637 0.0880 0.1340 0.1570 56 3 ~ 65 4 84 5 -.1892 -.0400 .0410 .0452 -.0589 -.0137 .0394 184 1 187 15 307 14 317 1 -. 1540 .3210 -.0283 320 8 -. 1872 .4520 .I560 ,2700 -.0320 .0850 -.2000 .0088 .0134 321 1 402 1 405 1 409 5 1 ArtREG: A Random Event Experiment TABLE 2 Continued 389 ' ~ 1 I 422 3 High L OW A High L OW A High Low A High LOW A High L OW A High L OW A High L OW A 2 1 1 3 3 4 4 6 7 2 2 2 1 1 1 25 33 58 62 63 125 89 87 176 28 27 55 33 34 67 10 6 16 885 907 1792 15 15 30 34 30 64 42 44 86 19 12 31 12 16 28 6 2 8 448 43 1 879 .0403 -.06 17 .0466 - 199 .O -.0667 .0146 .2430 -.0603 -.0990 -.0155 .3380 .I291 .I291 .0845 .0845 .0707 .0707 .I291 .I291 .I118 .I118 .2236 .2236 .0226 .0226 0.3 124 -0.4777 -0.1168 0.55 10 -0.2350 0.2235 -0.9433 0.2065 -0.5210 1.8823 * -0.4673 1.0005 -0.8855 -0.1386 -0.7242 1.5116 -0.5501 0.6799 -0.0145 -1.2527 -0.8960 501 7 507 10 509 3 515 4 599 1 1 0 1 44 43 53 -. 1230 -.0003 -.0283 All 98 I Chi-squared tests: High Low A 19.764 15.461 16.945 0.537 0.799 0.714 Individual operator overall success patterns: Nop. 21 H/L High Low A Z > 0.00 ( P I < .5) 11 9 11 Z > 1.645 ( P I < .05) 2 0 1 Z < 1.645 ( P I> .95) Note: Op. = operator; Ns= number of series; H/L = high or low random-event-generator driver; Ss = number of successful series; NR = number of runs; SR= number of successful runs; p = trial mean shift; o = standard error; Z= Z score = p/o; N op. = number of operators; PE = chance expectation. Significant at P < .05. (*I Significant at P > .95. R. G. Jahn et al. TABLE 3 Female Operator Results Op. Ns 3 1 H/L High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A Ss NR SR P CF Z 10 17 16 8 17 10 53 1 159 1 173 3 327 1 345 1 350 1 363 1 401 1 403 1 406 2 500 1 ArtREG: A Random Event Experiment TABLE 3 Continued 502 6 High Low A 3 5 4 2 4 4 0 1 1 4 2 3 56 63 119 40 39 79 8 10 18 88 95 183 89 99 188 10 10 20 19 21 40 9 8 17 9 9 18 36 36 72 8 10 18 10 11 21 36 43 79 888 917 1805 503 4 High Low A 504 1 High Low A 506 10 High Low A 5 10 11 High Low A High Low A High Low A High Low A High Low A High Low A High Low A 7 5 7 0 1 1 0 1 0 0 1 1 1 1 1 1 3 2 1 1 1 0 0 0 3 2 3 49 53 56 5 14 1 543 2 609 1 623 1 707 4 709 1 8 13 1 High Low A 830 4 High Low A High Low A All 97 Chi-squared tests: High Low A x2 (28 df 26.258 30.706 29.390 px 0.559 0.330 0.393 R. G. Jahn et al. TABLE 3 Continued Op. Ns H/L Ss NR SR P 0 Z Individual operator overall success patterns: Nop. 28 H/L High Low A Z > 0.00 ( P I< .5) Z > 1.645 ( P I < .05) Z < 1.645 ( P I > .95) 15 17 18 Note: Op. = operator; Ns= number of series; H/L = high or low random-event-generator driver; Ss = number of successful series; NR = number of runs; SR= number of successful runs; p = trial mean shift; o = standard error; Z = Z score = p/o; N op. = number of operators; $E = chance expectation. Significant at P < .05. (*I Significant at P > .95. database (Jahn et al., 1997), as well as to the exploratory data described in Appendix B, even though the low statistical power of this small data set imposes large error bars on this value. These comparisons are illustrated in Figure 6. It also is interesting that significant ZAvalues were attained by 4 of the 2 1 operators, compared to roughly one expected by chance, and that on a two-tailed TABLE 4 All Operator Success Summary 195 High L OW A 93 96 109 1,773 1,824 3,597 895 903 1798 -.0027 -.0009 .0160 .0160 -0.1691 -0.0583 -0.1608 Chi-squared tests: x2 (49 df High Low A 46.022 46.168 46.333 px 0.595 0.589 0.5823 Individual operator overall success patterns: Nop. 49 H/L High Low A Z > 0.00 ( P I < .5) 26 26 29 Z > 1.645 ( P I < .05) Z < 1.645 ( P I > .95) Note: Ns= number of series; H/L = high or low random-event-generator driver; Ss = number of successful series; NR = number of runs; SR = number of successful runs; p = trial mean shift; o = standard error; Z = Z score = y/o; N op. = number of operators; CE = chance expectation. ArtREG: A Random Event Experiment 1.o- 393 Female v Male o 0.8- 0.6v 52 rm v v ov 0 0 00 6'0 0 0 V 0 0 v 0.4 0 0.2- 0.0 0 I I I 5 10 Sqrt(Number of Runs) 15 20 Fig. 1. Runwise success rates by operators. basis over the separate high and low runs, 9 of the operators scored in the 11> 2 1.645 tails. In other words, despite the apparently inconclusive composite results, the individual operator performances display potentially instructive idiosyncratic effects. Interpretations and Speculations Given the equivocal character of the data derived from this sequence of ArtREG experiments, any conclusions and interpretations thereof must be re- 394 R. G. Jahn et al. 0 1 2 Sqrt(Number of Series) Fig. 2. Effect sizes by operators. 3 4 garded as tentative, speculative, and possibly even intuitive. Clearly, the composite results do not support the hypothesis that attractive and engaging feedback displays enable substantially larger anomalous effects, a conclusion that appears to be consistent with the early results of a number of our other visually appealing experiments. The attempt to assess the relative importance of the visual feedback on operator performance compared to that of the REG digital output stream, per se, also has been thwarted by the small effect sizes. Indeed, we now might speculate whether the blinding of the operator to the high ver- TABLE 5 Success Ratios by Picture (Rank Ordered) All S/N Anubis Apache Wave Petals Random Abduction Mask Bear India World 2 Egypt Toledo Arch Rug Shield Horserug World Acacia Hand Surf Calder Japan Leopard Park 33.25 (24 df) (Px = 0.099) &=33.61 (24df) (P, = 0.092) 2s Male Sat. tsltt SIN Zs Female Rank S/N Zs Rank xg = 21.70 (22 df) (P,= 0.478) Note: SIN = successful runsltotal number of runs; Zs = Z score of S/N; Sat. = successful runs saturatedltotal number of saturated runs; t,lt, = average number of trials for successful runslaverage number of trials for all runs. 396 R. G. Jahn et al. Sqrt(Number of Runs) Fig. 3. Runwise success rates by target image. sus low outputs of REG might be implicated in the reduction of effect size in the main group of experiments. Yet, as in many earlier studies, both "successful" and "unsuccessful," various details appear in the ArtREG data that if better comprehended could possibly point to more incisive experimental and theoretical strategies and to better understanding of the underlying phenomena. The somewhat better yield from the mystical or symbolic image subset also prompts a few intuitive speculations. First, it may be that the explicit artistic feedback characteristic of most of the pictorial targets, rather than enhancing ArtREG: A Random Event Experiment TABLE 6 Success Ratios by Picture Versus Random Background (Rank Ordered) Wave Anubis Mask Bear Apache Leopard India World Toledo Hand World 2 Calder Park Acacia Egypt All X 2 = 10.09 (1 5 df) (P,=O.81) Note: S/F = successful runslfailed runs; ZSIF Z score of S/F. = the resonance of the operator with the experiment, actually may inhibit such because of its specificity. That is, just as a fully random physical source appears to be requisite raw material for production of a more ordered digital stream in our standard REG experiments (Jahn et al., 1997), so the operator's consciousness may prefer less associative constraint on the imagery it employs to achieve its resonance with the experimental task than the fully articulated pictures allow. Only in the more vague and symbolic illustrations, such as those used in the ad hoc experiment, may some relief from this encumbrance be provided. Alternatively, it may be the symbolic, personalized meaning of the feedback, i.e., its particular relevance to the operator, that is the crucial ingredient in establishing a productive humanlmachine bond, and that the mystical subset carries more such individualized meanings to the operator. Or finally, all of these weak results may just be another indication that feedback, in any form, is not a major requisite in producing such anomalous effects. This interpretation would be consistent with the positive results of our Remote REG (Dunne and Jahn, 1992), Remote Perception (Nelson et al., 1996), and FieldREG experiments (Nelson et al., 1998), in which success is achieved even though no form of immediate feedback is available. Indeed, it is even possible that the anomalous effect sizes are fundamentally unamplifiable by any experimental strategy, i.e., that their scale is intrinsically constrained to at best a few parts per thousand, so that major statistical effects can be found only in very large individual or collective databases. In any case, these results have taught us that we must broaden our range of potentially important variables beyond those so far explored to include subtler R. G. Jahn et al. Sqrt(Number of Runs) Fig. 4. Runwise success rates for targets versus random. TABLE 7 Ad Hoc Experiment: Overall Run Score Summaries All No. of runs No. of successes No. of failures Z score (successes/failures) 404 200 204 -0.1990 High 202 96 106 -0.7036 Low 202 104 98 0.4222 ArtREG: A Random Event Experiment TABLE 8 Ad Hoc Experiment Operator Result 399 10 8 (F) 14 7 High Low A 6 5 6 (MI 17 (F) 21 10 High Low A 4 2 3 4 4 4 2 3 2 2 High Low A 5 (MI 41 10 High Low A (MI 196 1 High Low A 5 4 0 1 1 1 1 1 0 1 1 0 1 0 1 1 0 4 4 4 0 0 0 2 4 (M/F) 213 1 High Low A (M/M) 240 2 High Low A (M/F) 263 1 High Low A (M/F) 282 2 High Low A (M/F) 307 (M) 318 10 High Low A High Low A 2 (F) 373 5 High Low A (F) 412 1 High Low A High Low A 5 1 0 1 6 4 7 (MI 462 8 (M) High Low A R. G. Jahn et al. TABLE 8 Continued 500 (F) 516 (F) 549 (MI 551 (F) 552 (MI 555 (M) 6 High Low A High Low A High Low A High Low A High Low A High Low A High Low A 3 10 3 3 3 All 101 Chi-squared tests: x2 (2 1 df High Low A 31.762 18.204 27.089 px 0.062 0.636 0.168 Individual operator overall success patterns: Nop. 21 H/L High Low A Z > 0.00 (PI < .5) 8 11 10 Z > 1.645 (Pi < .05) Z<-1.645 (Pi > .95) Note: Op. = operator; Ns= number of series; H/L = high or low random-event-generator driver; Ss = number of successful series; NR= number of runs; SR= number of successful runs; p = trial mean shift; (3 = standard error; Z = Z score = p/o; Nop. = number of operators; P E = chance expectation. Significant at P < .05. (*) Significant at P > .95. ArtREG: A Random Event Experiment TABLE 9 Ad Hoe Experiment: Success Ratios by Picture (Rank Ordered) 40 1 Rank Apache India Egypt Random Anubis Mask Bear Wave All Prior rank SIN 2s X 2 = 10.486 (8 df) (P,= 0.23) Note: SIN = successful runs/total number of runs; Zs = Z score of SIN. personal factors, such as environmental influences, operator and experimenter attitudes, and the role of subjective meaning, in our future experimental designs if we are to acquire deeper understanding of these consciousness-related physical phenomena. Appendix A: ArtREG Illustrations All of the pictures utilized in the main body of ArtREG experiments and in the ad hoc subset are reproduced here in black and white to reduce printing costs. A few sets of full color reproductions are retained in our laboratory. Appendix B: Prior Explorations The formal ArtREG experiments reported in the body of this paper devolved from an earlier informal set of exploratory studies performed as the equipment was being brought on line and the protocols were being refined. These initial experiments closely followed our standard REG protocols, using series of 1,000 trials taken under the three directional intentions: high, low, and baseline. These were grouped in four sets of 250-trial runs per direction, with the operators blind to the randomly assigned directions. All runs required completion of 250 trials, even though picture saturation may have been achieved during the run. All told, 13 operators completed a total of 37 series, with the results displayed in Table 10. Although none of the operators, nor the group as a whole, achieved statistical significance in the high-low separation, the overall effect was in the intended direction, with an absolute size comparable to that characteristic of our much larger benchmark REG studies (Jahn et al., 1997), and hence regarded as propitious for more formal and extensive experiments. 402 R. G. Jahn et al. Based on the reactions of several of these operators, a number of changes in processing and protocol were implemented before the formal experiments were begun, namely: 1. The baseline direction was eliminated. 2. Runs were allowed to terminate on saturation in either direction. 3. The total number of trials per series remained at 1,000 for the high and ArtREG: A Random Event Experiment 403 Benchmark REG ArtREG Ad-Hoc Fig. 6. Effect-size comparisons. the low directions, but these included some shortened (saturated) runs, as well as full 250-trial runs. 4. Additional pictures were added to the library, and a few were removed. Acknowledgments The authors would like to acknowledge their indebtedness to Dr. Andrei Berezin of Moscow, Russia, who, in the course of an extended visiting appointment in our laboratory, first proposed an REG experiment with an artistic visual feedback display. Although the details of the ArtREG implementation have evolved considerably since his original conception of the idea, it was nonetheless seminal to the studies reported here. We also would like to thank the many operators who, without compensation or identification, have generously provided the databases of this project, Several other members of the Princeton Engineering Anomalies Research (PEAR) staff, other than the authors, have also played important roles in the conduct and interpretation of the experiments, as well as in the preparation of this report. In particular, Lisa Langelier-Marks has spent a considerable amount of time and effort implementing an effective and attractive technical text. The PEAR program gratefully acknowledges the enduring financial support of the Institut fiir Grenzgebiete der Psychologie und Psychohygiene; The Lifebridge Foundation, Richard Adams; George Ohrstrom; Laurance Rockefeller; Donald Webster; and numerous other private contributors. 404 R. C;. Jahn et al. 3. Wave 4. Petals ArtREG: A Random Event Experi~nent 1 1 . Egypt 12. Tolcdo 406 R. G. Jahn et al. 13. Arch 14. Rug 17. World 18, Acacia ArtREG: A Random Event Experiment 19. Hand 20. Surf I 2 1. Calder 22. Japan I 23. Leopard 24. Park R. G. Jahn et al. TABLE 10 ArtREG, Exploratory 7 1 High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A High Low A 10 16 21 1 41 1 78 3 84 1 161 6 171 1 173 1 174 3 182 1 406 1 813 1 All 37 ArtREG: A Random Event Experiment TABLE 10 Continued Chi-squared tests: 409 High Low A 10.450 6.439 8.736 0.657 0.929 0.793 Note: Op. = operator; Ns= number of series; HIL = high or low random-event-generator driver; Ss = number of successful series; NR = number of runs; SR = number of successful runs; p = trial mean shift; o = standard error; Z = Z score = plo. References Dunne, B. J., & Jahn, R. G. (1992). Experiments in remote humanlmachine interaction. Journal of Scientific Exploration, 6(4), 311-332. Jahn, R. G., Dunne, B. J., Nelson, R. D., Dobyns, Y. H., & Bradish, G. J. (1997). Correlations of random binary sequences with pre-stated operator intention: A review of a 12-year program. Journal of Scientific Exploration, 11(3), 345-367. Nelson, R. D., Bradish, G. J., & Dobyns, Y. H. (1992). The portable PEAR REG: Hardware and software documentation. Princeton Engineering Anomalies Research, Princeton, NJ-Internal Document, 92(1), 37. Nelson, R. D., Dobyns, Y. H., Jahn, R. G., & Dunne, B. J. (1999). Contributions to variance in REG experiments: ANOVA models and specialized subsidiary analyses. Princeton Engineering Anomalies Research, Princeton, NJ-Internal Document, 99(2), 29. Nelson, R. D., Dunne, B. J., Dobyns, Y. H., & Jahn, R. G. (1996). Precognitive remote perception: Replication of remote viewing. Journal of Scientific Exploration, 10(1), 109-110. Nelson, R. D., Jahn, R. G., Dunne, B. J., Dobyns, Y. H., & Bradish, G. J. (1998). FieldREG 11: Consciousness field effects: Replications and explorations. Journal of Scientific Exploration, 12(3), 425-454. Journal of Scientific Exploration. Vol. 14, N o . 3, pp. 411-420,2000 0892-33 10/00 O 2000 Society for Scientific Exploration Can Population Growth Rule Out Reincarnation? A Model of Circular Migration Johns Hopkins School o f Public Health Department of Population and Family Health Sciences 615 N. Wove St., Baltimore, MD 21205 Phone (410 ) 955-7807, Fax (410) 955-2303 e-mail: [email protected] Abstract- Many have tried to draw scientific inferences about the existence or nonexistence of an afterlife from observations undertaken in the natural world. A leading example of these attempts is the common refutation of the hypothesis of reincarnation as being inconsistent with human population growth. This paper demonstrates that the demographic refutation of reincarnation only holds if supplemented by substantial metaphysical assumptions. A simple Markov model of circular migration can account for population growth and remain consistent with a reincarnation hypothesis. Though the chief attraction of the circular migration model is its simplicity, the model also has implications about how long people would have to remain discarnate. Because multiple conflicting theories about the existence and nonexistence of an afterlife are consistent with the currently available data, there can be no conclusion regarding which is correct. An incorrect assessment of the limits to what can verifiably be known about an afterlife has plagued our predecessors with schisms, wars, and intolerance. It need not plague our successors. Keywords: reincarnation-population Introduction Edwards (1996) credits the early church father, Tertullian, for originating the notion that human population growth automatically invalidates the hypothesis of reincarnation. It is argued that a growing population of human bodies would somehow "run out" of souls with which to be incarnated. To accept such a claim would be to accept that one can observe the world and its population of living creatures and thereby make testable claims about the existence or nonexistence of an afterlife. The general answer to this question is of some interest, because the media has begun to feature several clips of professional scientists expressing "scientific" opinions about supernatural phenomenon. Can science really shed light on realms that are void of data? Despite the interest of these more general questions, this paper will be confined to examining the claim that the observed growth of the human population is inconsistent with the reincarnation hypothesis. The paper begins with a discussion of the facts of human population growth followed by two simple theoretical models to ac- 412 D. Bishai count for these facts. The two theories are compared. Finally, the reader will be left to judge whether the demographic refutation of reincarnation is as simple as has been commonly thought. Facts The facts of life are clear. Each human lifespan has a beginning and an end. Each lifespan can be measured as the interval between birth and death. Another unassailable fact is that the planet's human population is completing a phase of more than 500 years of unprecedented population growth. Our population's history can be divided into three phases. The modern phase dates to the middle of the second millennium and is characterized both by the highest growth rates and the best quality demographic data. The world's current population is approximately 6 billion and is sustained by an annual birth cohort of 135 m i l l i o n 8 5 million more than the annual death toll of 50 million. A sustained global decline in the fertility rate began over 20 years ago and shows no sign of reversing. Barring significant unforeseen economic or epidemiological setbacks, world population should peak at roughly 10 billion about the year 2050 (United Nations, 1998). Data on the historical size of the world's population is more sparse. The Roman empire, from Spain to the Near East, was estimated to be between 45 and 90 million persons during the reign of Augustus in 14 A.D. A very rough estimate of the world's population around 1 A.D. is about 300 million people (Haub, 1995). Historians of the Eastern hemisphere record at least two large fluctuations in population growth in this era. One fluctuation was caused by the Black Plague in which about one third of the population of Europe is thought to have perished (McNeill, 1976; Ziegler, 1969). The other population shock during this period can be attributed to the Asian Expansion led by Temujin of Mongolia. Without these two fluctuations, the historical era may have seen much larger population growth rates. As it was, overall population growth between 1 A.D. and 1650 A.D. was lower than for the period between 8000 B.C. and 1 A.D. There is even less data on the size of human populations prior to the development of writing. The current consensus is that the size of the world's human population in 8000 B.C. was about 5 million (U.S. Bureau of the Census, 1997). This implies a sustained growth rate of .05% per year between 8000 B.C and 1 A.D. Barely sustained growth would be a more apt description- -given child mortality rates of roughly 30%, every woman would have been required to have at least six children for the population to fend off extinction. A common question to be asked in the context of observations on the history of human population growth is, "How many people have ever lived on earth?" Any answer must be as arbitrary as selecting the starting point for our species based on the fossil record. At what historical point does one wish to accord the title "person" to the beings who animated the primate fossils that have been unearthed in the last 150 years. Although there is unassailable consensus that Population Growth and Reincarnation 413 the Cro-Magnon who painted the walls of caves in 30,000 B.C. were "people", it becomes murkier to assign this status to earlier primates whose only remaining artifacts are stone tools. It is unlikely that there will ever be an unassailable answer. Furthermore, as new fossils are unearthed, we will certainly learn more about which ancestor did what and when. For now, I suggest that the evidence of ritualized burials that show the intentional combination of artifacts with human remains offers as good a starting point as any for our species. The capacity to grieve, as evidenced by a ritualized burial, almost certainly requires the cognitive capacity to retain a mental representation of the counterfactual-an imagined world still animated by the dead loved one. The site bearing the earliest fossils with modem Homo sapiens morphology in a ritualized burial are from Qafzeh in modern Israel. It dates to the Mousterian era of the Middle Paleolithic, which is roughly 50,000 B.C. (Mellars, 1989). The skeleton was found in the fetal position, embraced by a large deer antler. With this arbitrary starting point, one may integrate the area under a series of exponential population growth curves to determine the number of human beings ever born. Using a starting date of 50,000 B.C., Haub (1995) offers an estimate of 105 billion human beings ever born. If these historical human lifespans conformed to the United Nations' high mortality pattern (United Nations, 1982), roughly 17 billion human lives would have ended before age 1, and 30 billion lives would have terminated before age 15. Those who see some religious significance to a cycle of human rebirth seldom grapple with the fact that nearly one in six human rebirths would have been followed by a "redeath" in less than a year. Theories As evidenced by the burial sites of our ancestdrs, theories to explain the facts of life are as old as man. A theory that appears to have been espoused by the vast majority of the 70 billion human beings who ever survived to adulthood is that there is a human soul that has some existence after the body has died. Accounts of the state of the disembodied human soul are many and varied. East and West differ over whether the soul returns to inhabit another living body or whether the soul goes on to some ethereal state of reward or punishment. Another theory is that there is no soul, only bodies made of atoms and molecules. Although this theory is sometimes called the scientific view, reflection shows that this theory is neither more nor less amenable to the publicly verifiable empirical tests required of scientific theories than are the more ancient religious teachings on the afterlife. To support their opinion, people that do not believe there are souls point to a lack of verifiable cases of reincarnation. 414 D. Bishai Fig. 1. Linear migration model. Materials and Methods Computer Simulation A computer simulation of the last 52,000 years of human population growth was programmed using Excel software under two different models of the nature of the afterlife. The models and the equations that represent them are discussed below. Models of the Human Population This paper will sketch out two competing accounts of the metaphysical implications of human population growth using Markov models. Markov models are a common tool in demography, epidemiology, and engineering and are used to model processes where there are transitions between states. In this application, the state of interest is the state of membership in the population of living human beings. Let us call this state A. Model 1. The linear migration model. The linear migration model (LMM) depicts the Markov process as shown in Figure 1. Individuals enter state A from an undefined source through a process called birth at a rate RBA.They exit state A to an undefined sink through a process called death at a rate RAB. The model predicts that human population growth will be accompanied by birth rates in excess of death rates. Mathematically, the prediction is dPA/dt> 0 H RBA > RAB. The assumptions required by the LMM are that human souls are continuously created from an undefined source and continuously destroyed in an undefined sink. Model 2. The circular migration model. An alternative model of the human population will be called the circular migration model (CMM). In the CMM, there are no ill-defined sources or sinks. Souls enter state A from state B through a process called incarnation at rate RBA.' Souls depart state A and re- The notational use of subscripts A and B may be made more useful by recalling that the Sanskrit "Atman" refers to the subjectively observable earthly soul and that the word "Bardo" is Tibetan for the afterworld. ' Population Growth and Reincarnation 415 Fig. 2. Circular migration model. turn to state B through a process called excarnation at rate RAB.Both incarnation and excarnation are observable processes documented publicly through birth and death certificates and recorded in vital registration databases. The model predicts that dPA/dt > 0 H RBA> RAB.In its simplest form, it assumes that human souls are neither created nor destroyed. More complex versions of the CMM not discussed here could accommodate the relaxation of this assumption. For simplicity, this paper invokes the assumption that the sum of incarnated and unincarnated human souls has been constant throughout history (PA+ PB= K, where K is constant). This is not a necessary feature of the reincarnation hypothesis, nor is it a necessary feature of the CMM. Many religious traditions hold that there are multiple states of heavens, hells, and purgatories in their accounts of an afterlife. The basic CMM, however, would gain little by incorporating such states. Indeed, Edwards (1996) singles out for scorn the multiple variants of the CMM with what he terms "noxious ad hoc assumptions" governing transitions of souls to astral planes, other galaxies, and animals. If there ever were census data or vital registration data available for these realms, the CMM could easily be enlarged to encompass them. Another feature that is not in the basic CMM is the creation or destruction of souls as depicted by the source and sink in LMM. The CMM could be enlarged to accommodate such features by grafting sources and sinks linked to state B. However, despite multiple conjectures by both spiritual and scientific scholars, there are no preliminary data on which to base such a hypothesis. We will use the available demographic data to assess both the LMM and the CMM. According to both models, all of the observers and all of the observations occur in state A. Indeed, it is with respect to data on the living human population that the models will be assessed. Table 1 summarizes the assumptions and predictions of the two models. Both models require strong assumptions about the eternity of the human soul. TABLE 1 Summary of the Models of Human Population Model Linear migration model Circular migration model (simple version) Predictions dPA/dt > 0 H RBA> RAB Assumptions Human souls continuously created and destroyed Human souls neither created nor destroyed Implication Undefined sources and sinks for human population Simple version requires that PA+ PB= Constant dPA/dt> 0 e RBA> RAB The LMM requires one to assume that the soul is ephemeral and finite. The basic CMM requires one to assume that each human soul is eternal. Both models predict the equivalence of population growth with disequilibrium in entry and exit to state A. The CMM is thought to be vulnerable in its simplifying assumption that PA+ PB = K where K is constant (Edwards, 1996), but relaxing that assumption does not lead to a fundamental change in the dynamics of the model. Results What do the Models Imply About the Dwell Time in Each State? In any first-order process, the reciprocal of the rate of outflow per inhabitant per unit time is a good approximation to the average dwell time of each inhabitant in a state. In both models, the rate of outflow from state A is characterized by TAB,which has the demographic interpretation of the crude death rate (deaths per 1,000 living persons per year). As an illustration, suppose one takes ~ A = 0.014, a number close to the crude death rate of the United States. B The average duration of life then is 110.014, or roughly 70 years. Table 2 presents the average human lifespans (or dwell times in state A), implied by applying both of these models to modern and historical crude death rates for the human population. Because observations are taken from state A, demographers have found it most natural to calculate crude birth rates, TBA, (births per 1,000 living persons per year) relative to state A, such that RBA= ~ B A PA= number of births into X A per year. In order to apply the time constant approximation to the CMM, it is necessary to transform r g such that it mediates an outflow and is expressed as ~ births divided by the population in state B. Simple algebra shows that such a transformation would be rBA' = TBA (PA/PB).Using the renormalized rate, the expression, RBA= rBA' X PB, is equal to the number of emigrants flowing out of B each year. The reciprocal of rBA', or (K - PA)/(PA X r g ) ~ the formula is used in the computer simulation to compute the state-B dwell times shown in Population Growth and Reincarnation 417 Table 2. However, there is no obvious value for K to use. A reasonable minimum value for K would be 10 billion souls. This figure is the projected peak of the human earthly population that will be alive at the end of the next century (United Nations, 1998). A reasonable maximum would be the number of humans who have ever lived (e.g., roughly 100 billion souls), because this figure would exhaust all of the human bodies into which the population of state B could possibly incarnate. Table 2 charts the implied mean dwell time in state B as a function of modern and historical crude birth rates under various assumptions for the constant sum of PA+ PB. The basic result is that changes in the dwell time of the unobserved state can reconcile reincarnation with population growth. This hypothesis was initially proposed by Stevenson (1974). Table 2 helps to formalize the implications of this earlier conjecture. Table 2 shows that historical fluctuations in the duration of stay in the unobserved state are consistent with historical changes in the observed human population. Although the dwell time in state A (e.g., life expectancy) has changed by a factor of 2 throughout human history, Table 2 shows that the simple CMM requires that the state-B dwell time change by a factor of as much as 2,000 over the same period to maintain population balance. The model predicts that the minimum possible average dwell time in state B in the year 2,000 would be around 30 years. Shorter dwell times for state B in the year 2,000 would require a smaller value of K and would be inconsistent with United Nation predictions that the world population will peak at around 10 billion. It should be stressed that the dwell time is a population average. Individual persons might experience either shorter or longer state-B dwell times, just as individuals might expect to live longer and shorter terrestrial lifespans than average. Discussion There are demographic methods to estimate the population dynamics of a population whose members migrate into an unobservable setting; however, these methods rely on interviewing immigrants about their transit time in the unobserved state. One could consider interviewing a sample of recent immigrants from state B to determine the average duration of stay in B. Because the reciprocal of the duration approximates rgA1,this information together with a count of the annual number of immigrants from B to A (RBA) would be sufficient grounds for an estimate of PB (Laska, Lin, & Meisner, 1997). Perhaps someday the database of interviews with children who recall past lives will be widely regarded as a suitable means of estimating the size and duration of stay of the discarnate population. Such return migrants would need to be asked how long they spent in the discarnate state to see if their responses accord with any of the predictions in Table 2. If nothing else, the database of interviews with children who recall past lives currently suffers from sample selection bias by all accounts. The children who recall past lives offer accounts TABLE 2 Population and Dwell Time in State A and B throughout History Stylized facts Crude birth rate (births per living personlyear) Crude death rate (deaths per living personlyear) Circular migration model State A population in persons State A dwell time in years K = 10 Billion State B population in persons State B dwell time in years K = 20 Billion State B population in persons State B dwell time in years K = I00 Billion State B population in persons State B dwell time in years Linear migration model State A population in persons State A dwell time in years State B population in persons Note: State A and state B can be identified respectively with incarnate and discarnate states. The variable K denotes the sum of population in state A + population in state B and is assumed constant in the variants of the simple circular migration model presented here. 50,000 B.C. 4000 B.C. 1650 A.D. 2000 A.D. 420 D. Bishai References Edwards, P. (1996). Reincarnation: A critical examination. Amherst, NY: Prometheus Books. Haub, C. (1995, February). How many people have ever lived on earth? Population Today, 23(2), 4-5. Keil, J. & Stevenson, I. (1 999). Do cases of the reincarnation type show similar features over many years? Journal of Scientific Exploration, 13, 189-198. Laska, E., Lin, S., & Meisner, M. (1997). Estimating the size of a population from a single sample: Methodology and practical issues. Journal of Clinical Epidemiology, 50, 1 143-1154. McNeill, W. H. (1976). Plagues andpeoples. Garden City, NY: Anchor Press. Mellars, P. (1989). Major issues in the emergence of modem humans. Current Anthropology, 30, 349-385. Stevenson, I. (1974). Some questions related to cases of the reincarnation type. Journal of the American Society for Psychical Research, 68, 395-416. United Nations. (1982). Model life tables for developing countries. New York: United Nations. United Nations. (1998). Worldpopulation projections to 2150. New York: United Nations. U.S. Bureau of the Census. (1997). Historical estimates of world population. Washington, DC: U.S. Bureau of the Census. Retrieved November 12, 1997, from the World Wide Web: http://www.census.gov/ipc/ www/worldhis.html Ziegler, P. (1969). The Black Death. London: Collins. Journal of Scientific Exploration, Vol. 14, No. 3, pp. 421430,2000 0892-3310/00 0 2000 Society for Scientific Exploration COMMENTARY The Mars Effect Is Genuine: On Kurtz, Nienhuys, and Sandhu's Missing the Evidence Znstitutfur Psychologie Gosslerstrasse 14, 0-37073 Gottingen, Germany 596 Villa Avenue Staten Island, NY 10302-1947 Abstract-Kurtz, Nienhuys, and Sandhu (1 997), instead of explaining the anomaly in their planetary data that we discovered (Ertel and Irving, 1997), launch a broadside attack against all claims of Gauquelin-type correlations. In our rejoinder, we provide evidence for highly significant Mars-effect results based on the skeptics' own samples. Regarding CSICOP's data problem, we suggested an inoffensive explanation, but Kurtz, Nienhuys, and Sandhu have, to date, precluded access to their original documents. Moreover, after looking at additional cues introduced by the skeptics themselves, we found the significance of CSICOP's data anomaly more conspicuous (p = .001) than with our earlier finding (p = .02). The skeptics' rejection of our evidence is thus unfounded. Keywords: Mars effect - planetary correlations - skeptics -CSICOP selection bias -IMQ-ISMQ Introduction An anomaly was uncovered in CSICOP's birth data of athletes. The validity of Kurtz, Zelen, and Abell's devastating verdict (1979180) on the Mars effect was therefore questioned (Ertel and Irving, 1997). Rather than reply to our critique, Kurtz, Nienhuys, and Sandhu (1997) take two tacks. First, they discredit all planetary correlations ever reported by the Gauquelins, Ertel, or Miiller. Second, they resort to ad hominems, accusing us of "tactics, tricks, tinkerings, irresponsible insinuations," usage of "abstruse" methods and "statistical manipulations," etc. We will deal with the first only, except where the second can be addressed by factual corrections. We also provide new results regarding our original question: Where did the anomaly in CSICOP s data come from?' 7 The Kurtz, Nienhuys, and Sandhu (1 997) article ignores nearly all of the 50 comments on an earlier draft we offered prior to its publication. 422 Suitbert Ertel & Kenneth Irving The Athletes' Mars Effect Our critics' attack on planetary effects fails in view of compelling counterevidence: Gauquelin's Mars-effect results are replicated by the data of his most critical opponents. Frequencies of athletes' births occurring while Mars is crossing the 36 sectors of its daily circle are given with deviations from expected frequencies (Figure I ) . ~ dashed curve is based on Gauquelin's The total (N = 4,384), the solid curve on the skeptics' total (N = 1,664) to which Committee PARA (CP) (Belgian), CSICOP (United States), and CFEPP (French) contributed (by N = 535,408, and 1,066, respectively), with overlap cases among skeptics excluded. It can be seen that the skeptics' and Gauquelin's results hardly diverge, and Kendall's Tau correlation between the two curves is highly significant (r = 3 . 4 4 , = .0003).' ~ Is there a Mars effect, i.e., an above-chance deviation of people born with Mars in "sensitive" sectors (generally designated as G%)? Narrow and Gauquelin-extended definitions of planetary G% have been proposed: 1.The narrow (classical) definition is as follows: The percentage of births with the planet in primary sectors-i.e., sectors 1, 2, or 3 (region after rise) and 10, 11, or 12 (after culmination). "Main sectors" (abbreviated M ) is an alternative term. 2.The Gauquelin-extended definition is the percentage of births as per the narrow definition plus the percentage of births in initial sectors 36 (just before rise) and 9 (just before culmination) (abbreviated I). 3.According to the Nienhuys-extended definition, the percentage of births in sectors 19, 20, and 21 (after set) and 28, 29, and 30 (after lower culmination) are added to the percentage of the Gauquelin-extended definition. Nienhuys called the set and lower culmination sectors, along with initial sectors 9 and 36, "secondary sectors" (abbreviated S ) . Is Definition 3 justified? Gauquelin had noticed early on that births of sports champions clustered in secondary sectors, although less conspicuously than in primary sectors. However, he never considered them, as in Definition 3, nor did he mention them in the course of the Mars-effect debate with skeptics. Secondary sectors were also generally neglected in the skeptics' studies. Definition 1 was generally taken as binding. Figure 1 shows deviations from expectancies, not expectancies; zero deviations thus represent a straight line (dotted horizontal) even though the expectancies are not quite uniform as in Figure 2 for births of physicians (expectancies are almost identical for Figure 1 athletes and other samples). Thirty-two percent of the skeptics' athletes are part of the larger Gauquelin sample. But this does not explain why the skeptics' and Gauquelin's Mars-effect deviations are alike. If the effect were due to Gauquelin's selection bias, it should not reappear in a sample gathered entirely by those highly skeptical of the effect. The Belgian sample was mildly affected by Gauquelin who assisted the data collectors (Ertel's discovery of 1996). However, indications of a Mars effect are present even after amending the sample (by adding data formerly excluded from the sample) or by analyzing the skeptics' data without the Belgian contribution (Ertel, 1998). The Mars Effect Is Genuine Mars sensitive sectors m Rise Culmination t Primary m Set Lwerculm &av Skeptics'data (N=1,664) : Gauquelindata(N=4,384): a@ o orig -&hed &hed ------ Fig. 1. Count of athletes' births (deviations from expectancy) across 36 Mars sectors, primary sectors (rise, culmination), and secondary sectors (set, lower culmination). Samples: skeptics (CP, CSICOP, CFEPP, N = 1,664), solid line and full circles; Gauquelin (published and unpublished, N = 4,384), dashed line and open circles. Interestingly, Nienhuys observed a considerable surplus of births in secondary sectors among data that Gauquelin urged the CFEPP to consider for its study (Benski et al., 1996, p. 125ff). He took this as support for the view that Gauquelin's data selections were biased without realizing that such insinuation was astray, because by adding Gauquelin's secondary sector cases, the CFEPP's Mars effect (by the agreed-upon Definition 1) cannot increase; on the contrary, the effect drops. Because Nienhuys took secondary sectors to support the skeptics' case, we deem it legitimate to consider them too expecting that via Definition 3, they will actually add pertinent information. Mars birth percentages (G%) for the above three definitions are given in Table 1, separately for the Gauquelin, the skeptics , and both samples combined. (The raw data, birth frequencies across all 36 Mars sectors, are available on a web-site.4) The binomial test shows that for the combined sample 7 Check with Internet URLs from Jim Lippard and Ken Irving: http://www. discord.org/skeptical/ astrology and http://www.primenet.com/kirving. Quite a few articles and discussions on the Mars-effect debate, book reviews, and a chronology of pertinent events are accessible there. 424 Suitbert Ertel & Kenneth Irving TABLE 1 Main Results Based on the Gauquelin and the Skeptics' Data Sets for Mars-Effect Definitions 1,2, and 3 (Narrow, G-Extended, and N-Extended) Data Gauquelin Skeptics Definition 1 2 3 1 2 3 1 2 3 N Obs. Exp. Mars G% p Kappa Combined Note: N = total number; Mars G% = percentage of cases with Mars in G sectors; Obs. = observed frequencies; Exp. = expected frequencies (percentages 17.4,23.2, and 39.0 for definitions 1 , 2 , and 3 respectively);~ error probability; Kappa = effect size. = and the Gauquelin sample alone the Mars effect is statistically very significant, whatever its definition. For the skeptics' sample alone, the Mars effect reaches this level only with Definition 3. The skeptics' lower G% and less extremep values are due, above all, to the inclusion of less eminent athletes (Ertel and Irving, 1 9 9 6 ) ~ apart from the smaller size of their sample. Regarding effect size, though it is lower for the skeptics' G% with Definitions 1 and 2, it is larger with Definition 3. In view of the above results and of additional evidence for the Mars effect in the skeptics' data (Ertel, 1994, 1996), our critics' allegation that the Mars effect is a mere "illusion" (Nienhuys in Benski et al, 1996) should be rejected. The Physicians' Saturn and Mars Effects Kurtz, Nienhuys, and Sandhu (1997, footnote 10) discredit Miiller and Ertel's (1994) test of planetary effects with members of the French Acadkmie de Mkdecine. The aim of that study was to replicate Gauquelin's 1955-1960 planetary correlations, based on the 1939 edition of the academy's directory (N = 576 members), now using the larger 1972 edition with N = 9 15 members. Counts of the physicians' births across Mars and Saturn sectors are shown in Figure 2. Birth excess occurs in rise and culmination zones. The error prob- Effects by Gauquelin's selection bias were undone by reuniting his unpublished cases (N = 1,503) with his published cases (N = 2,880). By contrast, the skeptics' samples remained uncorrected. Benski et al. (1996) refused to correct them, even in those cases where they acknowledged Gauquelin's criticisms as justified. The Mars Effect Is Genuine 425 SATURN . a -smoothed (FK-2) ------- MARS expected smoothed (FK-2) - - - - - - - expected Fig. 2. Count of births of members of the Academie de Medecine across Saturn and Mars sectors. (Saturn: solid line and full circles; Mars: dashed line and open triangles; expected frequencies: thin lines, Saturn solid, Mars dashed.) (From Miiller and Ertel, 1994, N = 1,083 .) abilities for Mars G% and Saturn G%, using Definitions 1, 2, and 3 are .06, .05, and .08, and .00007, .0005, and .03, respectively.6 Birth frequencies across 36 Mars and Saturn sectors, as deviations from expectancy, are significantly correlated (rho = 0 . 4 3 , = .004). Miiller and Ertel's study (1994) was in ~ no way influenced by Gauquelin. Kurtz, Nienhuys, and Sandhu's (1997) discounting of planetary effects in this sample as due to a "Gauquelin bias" thus cannot apply. Eminence Trends Regarding tack two-the attempt by Kurtz, Nienhuys, and Sandhu (1997) to raise suspicion about their critics' trustworthiness-we restrict ourselves to the eminence issue. "The eminence hypothesis falls dead flat" (Kurtz, Nien- For physicians, unlike athletes, significance levels of G-sector percentages decrease from Definition 1 to Definition 3. Relative contributions of primary and secondary sector frequencies to overall effect indicators thus vary among professions. The problem, although interesting, is not crucial. 426 Suitbert Ertel & Kenneth Irving huys, and Sandhu, 1997, p. 32). Question: Does G% increase with eminence in Gauquelin's data only (Ertel, 1988)? Answer: No. Eminence trends are also present in the skeptics' data, as shown in Figure 3. The three slopes are based, from bottom up, on Mars-effect Definitions 1, 2, and 3, and as in the case of general effect size (see above), the correlations improve as the definition is extended. In fact, by Definition 3, all six G percentages are above the line of chance expectancy and the regression is almost perfect. For the skeptics' data, eminence trends by Kendall's S for ordered contingency tables are significant (one-tailed tests), error probabilities beingp = .003,p = .001, andp = .0003 for Mars-effect Definitions 1,2, and 3, respectively.7 The results in Figure 3 are based on Ertel citation counts. Supposing these counts were flawed, as Kurtz, Nienhuys, and Sandhu (1997) would have it, the skeptics' own sources of eminence information should not replicate Ertel's results. Pertinent tests, however, checked by six independent researchers, strongly confirm them (see Ertel, 1998199). In view of these results, Kurtz, Nienhuys, and Sandhu's (1997) quibbling on Ertel's eminence procedure lose weight. Our critics merely stir up error variance by pointing at alleged faults of our use of biographical dictionaries (that they also contain items "not related to sporting achievement," etc.), disregarding the simple fact that part of any observed variance is error variance. Objectivity, i.e., the desired exclusion of subjective judgment, demands an inevitable price, an increase of error variance which, however, is unsystematic and therefore harmless and tolerable. The question at issue is not the presence of error variance and a corresponding decrease of precision, but rather how to explain highly significant covariance between citation counts and G% despite such a decrease of precision.* The Anomaly in CSICOP's Data We remind readers of our earlier discovery (Ertel and Irving, 1997) that the CSICOP data showed an inflated IMQ (initiallmain sector quotient) of 1.58 (IMQ was introduced in Ertel and Irving (1 997), and it represents the birth frequencies in the initial sectors divided by the average of birth frequencies in In Ertel (1996), using all Gauquelin professional data, differences between Definition 2 and 3 are investigated extensively. Definition 3 yields greater effect sizes generally while significance levels do not increase much, if at all. The choice among trend tests is arbitrary, but the results are generally alike. Significance levels by Kendall's Tau correlation for narrow, G-extended, and N-extended trends are .014, .014, and .005, respectively. For Spearman's rho, significance levels are ,005, ,011, and .009. Nienhuys criticized the eminence trend with Definition 3 as being "too good to be true." He alleged that it was due to "manipulations" performed "not blind," which in his view was "an alarming revelation" (Nienhuys, 1997). A charge of fraud is unmistakable. Unfortunately, the editor of Skeptiker did not allot any space for Ertel's defense consisting of abundant counterevidence. A reply appeared elsewhere (Ertel, 2000). Space allotment does not allow for providing grounds for rejecting Kurtz, Nienhuys, and Sandhu's (1997, p. 29) contention that Ertel and Irving's (1996) use of dictionaries was inconsistent. The Mars Effect Is Genuine - - - - - - Gauquelin data Skeptics' data Mars &ect percentby definition: Nenhysextenckd (vvithexpe&xy3) Ququelin abendxl (\nith qedancy 2) -(dh=P-wl) A 0 a Fig. 3. Percentage of births with Mars in critical G sectors (Definitions 1 , 2 , and 3), breakdown for six eminence ranks. Data: The skeptics' athletes total (N = 1,664). Expectancies for Mars-effect Definitions 1,2, and 3 are given as horizontal lines. Dashed slopes represent regression lines based on Gauquelin athletes (N = 4,384). the main sectors). Aere, the expected IMQ is 0.95. In other words, main sector athletes in the CSICOP data, which would be considered as desired candidates in regard to Gauquelin's Mars-effect hypothesis, seemed to be missing (i.e., missing denominator cases M raised IMQ from 0.95 to 1.58) compared to cases with Mars in the seemingly harmless initial sectors, who remained unaff e ~ t e dWas this due to rare chance, despite statistical significance ( p = .02), .~ or had CSICOP researchers excluded main sector athletes (leaving initial sector cases, not yet relevant at the time of their experiment, untouched), thus According to our estimate (details must be renounced), a deficit of about 25 Mars-positive cases exists in CSIOP's sample, which estimate is conservative (by adding 25 such cases, the corresponding IMQ would come down from an inflated 1.45 to an almost uninflated 1.09). 428 Suitbert Ertel & Kenneth Irving TABLE 2 IMQs and ISMQs for CSICOP's and ComitC PARA'S Data Sample CSICOP (N = 408) Comitk PARA (N= 535) Ordinary people (N= 24,614) Sector category Initial Secondary Main Initial Secondary Main Initial Secondary Main f 29 115 55 27 121 118 1,341 5,203 4,211 Average per sector IMQ ISMQ P Note:f = summed birth frequencies; IMQ = birth count average across initial sectors divided by birth count average across main sectors; ISMQ = as per IMQ, nominator includes births in secondary sectors. Ordinary people results are used for expectancies. Initial sectors are 9 and 36; secondary sectors are 9,36, 19,20,21,28,29, and 30 (initial sectors are included); and main sectors are 1,2,3, 10, 11, and 12. Column 4 represents the average er sector for the two initial, eight secondary, and six main sectors. Column 6 represents X goodness-of-fit test (expectancies from ordinary people). a Examples for the first two values of column 5 are as follows: IMQ = 14.519.2 = 1.58 and ISMQ = 14AL9.2= 1.57 (numerator and denominator values taken from column 4). * significant. very significant. Y lowering G% of Definition 1, while increasing tampering detector IMQ? We did not feel entitled to draw ultimate conclusions. The question, legitimate in view of our observations, is still in want of an answer. Fortunately, more pertinent information is provided by CSICOP's data. The "I" of IMQ is based on only two neglected sectors (9 and 36) considered in Mars-effect Definition 2. But, as noted earlier, Nienhuys introduced secondary sectors 19,20,2 1,28,29, and 30 in the Mars-effect debate, the sectors we added in Definition 3. We therefore feel entitled to replace IMQ with ISMQ (initial plus secondary sector birth counts divided by main sector birth counts). If CSICOP's inflated IMQ were fortuitous, as Kurtz, Nienhuys, and Sandhu (1997) would have it, the ISMQ-whose numerator IS (initial and secondary sectors) represents an average of births per sector for more than twice as many neglected sectors-should drop and might no longer be significant. The results, provided in Table 2, contradict Kurtz, Nienhuys, and Sandhu's "rare chance" hypothesis, as the ISMQ does not drop (ISMQ = 1.57, IMQ = 1.58). The ISMQ's inflated level, based on more nominator cases than the IMQ's level, is in fact more significant (p[ISMQ] = .001 versusp[IMQ] = .02) (see Table 21° for the rest of the results). The Mars Effect Is Genuine 429 TABLE 3 Counts of CSICOP Athletes' Births in Main, Initial, and Secondary Sectors (Columns 1 4 ) , IMQ and ISMQ With Chance Probabilities (Columns 5-43), Breakdown for Unproblematic Batch 1 (Row 2) and Problematic Batches 2 and 3 (Row 3), and Differences (Row 4) Batch 1 2 3 4 Total Batch 1 Batch2 & 3 Difference batch (2 & 3) vs. batch 1 N 408 128 280 - Main Initial Secondary sectors sectors sectors IMQ 55 25 30 29 6 23 115 35 80 - PIMQ ISMQ 1.57 1.05 2.00 0.95 PI~MQ 1.58 0.72 2.30 1.58 .026* NS .001** .02* .001** NS .001** .04* - - Note: IMQ = birth count average across initial sectors divided by birth count average across main sectors; ISMQ = as per IMQ, nominator includes births in secondary sect0rs.p by chi square in rowf*l to 3. p by Fisher's exact test in row 4. significant. very significant. ISMQ analysis allows for another scrutiny. As noted earlier, CSICOP researchers' data anomaly appeared in their second and third batches only, collected with an apparent fear of failure, not in their first batch collected with hope of success and thus without bias. Batch 1 results were surprisingly proGauquelin. An inconspicuous IMQ and ISMQ (the two bias indicators) is therefore expected for Batch 1 while the two quotients should rise with Batch 2 and 3. Results in Table 3 are as expected (for IMQ, see columns 5 and 6 , and for ISMQ, see column 7 and compare lines 2 and 3). An increase of IMQ and ISMQ from Batch 1 to Batch 213 is apparent in line 4 of Table 3, and both shifts are significant by Fisher's test. l1 Despite an increase of evidence for the reported anomaly in CSICOP's data, it should be maintained that one cannot be absolutely certain that they are due to case selections corrupted by knowledge of Mars positions. But this question may be resolved by checking certain documents in the files at CSICOP's headquarters in Buffalo, New York. We suggested, "CSICOP might invite ' To avoid laborious randomization, ISMQ's significance was based on Fisher's exact test. O Gauquelin's "ordinary people" (N = 24,614, see table 2) served as controls. Fisher's significance level for CSICOP's IMQ ( p = .02) compares to our formerp = .02 from randomization and may therefore be trusted. An unexpected large negative deviation of IMQ for the Belgian (Comite PARA) data disappeared with the ISMQ and was thus apparently random. The only still unexplained anomalous results (IY? and ISMQ) are CSICOP's. Kurtz, Nienhuys, and Sandhu (1997) attempted to reject our IMQ indicator through use of a computer simulation. A critique of this case must be renounced here. (The respective section of our paper is obtainable, on request, from S.E.) 430 Suitbert Ertel & Kenneth Irving critical non-CSICOP researchers to check the original lists of data and the correspondence with birth registry offices." Has CSICOP acted accordingly? To date, we know of no independent examination; one CSICOP-friendly associate who was invited to check with CSICOP's files eventually resigned. Conclusion The anomaly in CSICOP's database remains unexplained. Kurtz, Nienhuys, and Sandhu have ignored our readiness to invalidate, by finding appropriate proof among CSICOP documents, one reasonable, though possibly correctable, explanation. But one need not insist on resolving this ambiguity. It is in fact inconsequential because, as Figures 1 and 3 and Tables 1 through 3 have shown, CSICOP'S data, despite obvious distortions, still contribute to an unambiguous overall Mars effect. So do the data of the French Skeptics who wanted to get rid of it (Ertel, 1998199). In our view, the evidence for Gauquelin's Mars effect today is stronger than ever before. It seems so strong, that the skeptic committees' allegation that "it isn't there" is almost as enigmatic as is the fact that it is there. It is time to face the reality of Gauquelin's tenacious planetary phenomena and to design the next stage of research . References Benski, C., Caudron, D., Galifret, Y., Krivine, J. P., Pecker, J. C., Rouzt, M., Schatzman, E., & Nienhuys, J. W. (1996). The Mars effect: A Fench test of over 1,000 sports champions. Amherst, NY: Prometheus Books. Ertel, S. (1988). Raising the hurdle for the athletes' Mars effect. Association covaries with eminence. Journal of Scientific Exploration, 2, 53-82. Ertel, S. (1994). Mars effect uncovered in French skeptics' data. Correlation, 13(2), 3-16. Ertel, S. (1996). How to suppress the Gauquelin Mars effect. Strategies of concerned committees. Correlation, l5(1), 2-16. Ertel. S. (1998). Man tut sich schwer mit dem Mars-Effekt. Zu einer Buch-Rezension N. Rohrkohls. Grenzgebiete der Wissenschaft 47,269-274. Ertel, S. (1998199). Is there no Mars effect? The CFEPP's verdict scrutinized with the assistance of six independent researchers. Correlation, 17(2), 4-23. Ertel, S. (2000). Debunking with caution: clearing up Mars effect research. Correlation 18(2), 9 - 41. Ertel, S., & Irving, K. (1996). The tenacious Mars effect. London: Urania Trust. Ertel, S., & Irving, K. (1997). Biased data selection in Mars effect research. Journal of Scientific Exploration, 11, 1-1 8. Kurtz, P., Zelen, M., & Abell, G. (1979180). Results of the U.S. test of the "Mars effect" are negative. The Skeptical Inquirer, winter, 10-26. Kurtz, P., Nienhuys, J. W., & Sandhu, R. (1 997). Is the "Mars effect" genuine? Journal of Scientific Exploration, 10, 1-2 1. Miiller, A., & Ertel, S. (1994). 1,083 Members of the French Acadkmie de Midecine. Astroforschungsdaten (Vol. 5). Waldmohr: A. P. Miiller. Nienhuys, J. W. (1997). Ertel's "Mars-Effect": Anatomie einer Pseudowissenschaft. Skeptiker, 10, 92-98. Wunder, E. (1997). Der "Mars-Effect": Besteht noch Forschungsbedarf! Skeptiker, 10, 99-105. Journal of Scientific Exploration, Vol. 14, No. 3, pp. 431-446,2000 0892-33 10100 O 2000 Society for Scientific Exploration 1 I Bulky Mars Effect Hard to Hide: Comment on Dommanget's Account of the Belgian Skeptics' Research Institut fur Psychologie Gosslerstrasse 14, 0 - 3 7073 Gottingen, Germany Abstract- In 1997, Jean Dommanget gave an account of investigations on the Mars effect claim conducted by the Belgian skeptics' committee PARA. Gauquelin had shown that their data revealed a strong Mars effect. Dommanget rejects Gauquelin's claim, alleging justification through the committee's "model of the mechanism." The present paper scrutinizes the committee's model using a simpler data set to facilitate its intelligibility. It turns out that the committee's model is misconstructed. The Belgians' arguments resort to confused complexities through which apparent evidence for a Mars effect is dispelled. Gauquelin's early but insufficiently founded objections to the committee's critique now appear fully justified. The Belgian committee is urged to clarify persisting contradictions in the early history of its resistance to Gauquelin's discovery. Keywords: Mars effect - planetary correlation skeptics -Comite Para - Gauquelin claim - 1. Introduction Jean Dommanget, chairman of Comitk PARA (CP), the skeptical Belgian committee for the scientific investigation of claims of the paranormal, has given an account of the committee's early research into the Mars effect (1 968-1976) and of a sequel to that work (Dommanget, 1997). Their research started with a replication of Gauquelin's Mars effect claim on a sample of 535 sports champions, through which Gauquelin's prediction found strong support. CP, however, eventually found problems with Gauquelin's Mars effect calculation, above all with expectancy (Committee PARA, 1976). This issue is of prime importance because the Mars effect is defined by a significant deviation of observed birth frequencies, partitioned across 12 diurnal Mars sectors, from corresponding expected birth frequencies.' CP proposed "a model for the theoretical mechanism of the purported phenomenon" (p. 275). Their model, the Belgian critics allege, invalidates More precisely, a planetary effect is defined by significantly deviating birth frequencies occurring while the planet occupies certain sectors, e.g., Mars sectors 1 and 4 on a 12-sector scale, for ComitC PARA'S experiment. 432 Suitbert Ertel Gauquelin's expectancy calculations. If this were true, the Mars effect claim and all other planetary claims of the Gauquelins must be abandoned. A scrutiny of CP's position, as presented in Dommanget's account, is therefore urgent. A note on the intelligibility of CP's model is prompted by Dommanget's repeated complaint that "misunderstandings have unfortunately occurred with nearly all...authors who started studying this question. Many authors did not understand (or did not want to understand) ...[our] position"2 (p. 279). Dommanget should consider, however, aside from possible limitations of his readers, possible deficiencies of his own presentations. After wading through his puzzling text, after repeated queries addressed to him (to some of which he kindly replied), and after receiving answers to pertinent queries from others who labored through his a r t i ~ l e a ~ , review is now due. Novices in the field might welcome my clarification of elementary concepts. Every test of an association between planetary positions and births of professionals requires expected birth frequencies along the planet's daily path, subdivided among planetary sectors 1 through 1 2 . These serve as a reference ~ to evaluate observed birth frequencies in individual sectors or sector zones (combinations of sectors). One broad consensus, encompassing both detractors and supporters of the Mars effect, exists on the rationale underlying the necessary calculations: Expected birth frequencies must accommodate astronomical and demographic factors. These two sources of intrusive variance, not at first familiar to many readers, will become intelligible by using an easier to understand analogous problem invented for the purpose. 2. Astronomical Factors The Gauquelin Approach Suppose someone claims that eminent singers are more often born between sunrise and sunset than would be expected by chance. Doubt prompts us to test this claim, so we collect birth data (date and hour) of N = 535 singers. Note Dommanget (1997) does not refer in his article to his correspondence with astronomer George Abell who apparently did understand CP's model and who was the first to discover certain flaws in it. Abell, in a letter to Dommanget, dated October 16, 1982, wrote, "I must say that I have read and reread your paper, and I think I understand it but am still perplexed about a matter, as I shall explain below." I am grateful for Geoffrey Dean's, Ivan Kelly's, Ken Irving's, and Marcello Truzzi's suggestions for improvement. ~ u to the earth's rotation-one turn in 24 hours-Mars seems to circle the earth. The planet's full e circle of 24 hours, seen from, say, Brussels, can be divided into 12 sectors: six above and six below the horizon. For example, champion Jean de Bie was born in Brussels at 17h00, while Mars, seen from his birthplace, was in sector 9. Had de Bie been born at the same time (17h00), but 4 months later, he would have had Mars in sector 10. Another 4 months later, again at the same time, Mars would have been in sector 11, etc. Bulky Mars Effect Hard to Hide 433 that the claim of our fictional interest refers to only two solar "sectors" (roughly day and night), instead of to 12 Mars sectors, as in CP's research on sports champions, which demands more attention. What do we expect by chance? Depending on the calendar day, the sun can be above the horizon for periods of varying duration between very short, roughly 8 hours (winter minimum), and very long, 16 hours (summer maximum), assuming geographical latitude of 50 N. For each singer's birth date and location, we therefore need to know at what time the sun rose and set. This information is needed to calculate sun-above-horizon proportions for each individual birth date, i.e., the probability ,pup of the sun's presence above the horizon (subscript a for "astronomical" and up for "above horizon"): hup is the difference between the sun's set and rise time on an hourly scale. Probabilities ,pupare diurnal (daily) proportions of sun-above-horizon periods. They vary, taking the above hup range 8 to 16 hours, between 0.33 (winter minimum) and 0.66 (summer maximum). For each singer's birth date, we obtain one sun-above-horizon proportion. Summing them over N = 535 birth date occurrences, we obtain the expected number of births with the sun above the horizon, a Nup. The expected number of births with the sun below the horizon is simply a ATdown = 1 - a Nup.Equation 2 accommodates the astronomical factor (index a) arising in our project from the sun's season-dependent variation of its daily circle. Demographic complications are brought in be10w.~ Michel and Francoise Gauquelin's method to obtain expected birth frequencies for planetary sectors accommodates astronomical factors in the same way as our example, though using Mars, not the sun, and dividing the diurnal motion into 12 sectors, not two as in the case above. Critical astronomers and statisticians (Abell et al, 1982; Couderc, Porte, Rawlins, 1978), most of them members of skeptical committees, have approved of the Gauquelin procedure, in some cases after subjecting it to detailed scrutiny In view of problems with CP's approach, to be discussed, it should be emphasized again that the total of expected births of a sample requires one probability for each individual of the sample. Details on Gauquelin's algorithm are provided by Gauquelin and Gauquelin (1 957), Gauquelin (1 988a, p. 203-217), and Gauquelin (1988b). 434 Suitbert Ertel CP's Approach CP'S researchers, however, contest Gauquelin's rationale underlying his expectancy calculation. Their arguments are puzzling, and unraveling them requires space and time. CP's model is based on a matrix, an arbitrary and unusual construction. The matrix comprises six rows and nine columns, i.e., n = (6) x (9) = 54 cells, whose entries are so-called configurations, tagged Ck(see their table 2, Dommanget, 1997, p. 285). Configurations are frequencies of Mars in 54 combinations of Mars and Sun daily positions. They are calculated for the first of each month of the experimental period, i.e., for the 888 months making up 1872-1945. Configuration, by the way, is a misleading term to denote what is actually a cell unit in the matrix, just as the term wall would be misleading to denote a brick.7 Readers become confused all the more because Dommanget occasionally uses the term configuration to denote the 12 sectors of the daily Mars circle (which is quite different from what is represented in table 2). In such cases, configuration would mean "the set of the twelve classes [class is Dommanget's unusual term for sector]" for any particular date (Dommanget, 1997, p. 283). Why did CP set up table 2 at all? We will postpone this question until we are through with all necessary technicalities. Our understanding of Dommanget's table 2 is facilitated by creating an analogous table for the singers project. We follow his rules and sort all sunabove-horizon periods, observable within the time span 1872-1945, into eight intervals (irrespective of whether the singers' birth dates actually occurred in the sun-above-horizon periods) starting with the shortest possible 1-hour period 08h00-08h59, 09h00-09h59, etc. up to the longest, 15h00-15h59. Next, we partition each year of the observational period, 1872 - 1945 = 74 years, into 12 calendar months (January, February, March, etc.). A matrix of 8 rows (representing eight sun-above-horizon periods) by 12 columns (representing 12 months) is thus established, comprising n = (8) x (12) = 96 cell units Ck(k equals 1-96). Next, following Dommanget, the cells of the matrix are filled with counts of days. One day from each month of the observation period is entered into the table; Dommanget selected the first of each month, so we follow suit. To give an example, on January 1,1872, the sun was 8h07 above the horizon (obtained from computer software or from an ephemeris), so we put 1 in the first column for January and first row for 8h00-8h59. The last entry is December 1, 1945 (column 12, December and row 1, 8h00-8h59 sun-above-horizon duration on January 1, 1945). The table eventually contains (1 2 ) x (74) = 888 days distrib- Bulky Mars Effect Hard to Hide 435 uted among 96 Ck cells comparable to Dommanget's sample of 888 days, which in his case were distributed among 54 Ckcells. In Dommanget's table 2, Mars '-stay-in-sectors periods (rows) correspond to the sun-stay-above-horizonperiods of our singers table, and his Mars-Sun distances (columns) correspond to our months. To make the analogy more explicit, month may be roughly conceived as distance between the sun's present and its winter solstice position on the ecliptic. These technicalities are needed to understand the essentials of CP's procedure. Table 2 of the singers study, when filled with day counts, shows that sunabove-horizon periods are longest in summer months and shortest in winter months with gradual changes between them. We are aware of seasonal variation, of course, without consulting this matrix. Many people, however, are not aware of Mars' celestial motions. They will discover in table 2 of CP's study, when filled with day counts, that Mars-sector periods manifest lawful variation no less than the sun's "sector L" periods. That is, Mars, too, has "seasons," although roughly twice as long as solar seasons. This might be interesting for novices in celestial motion. But how to get from here to expectancies for births in Mars sectors? CP's model, taking shape with their formula 6, was constructed to somehow transfer Ckcounts onto sectored birth expectancies (Dommanget, 1997, p. 286). Their figure 5, a pictorial aid with many arrows, is introduced to make the intended transfer visually persuasive. But this figure is puzzling, not persuasive, for three reasons. First, to arrive at expectancies for births in Mars sectors, there is no need to adopt CP's Ck statistics. In my view, which I share with CSICOP astronomer George bell,^ CP researchers are forcing superfluous data about celestial Mars movements on human birth expectations. Resuming the singers study, we can straightforwardly and safely state that the sun-above-horizon proportion, e.g., for May 2, 1903, when Bing Crosby was born (4 p.m., in Tacoma, Washington, at 47.14N, 122W26) is determined solely by sunrise and sunset in Tacoma on May 2, 1903. Secular (seasonal) changes of the sun's rise and set before and after Crosby's birth date, i.e., the day count statistics of Dommanget's table 2 over 74 years of observations, are entirely irrelevant for the sun-above-horizon period on Crosby's birth date and place. The same applies to Mars-in-sector pe- In one respect, Ck entries (day counts) for CP's champions project (Mars) differ from Ck entries (day counts) for the singers project (sun): CP's Mars day counts in the champions project vary on the secular time scale: Analogous sun day counts for the singers project change within years only, not over the years. This, however, is of no account here. Abell (in a letter to Truzzi, November 10, 1982) already pointed at the irrelevance ofpCk: "The pCk...is simply the probability of obtaining a particular configuration of Mars in the sky at the time of the athlete's birth. I think it may be relevant if one assumes a statistical average for Mars distributions, but not if the theoretical frequencies are calculated, as I have indicated in a letter to him." g 436 Suitbert Ertel riods and sports champions. CP's model, in a generalized perspective, amounts to a nonsensical claim, namely that effect Et, caused by Ct at calendar time t, is affected, aside from C , by Ca,...Ct-3, Ct-2, Ct-1, Ct+1,Ct+2, Ctlt3,.. + . Cz, i.e., by the entire time series ta through tz of continuous Ci occurrences, past and future, of which Ct is one temporal slice in the middle. Second, understandably therefore, CP's model cannot be put into operation. The transferring device of their formula 6 is incomplete. It does not and cannot indicate how Ck-the information in their table 2-should be translated into expectancies. The arrows in Dommanget's figure 5 (1997) are misleading pictorial substitutes for a missing formalization, which in turn seems to be due to a confused conceptualization. Third, we have shown above that calculations of expected birth frequencies for N = 535 champions must be based on Mars-in-sector durations for each champion's birth date. In CP's model, however, expressed by their formula 6, no link to those 535 individual birth date parameters exists (index i refers to sectors, index k to day counts of their table 2, there is no index for individuals): For each of 12 sectors i, one probability pi is formalized by summing over 54 Ck-based probabilities instead of across 535 person-based probabilities. in What purpose does the expressionpi(Ck CP's formula 6 serve?, Apparently, it should bring k-indexed Ck-day counts and i-indexed Mars-sector observations together. But readers are not told to which observations pilCkrefers, and they will hardly dare to question this construction, which Dommanget refers to as "well known" even though no one can have encountered an odd formula such as this one before. More discussion was needed to fully dissolve the intricacy. The main point is that in CP's model, no variable referring to N = 535 birth dates is provided on which, as pointed out above, expectancy calculations must be based. Whatever the model may be, N individual expectancies must be summed and they must add up to some value around 535112 = 44.5. Why did CP ignore individual birth date information? Dommanget's answer to this brings deeper factors to light. He holds that a researcher who wants to "compute the theoretical histogramlo...is not allowed to make use of the sample itself" (Dommanget, 1997, p. 279). For him, using "the sample itself to establish the reference diagram [is a] disadvantage" (Dommanget, 1997, p. 292). This is the exact opposite of what proper methodology demands: Pertinent information obtained from the sample itself--i.e., from each lo ~ o m m a n ~ e t ' s terms "theoretical histogram" and "model of the mechanism" sound substantive and significant, yet they are ambiguous. Replacing "theoretical histogram" with "expectancies" increases conceptual precision. Bulky Mars Effect Hard to Hide 437 individual in the sample-is neither a disadvantage nor an advantage, rather it is an indispensable source of reference. Expectancies are required for an actual sample and not, say, for the entire world population. This goes without saying among experts calculating expectancies, whether by randomization, Monte Carlo, or by bootstrapping procedures, etc. Being based on misconceived statistical presumptions, CP's model is not appropriate to accommodate birth-related astronomical factors. 3. Demographic Factors The Gauquelin Approach Aside from astronomical variance ( a ) , the average variance of birth frequencies over hours of a day (d) must be considered: People are not born with uniform frequency day and night. Hourly birth rates for natural births generally peak before sunrise, decline during the daylight hours and are lowest in the afternoon. This so-called demographic or d factor, however, is easily accounted for with the help of average diurnal birth counts, occasionally called nycthemeral distributions. Let us use the distribution of ordinary people from Gauquelin studies (No, = 24,6 14)" and calculate from them expected birth rates as needed. For our-simplified singers example, the question in need of an answer for each birth is this: Which proportion of a day's general birth output is obtained between the time of sunrise and sunset occurring on the subject's birth date? This is the demographic birth proportion required to adjust the astronomical birth proportion as explained above. We proceed as follows. Above we obtained, for Bing Crosby's birth date May 2, 1903, a sun-above-horizon proportion of ,p = 0.60. On that day, the sun rose at 5:53 a.m. and set at 8:18 p.m. Next, we take birth counts of ordinary people, i.e., for the time section between 5:53 a.m. and 8:18 p.m. (sunabove-horizon considerations are here irrelevant), and find that this temporal section contains 59.5% of the total. Thus, the probability for a person to be born on May 2, 1903, with the sun above the horizon was . ~ d p= 0.595. In a similar fashion, just as for Crosby's birth date, we obtain .ldp values for each singer of our sample. The expected number of births with the sun above the horizon, demographic condition adjusted, is obtained analogous to Equation The expected number of sun-below-horizon births is simply aldNdownN = The Gauquelins collected the data for their so-called heredity studies attempting to test possible similarities of planetary positions between parents and their offspring. '' 438 Suitbert Ertel oidNup. Equation 2 considered astronomical factors only while Equation 3 takes into account astronomical and demographic factors. Equation 2 served merely as a didactic first step. We have made it clear how Gauquelin's expectancy algorithm works in principle, using singers and two simple solar-sector expectations. For sports champions and 12 Mars-sector expectations, the same principle applies. Additional complications are few and can be passed over here.12 CP 's Approach How does CP's "model of the mechanism" accommodate demographic factors? Surprisingly, this model (their formula 6) ignores demographic conditions. This is puzzling because Dommanget (1997) actually refers, in passing at least, to "taking into account the nycthemeral distribution" (p. 285). He seems to make readers believe that demographic variation has been incorporated into the model. He even refers to Gauquelin's nycthemeral birth curve (Dommanget, 1997, figure 4) as ifCP's formula would account for it, but it does not. Moreover, remember that Dommanget's main objection to Gauquelin's algorithm points at allegedly wrong demographic assumptions, not wrong astronomical assumptions, i.e., CP objects to the neglect in Gauquelin's algorithm of secular inconstancies of the diurnal birth rhythm. But on close inspection, one finds that CP's criticism of Gauquelin's demographic assumptions does not result from their formalized model. Rather it is a mere verbal patch to cover a hole in what CP is really saying. In view of CP's verbalized concern with demographic factors, no one would expect that these factors are entirely neglected in their own formula. Dommanget's summary of CP's position turns the facts upside down: "[CP established] a model for the theoretical mechanism of the purportedphenomenon. [From this analysis,] it clearly appears that the theoretical principles proposed by Gauquelin to support his research have to be rejected because they do not correctly take into account the fundamentals of the problem-the secular and diurnal sociodemographic factors (Dommanget, 1997, p. 275f). However, CP's "theoretical mechanism" cannot logically provide any grounds to reject Gauquelin's procedure. The reason is that it is unrelated to sociodemographic factors. Can CP's "only correct formula" (Dommanget, 1997, p. 290) be put to a test? This is a reasonable question, and a straightforward response would be to just test its expectancy output. Yet, here we encounter an astounding deficiency. An expectancy output from CP's formula has never been provided, despite early and repeated demands, especially by Gauquelin (1982): "He [Dom" l 2 Unfortunately, Dommanget's account of the Gauquelin procedure is as unintelligible as that of his CP procedure. Bulky Mars Effect Hard to Hide 439 manget] should publish his own theoretical...distribution, [but] he never did so" (p. 77). Dommanget, when challenged by such demands, used to object that expectancy computations are "unfortunately impossible," since according to CP's model, all factors contributing to diurnal birth curve variation must be considered, but that somelmany factors are "unknown" (Dommanget, 1997, p. 294). But we may simplify the test by creating a fictitious sample without any diurnal birth curve variation, assuming that birth frequencies are evenly distributed over 24 hours of the day and that no secular variation occurs. What is the expectancy output of CP's "only correct formula" for the simplest of all conceivable conditions? There is no such output. CP's formula manifests its uselessness convincingly under most favorable test conditions. l 3 CP's Main Objection to Gauquelin's Demographic Conception CP was concerned that observed birth peaks on a Mars-sector scale might be due to secular variation of diurnal birth distributions. It must be conceded that such demographic artifacts, even though improbable, are at least conceivable. The CP's concern over them does not logically result from their model, but one may look at them regardless. So I checked CP's claim, but on close examination of their criticism of secular variation, I did not find any evidence for it. In contrast, elementary logic and empirical facts contradict it, as is shown in what follows. First, recall the Gauquelin claim that whenever deviations of birth counts from chance occur, they invariably do so in key sectors 1 (rise of the planet) and 4 (culmination), irrespective of planet (Mars, Jupiter, Saturn, Moon, etc.) and profession (champions, actors, physicians, etc.). CP's suspicion that Gauquelin planetary effects might be due to secular demographic variation thus implies the contradiction that invariable deviations of birth counts in invariable sectors arose from variable and unpredictable demographic causes. This is like claiming that weekly maxima of accident cycles on Saturdays were due to weather factors. No doubt, the weather is known to affect accident occurrences, but its influence for particular weekdays is entirely unpredictable. The second argument is empirical. Dommanget, aside from ignoring constancies of birth frequencies for Mars sectors 1 and 4, tried to emphasize inconstancies of birth frequencies over Mars sectors 1 to 12. He used his figures l 3 Dommanget (1997) attempts another way out of CP's dilemma: "It is surprising that nobody seems to be aware of the need for computing the theoretical histogram" (p. 290). However, Abell computed a theoretical histogram (expectancies) in 1982 (Dommanget received the result by letter), Gauquelin published more than a dozen volumes of data including expectancies, and Miiller, Ertel, CFEPP, Nienhuys, and Pottenger computed expectancies that agreed quite well with one another. Dommanget is aware of at least some of these successful efforts. CP is the only party not providing expectancies. Dommanget thus turns crucial facts upside down, thereby misleading inattentive readers and puzzling attentive readers. 440 Suitbert Ertel 9 through 11 to demonstrate how the distribution of births over Mars sectors varied over the years. In his figure 9, e.g., we find a breakdown of birth rates over Mars sectors for three successive secular periods. From eyesight, Dommanget (1997) concludes that "there is a dependency of the results from the secular sociodemography" (p. 290), meaning that (a) over secular periods, birth frequencies for Mars sectors are inconstant, and (b) this is caused by corresponding secular variation of daily birth distributions. Dommanget's focus is on Claim 1 only, the alleged effect. He does not consider Claim 2, the alleged cause, which is a remarkable omission. Worse, he actually fails with both claims, as shown in Figures l a and b. Figure l a represents birth rates of CP champions on the Mars-sector scale after a breakdown of the total into Dommanget's three successive cohorts. His figure 9 (Dommanget, 1997, p. 291) showed the same three curves, though side by side and small, which impedes perceiving invariant features. Such features manifest themselves clearly by an overlay display, as in Figure la: Peaks in Mars key sector 1 are invariably predominant. The synchrony of key sector 4 peaks is also present, although less conspicuously. More to the point, the three birth count distributions are significantly correlated among each other (by Spearman's rank correlation rs ): In view of the smallness of the three subsamples, each subdivided over 12 sectors (leaving on the average circa 15 champions per sectorial unit), the indication of stability in Figure 1a is remarkable. Dommanget did not calculate such correlations, and what he did instead actually veils them. For his figure 10 (p. 29 I), he doubled the subdivisions of the same data (using six cohorts instead of the three of his figure 9), thus throwing even more random noise into his readers' eyes, with no additional information provided. As noted above, Dommanget claims that Mars-sector variation (the alleged effect) is due to diurnal birth time variation (the alleged cause), but he did not actually investigate the latter. Making up for this neglect, we show pertinent results in Figure lb. For the three Dommanget cohorts, birth frequencies are plotted over time of the day. Dommanget would predict that variation of birth rates over time of the day is more marked than variation of birth rates over Mars-sector position, because causal oscillations should be stronger than effected oscillations. Figure lb, compared with Figure la, however, does not bear this out. Birth + Fig. 1 Birth counts of CP's athletes, separately for three CP-defined successive cohorts, (a) across 12 Mars sectors, (b) across 12 daytime periods. Bulky Mars Effect Hard to Hide e e -----.Cohort 1 Cohort 2 0 ....... Cohort 3 Rise t t Culmination t s t t Lowr Culmin. Mars sector (b) 30 ----.Cohort 2 ....... Cohort 3 -Cohort 1 ;: ;: 3 Time of the day 442 Suitbert Ertel rate variation is smaller, not larger, for hours of the day (Figure lb) than for Mars-sector position (Figure la). It is inconceivable that a strong effect (represented by ragged Mars-diurnal (sector) distributions with large peaks and troughs) could ever result from a weak cause (a smooth solar-diurnal distribution). Only one of 3 x 12 = 36 counts is conspicuous, a statistical fluke probably. Dommanget (1997) concluded, again from mere eyesight, not from measurement, that there exists a "dependency of the shape of the histogram [births over Mars sectors] on the secular distribution of the births dates." This, he concludes, "confirms the role of the secular sociodemography" (p. 294). Figures l a and b show that his conclusion is wrong. The evidence leads to the opposite conclusion. 4. Missing Result: Expected Birth Frequencies As noted above, CP's work of 1967-1997 did not culminate in the publication of expectancies. Surprisingly, however, as early as 1970, before inventing their "model of the mechanism," CP actually did calculate expectancies using a widely accepted method. However, they did not publish these results. In his "rough historical sketch," Dommanget skipped this period of CP's intensive research. Why did he leave this out? Let us first look at CP's unpublished expectancies (see Figure 2). The solid line in Figure 2 shows, first, observed birth rates of CP's champions. Full circles represent CP's expected birth rates. These were obtained by shifting the birth time from each actual champion to the next in line when listed alphabetically. Birth year, month, and day remained unchanged. Full circles in Figure 2 represent the average sector positions of nine such successive shifts. They are taken from an unpublished Dommanget document. l 4 CP's expectancies in Figure 2 may be compared with expectancies obtained by Gauquelin's classical technique (explained above). This technique yields birth rates of a fictional control group whose size equals that of the experimental sample (small open circles). Small open triangles represent expectancies obtained by using another strategy. Expectancies were here obtained by summing sector positions of ordinary people (averages of a noneminent population). Ordinary people may be regarded as a genuine control sample, which, however, is much larger than the experimental (champions) sample. (The Zelen test method is more precise, but of the same kind.) It can be seen that expectancies by CP's shifting procedure hardly differ from those obtained by the two other methods.15 l4 Source Dommanget, "Tableau extrait d'une note de M. Dommanget," 1970, an unpublished appendix of a letter by de MarrC, a copy of which was transmitted to Gauquelin, Hoebens, Hovelmann, and Ertel. l5 CP's randomized expectancies show larger variance than the other two, which apparently indicates that nine randomizations are insufficient for obtaining an average curve as reliable estimate. The CFEPP ran 80 randomizations; Ertel, 200. Bulky Mars Effect Hard to Hide Observed: Comite PARA athletes (N=535) Expected 1: Gauquelin's standard calculation . ..Am- Expected 2: Gauquelin ordinary people (N=24,614) 4 Expected 3: Cornite PARA'S randomizations of their data 443 - -c- - Rise t Culmination t t Set Lower Culmin. t Mars sector Fig. 2 Observed birth counts (%) of CP's athletes across 12 Mars sectors and pertinent expectancies, CP's unpublished ones included. The Mars effect controversy is certainly worth being settled and CP could help at that by clarifying the following details of their investigation: 1. In his paper, Dommanget (1997) contends that the results of CP's randomization experiments, counterexperiments using his terminology, "were ready for publication in the years 1977-1978 in our N.B. No. 44" (p. 292). But documents exist showing that Dommanget had tabulated the results of these experiments already by 1970.16CP member Luc de Marre expressly confirms this in a letter to Kurtz (November 8, 1975). He informs Kurtz that the counterexperiments were conducted "be- l6 The table was attached as a copy of a letter (April 2 1, 1975) by CP's member Luc de Marre addressed to Jean Dath, chairman of CP at that time. 444 Suitbert Ertel tween 1967 and 1970." De Marrk complained that "during three years [1970-19731, the PARA committee refuses to publish the results of this investigation," and he resigned from the committee in protest. 2. According to Dommanget (1997), CP's counterexperiments "were conducted by PARA.. .on the basis of its theoretical conception of theproblem [italics added]" (p. 292). This is puzzling for two reasons: First, as just noted, the counterexperiments were conducted right after CP's data collection in 1968. Their theoretical model was developed later, around 1973-1975 (de Marrk). CP's experiments thus cannot have been conducted on the basis of CP's "theoretical conception of the problem." Second, following Dommanget, CP's theoretical conception of the mechanism contradicts the assumptions on which their own counterexperiments are based, i.e., CP's own experiments have "the disadvantage of using the sample itself to establish the reference diagram" (Dommanget, 1997, p. 292). Even on logical, not only temporal grounds, therefore, the committee's model cannot have served as the basis for their counterexperiments. 3. The results of CP's counterexperiments, which were "ready for publication in 1978 (N.B. No. 44)," were not published, according to Dommanget (1997), "because no one seemed to accept our analysis of what was needed to understand these experiments" (p. 292). However, the French skeptics (CFEPP), Nienhuys, and Ertel, all used the CFEPP's early technique independently without any consideration of the CP model of 1976. An acceptance of this model is not needed for understanding and applying their procedure of 1970. 4. According to Dommanget (1997), research on secular variation (shown in his figures 9 through 11) was part of the CFEPP's "research conducted since 1976" and this "was not communicated to anyone" earlier because "they would have needed an understanding of the mechanism of formation of the histogram expressed by our model" (p. 289). Eventually in 1997, while publishing this research, Dommanget apparently credits his readership with somehow understanding their difficult mechanism. But why does this author still exempt from publication the results of those counterexperiments of 1968-1970? We were told that they had not been published because, in a similar vein, the CP deemed the public to be too uninformed to understand them. But now, although crediting the present readership with being capable of understanding CP's idea of secular variation, Dommanget withholds the results of CP's counterexperiments, which in 1976 became model inconsistent. Can anyone be blamed for gaining the impression that CP swept its model-inconsistent results of 1970 under the carpet, concerned over a possible detection? Dommanget should be prepared to dissolve the above contradictions and make all pertinent facts transparent. This would be in keeping with his own Bulky Mars Effect Hard to Hide 445 philosophy of science, because he contends that "the committee is only interested in seeing science progressing on a firm, stable, and rocky basis, whoever could be right" (Dommanget, 1982b, p. 66). 5. Conclusion Dommanget (1982a), in published and unpublished Mars effect disputes, often challenged researchers either to accept the Belgian committee's model or to "indicate without possible doubt.. . on which precise point this analysis could appear erroneous" (p. 74). Our scrutiny of CP's work has revealed quite a few points of error. The conclusions to be drawn from those flaws have been anticipated by previous authors, to whom due credit is given in what follows: 1975: R. Chauvin (professor of biology at Sorbonne, in a letter to J. R. Dath, April 2 1, 1975, then chairman of CP) said, "I conclude: You did observe the same significant results as M. Gauquelin.. . You did not calculate the theoretical expectancies.. . Your counterexperiment apparently removes the two main objections that you raised to Gauquelin's method." In 1975, L. de Marrk (just-resigned member of CP, in a letter to Kurtz, November 8, 1975) said, "It looks as if the committee's majority, conducted by its president, were seeking, under pseudoscientific pretexts, to hide a fact that they themselves have verified.. .: the relation existing between Mars and the champions." In 1982, G. Abell of CSICOP, U.S. skeptics, astronomer, in a letter to M. Truzzi (July 21, 1982) said, "I admit that I cannot understand CP's [negative] points about [Gauquelin's] expected distribution in light of the [positive] results of the Zelen test [data comparable to Gauquelin's ordinary people data]. I think they [CP's objections] are just wrong." In 1982, M. Gauquelin wrote, "Why did Dommanget never publish his own expected.. . [birth] frequencies? Many people asked him, too. [His]. .. answer: 'Such calculations...were never done.. .it is impossible.' The so-called mathematical model.. . was created to hide the Mars effect and to mislead the general reader" (p.68). l 7 1998: J. W. Nienhuys (Dutch skepter): Nienhuys considers possible variation of theoretical expectancies, as might occur by Dommanget's secular factors, as minimal ("unimportant") (here visualized in Figure lb) in view of l7 For me, it is surprising that Gauquelin did not look deeper into CP's misconceived model. His position would have been considerably stronger had he demonstrated its absurdity in detail. Abell made the first unraveling steps, which Gauquelin missed to continue. Dommanget's note to Truzzi on September 20, 1982, was apparently true: "We know that M. Gauquelin never understood the real significance of what we call thepCk9'Understandingpck means piercing an absurdity veiled by some stupefying formalism. 446 Suitbert Ertel those large ups and downs of observed key sector frequencies, as were manifest in the data of the Belgian committee (see Figure la). Nienhuys' own reservation about the Mars effect has different grounds (Wunder, 1998, p. 82). 1998: E. Wunder (German skeptics GWUP): "An answer to the question whether the Mars effect does or does not exist does not necessarily presuppose the solving of the theoretical problem brought forward by Proffessor Dommanget [S.E.'s translation]" (Wunder, 1998, p. 82). The CP's approach has in fact never been endorsed by anyone, not even by skeptical committees outside Belgium. The present review aimed to show in hindsight, and in detail, why the above-quoted reactions were justified. Nevertheless, CP's Mars effect research, as well as efforts by other skeptical committees on this issue, should be acknowledged. Their continuous resistance to the Mars effect claim has challenged rigorous tests by Gauquelin, Miiller, and Ertel (see Ertel, 1998, 1999; Ertel and Irving, this issue). The planetary anomaly, if it were due to mere imagination, would certainly have disappeared under such scrutiny, but it did not disappear. For readers entering this arena for the first time, the dynamics of this controversy might be helpful to understand the challenge of this phenomenon, which deserves extensive research in the future. References Abell, G. O., Kurtz, P., & Zelen, M. (1982). The Abell-Kurtz-Zelen 'Mars effect' experiments: A response. Skeptical Inquirer, 7(3), 77-82. Committee PARA. (1976). Considkrations critiques sur une recherche faite par M. M. Gauquelin dans le domaine des influence planktaires. Nouvelles Brkves, 43,327. Dommanget, J. (1982a). Comments on Patrick Curry's research on the Mars effect. Zetetic Scholar, 9,73-74. Dommanget, J. (1982b). On the Mars effect. A last answer to M. Gauquelin. Zetetic Scholar, 10, 66. Dommanget, J. (1995). Personal communication. Dommanget, J. (1997). The "Mars effect" as seen by the committee PARA. Journal of Scientific Exploration, 11(3), 275-295. Ertel, S. (1992). Mars effect survives critique of Dutch skeptics. In Nienhuys, J. W. (Ed.), Science or Pseudo? The Mars effect and other claims: Proceedings of the third EuroSkeptics congress (pp. 185-203). Utrecht: Skepsis. Ertel, S. (1998199). Is there no Mars effect? Correlation, 17, 4-23. Ertel, S., & Irving, K. (2000). The Mars effect is genuine. On Kurtz, Nienhuys & Sandhu's missing the evidence. Journal of Scientific Exploration, 14, 421-430. Gauquelin, M. (1982). A reply to the statement of committee PARA. Zetetic Scholar, 10, 75-77. Gauquelin, M. (1988a). Written in the stars. Wellingborough, U K : Aquarian Press. Gauquelin, M. (1988b). Is there a Mars effect? Journal of Scientific Exploration, 2(1), 29-52. Gauquelin, M.F., & Gauquelin, M. (1957). Mkthodes pour ktudier la &partition des astres dans le mouvement diurne LERRCP, Paris. Rawlins, D. (1 978). Memorandum on the relation of Mars' solar proximity to Michel Gauquelin's Mars-sports results and claims. Phenomena, 2.3-2.4, 22. Wunder, E. (1998). Reply to Prof. Dommanget's letter to the editor. Skeptiker, 11(2), 82. Journal of Scientific Exploration, Vol. 14, No. 3, pp. 447454,2000 0892-33 10/00 0 2000 Society for Scientific Exploration ESSAY What Has Science Come to? Max-Planck-Institut fur Astrophysik 0-85740 Garching bei Miinchen, Germany Abstract- Some criticisms of the fundamental processes in modern science are made. They are illustrated by references to examples over a range of different fields. Keywords: science - religion -enlightenment Introduction Fifty years ago, one could hardly avoid falling into a heated argument on whether science and religion were compatible. Today, that is a dead subject. As a scientist, I just assumed people had come to realize that science is what works and religion was based on myths and guesses. But now an astonishing realization has begun to dawn on me-religion has prevailed! Science has become religion! Let me quickly add that this is not the view of the influential establishments of science and religion. The two, as Stephen Jay Gould (1 999) proclaims in his new book Rocks of Ages, represent "a respectful noninterference-accompanied by intense dialogue between two distinct subjects." Another recent book about religion, however, gives the game away in its title: Seduced by Science (Goldberg, 1999). The point this latter book misses, however, is that although religion may have borrowed some of the jargon of science, science, more importantly, has adopted the methods of religion. This is the worst of both worlds. Rather than going on at length about how both approaches to enlightenment are correct, it would be more useful to explore why both are so incorrect! Of course, there is a questioning, exploring side to both science and religion, which in the beginning was vital to humanity, but what most people accept today as fundamental scientific knowledge is barely distinguishable from what organized religion became some centuries ago. The fatal part of the latter was dogma unsupported by replicable experiment. The most damaging aspect of science today is widely promulgated theories that are contradicted by observation and experiment. In both cases, a story is mandated by authority and then defended by educational, economic, and sociopolitical agencies. 448 Halton Arp Some Examples of False Premises Of course science claims to be based on facts and that contending theories must rise and fall as new evidence emerges, but the changes tend to be extravagantly hyped variations based on the same unquestioned basic assumptions. Let us examine for a moment the current all-encompassing science of cosmology, i.e., the physics of the universe. The big bang theory proclaims that the whole universe created itself instantly out of nothing. I believe there are many observations by now that disprove this, but even supposing for a moment it were true, would it be essentially different from the religious belief that God created the universe at some time in the past? Does calling the event "faster than light inflation" explain any more than calling it a miracle? In fact, scientists seem to have borrowed heavily from the religious concept of "immaculate conception." The Vatican has supported the big bang theory since they alertly sense a place for a "creator." And when Stephen Hawking solves the riddle of "How did the laws of physics by which the universe was formed exist before it was formed?" by saying, "They existed in imaginary time," he might just as well have said, "Any information content in that statement is purely imaginary-I don't know any more than the church does!" Creationism and Science Education? One of the crusades of academic science is against religious creationism. Periodically there arises a messianic need to save the general public from the ignorant belief that humans were created in their present form some short time ago, say, 8,000 years or so. They should blush with shame. Their big bang cosmology, aside from a small quibble about timescales, is the most blatant form of creationism. The claim is that not just humans but the whole universe was created instantaneously out of nothing. Many scientists are outraged that the Kansas Board of Education has banned the big bang theory, but they overlook the point that it was brought on by their own efforts to ban religious creationism; as in most religious wars, they tried to ban the heretic beliefs. As for Darwinian evolution, they did not see that it was not a valid theory until it confronted openly the opposing claims. The essence of true education has been mutilated by all three participants in this sorry spectacle. Academic science fails for trying to ban religious creationism, religious creationism for trying to ban evolution theory, and big bang creationism and the board of education for trying to ban discussion of the whole subject. Only if evidence and arguments on all sides are discussed can students make up their own mind what is the most likely truth-probably something quite different from that of any of the current partisans. More Sacred Oxymorons Another example of a basic premise that is self contradictory is "dark matter." Because extragalactic astronomy interprets all redshifts as velocities, it What Has Science Come to? 449 has to invent huge quantities of invisible matter to explain all the supposed motions that are observed. Never mind that some scientists have shown the postulated dark matter cannot account for rotations, the only redshifts that are certainly velocity. Never mind that the remaining large redshifts have been argued to be nonvelocity, and never mind that the supposed dark matter has never been detected. Nevertheless, enormous research projects are funded on the assumption that more than 90% of the universe is unobservable. This violates the basic definition of science, namely that it deals with the relation of real observations to each other. Most citizens stopped describing angels some centuries ago. But consider the most fundamental precept that underlies modern science: gravity. Let us ask a few simple-minded questions about this force: It is supposed to be a force that attracts one body to another, but how does the sun pull on the earth and vice versa? Are there invisible elastic bands pulling them together? Does the exchange of electromagnetic particles cause a force pulling them together rather than giving them an impulse apart, as one might expect? Obviously, gravity acts much faster than the speed of light; otherwise, the earth would be orbiting around a point where the sun was 8 minutes ago. But in the simplified pantheon of scientific saints, Einstein said that information cannot be communicated faster than the speed of light. So he resorted to having masses "curve space." Bodies ran on invisible, prefixed tracks in space, but how can you curve nothing? Over 100 years ago, a physicist named Le Sage pointed out that a universal sea of faster-than-light particles (or wave particles, for generality), by pushing on all bodies, would produce essentially the same equations as that of "attractive" Newtonian gravity. Gone are the complexities of multidimensional space-time that renders general Relativity comprehensible to only a chosen few. Did someone mention observational tests? Over the last third of a century, it has turned out that extragalactic objects are not necessarily rushing away from each other in an expanding universe. The observations indicate that new galaxies are being continually created. If they are created from low-mass particles and evolve into normal-mass galaxies, then their early redshifts are not indicative of high-recession velocity but instead indicate that matter composed of low-mass, young particles emits weak, redshifted photons (Arp, 1999). In 1977 the Indian astrophysicist, Jayant Narlikar, showed there was a more general solution to the field equations of general relativity. This yields intrinsic redshifts directly as a function of the age of the object-in agreement with empirical observations. Mathematically, it is a transform of the usual special solution that describes our small sample of space and time. But physically, in cosmological realms, it is nonexpanding, continually creating, and indefinitely large-totally opposite the current big bang paradigm. The solution is very simple and requires no space geometry (Riemannian) terms. Those very complicated terms are no longer needed to fit the observa- 450 Halton Arp tions. General relativity turns out to be a local theory. The simpler underlying theory is driven by communication of all parts of the universe with all other parts (per the conclusions of Ernst Mach, another 19th-century physicist and philosopher). There are no "singularities" where physics "breaks down," as plagues the present theory. The fundamental property of matter is its age (how much of the universe it has communicated with). That determines the rate at which its clocks run (another way of saying what its redshift is). Space is obviously filled with wave particles (ether, if you will), but it is not sensible to give them coherent geometric properties. Certainty in Science The alternative interpretations I have sketched above may seem outrageous to anyone schooled in currently accepted physics. Moreover, because my major criticism of today's science is that it is impossibly authoritarian, I cannot claim that these new ways of understanding the observations are, with any certainty, correct. The major reason for my advancing them is to demonstrate that there are possible ways of simply and rigorously connecting the data that are enormously different from the currently obligatory theory. The usual establishment excuse that "there is no other possible way of explaining the observations" simply cannot be used to prop up a theory that has been devastated by the empirical evidence. The only thing we can be certain of is that the old theory has been disproved. New working hypotheses such as those I have outlined above can be tried, modified, and perhaps, inevitably, discarded completely. Theory is only an attempt to simplify the connection between currently known facts. The prime responsibility of science is to keep in mind that there is never certainty and the most important obligation is to keep testing the fundamental assumptions. Suppression of Evidence The most harmful aspect of what science has become is the deliberate attempt to hide evidence that contradicts the current paradigm. Most scientists give ritual obeisance to the dictum that "one can never prove a theory, only disprove it." In a quite human fashion, however, they act in an exactly opposite manner-judging that "if an observation disagrees with what we know to be correct, then it must be wrong." The tradition of "peer review" of articles published in professional journals has degenerated into almost total censorship. Originally, a reviewer could help an author improve his article by pointing out errors in calculation, references, clarity, etc., but scientists, in their fervid attachment to their own theories, have now mostly used their selection as a referee to reject publication of any result that would be unfavorable to their own personal commitment. The intensity of the feelings involved can be judged by the frequent recourse to personal invective in the reports to the editor (which the editors, joining in the What Has Science Come to? 451 , spirit, pass on to the authors). The only comparable interaction I have heard of is the passionate wars between different religious doctrines of past centuries. The press, of course, only reports news from established academic centers that have a strong financial and prestige interest in glorifying the status quo. The result is that real investigative science is mostly now an underground activity. Independent, often self-supported researchers are publishing in privately supported, small-circulation journals. It is difficult to say whether "big science," like the medieval church, will slowly erode in influence over many generations, or whether there will be a sudden rebellion with scandal and corruption being reported by investigative journalists. The Humanities Fail to Counterattack One characteristic of an institution that has long gone unchallenged is arrogance. The physical sciences manifest this quality particularly toward the social sciences or the humanities, scathingly referred to as the "soft sciences." A few years ago, the now famous "Sokal hoax" (Sokal, 1996; Sokal & Bricmont, 1998) was initiated. An article intended to be nonsense was cast in pseudosocial-science jargon and accepted in a humanities journal. After the hoax was exposed, it was heralded as proof of the unrigorous character of the social sciences. It is undeniable that there is a pervasive use in all of academia of complex, specialized terms, which when examined do not yield much significance. It should be challenged wherever possible. For example, Serge Lang, a mathematician and member, has in the past challenged puffery by social scientist members of the National Academy of Sciences, but conversely the nonphysical sciences should not pass by the opportunity to criticize a much greater breach of rigor in the "hard sciences." After all, to get the whole universe totally wrong in the face of clear evidence for over 75 years merits monumental embarrassment and should induce a modicum of humility. It is not enough for deconstructionists to complain that our culture is dominated by dead white men. The important point is that the dominators (at least the ones generally revered in the hard sciences) got it completely demonstrably wrong. Consider at this moment, e.g., the pinnacle of modern physics: String Theory. From an article in the Los Angeles Times of November 16, 1999, we can select a few dazzling quotes. Famous physicists from the Institute of Advanced Study in Princeton: "Space and time may be doomed" and "time can be extinguished like a blownout flame." From Columbia University: "Strings are shards of space and time." From the Institute for Theoretical Physics: "Where are we? When are we?" And finally: "Today's physicists are in possession of what may well be the Holy Grail of modern science." If I stand back a little from this, it sounds like a religious frenzy-like speaking in tongues. Where is the hard rebuttal from the postmodernists, and I 452 Halton Arp perhaps most important of all, the analytical criticism and discussion from the generalists? Is it not time to move on from partisan authority to open-minded, sensible curiosity? Medical Science and Biology At first sight, the biological sciences would seem to be immune to plunging off the track with invalid paradigms. After all, the experiments are easily repeatable and either they work or they do not work. But the cutting edge of research passes through academia. Some very bad mistakes occur. For example, the problem of how AIDS arose in humans. In the 1950s, polio vaccines were grown in kidney-cell cultures from monkeys. They were administered to massive numbers of people in central Africa around 1955. Some years later AIDS, caused by viruses present in certain types of monkeys, began to decimate African peoples and soon spread around the world. It has taken over 44 years for science to discuss the probable cause of AIDS, as due to using cross-species cellular material as a substrate for live viral vaccines. The evidence began to be clear 10 years ago (Cribb, 1996; Curtis, 1992; Hooper, 1990, 1999) but was not discussed in the preeminent interdisciplinary science journals such as Science and Nature. Toward the end, there were deprecatory references to some "scientifically implausible" theories and once the monkey viruses were inescapably identified, detailed scenarios of the African people eating "bush meat" (monkeys) as an explanation for the (sudden) epidemic (Hahn et al., 2000). It should not require prominent expertise to face the fact that various species carry viruses to which they are immune, but which can be deadly to other species. Is it permissible to simply say, as Hooper quotes one investigator of that time, "We were acting in full innocence, not understanding what sort of Pandora's box we were opening?" Is it permissible to give another human being a substance and say, "Here take this, scientific knowledge has established that it will protect you from debilitating and anguishing disease?" In the course of events, something more seems now to be called for. . Physics in Action One might think that theoretical physics, that realm of what people think reality is, cannot do much practical harm. Wrong. Beyond the invention of racethreatening bombs, there is the trying out on human beings of naive experiments and ill-thought-out products, which if given a moment of common sense reflection, would have revealed the possibility of truly horrible personal consequences. The Plutonium Files by Eileen Welsome (1999) is the latest in a now long list of accounts of how radioactive substances were tested by the government, the military, and scientists on unsuspecting subjects. Welsome's personal accounts of the victims make the incidents poignantly real. What Has Science Come to? 453 Of course, everyone who builds a device to improve the lot of mankind knows it may be used by some to do great damage. But it seems to me that the guardians of fundamental knowledge, the universities and research institutes, could set a much better example of responsible testing of their theories and public announcements. Most graduates of top-ranked research departments have unfortunately been treated to an ongoing spectacle of prominent personalities publicizing their own theories while ignoring or suppressing obvious observational disproofs. Getting the answer that will do the investigator the most good is not necessarily the answer that will do the society the most good. There are well-known departments that are almost completely preoccupied with personal issues of tenure and competition but where the real issue is whether there is any professional competence. Psycho-Science Sometimes it is better to just sleep through a whole era of scientific advance. In November 1999, I encountered a news note that brain researchers were having second thoughts. It seems that for decades, they had been measuring electrical activity in various regions of the brain and had come to the conclusion that the Freudian psychoanalysis was wrong. There was no activity in parts that should have supplied stimuli while dreaming! But now, they had measured again, and in different parts, and announced that they were not so sure of their original conclusion. Imagine, more than a century of comparison and study of carefully recorded subconscious and conscious human states had been overturned by needle twitches. Then psychoanalysis had been reinstated (maybe)-and I had been blissfully unaware of the whole drama. The Way It Began and How It Could Be The great irony is that both science and religion started as legitimate inquiries into the nature of existence. The earliest impulse of beings must have been to observe. Religion noted the inside feelings and dreams. But probably some of the feelings were misunderstood fears, disguised images, and illusions, which then became institutionalized by charismatic personalities. Science, on the other hand, tried to record events objectively. But perhaps similar subconscious assumptions crept in and influenced all subsequent interpretations. Again, as science organized, authority figures became associated with the "laws" they were credited with discovering. Organized religion succeeded in killing a great number of people down through the ages on issues that were labeled "belief and heresy" but were probably more fundamentally concerned with personal profit and power. Science has arisen some centuries later in less bloody societies but has killed and delayed many new ideas and discoveries and has made many mistakes, for perhaps basically the same reasons. 454 Halton Arp What do we do now? Perhaps community-based organizations such as the Unitarian or Quaker churches, which are almost absent theoretical dogma but instead concentrate on useful service to a society of real people, is the most desirable direction. Perhaps it is necessary to discover morality empirically. Of course, today, for science in particular, electronic communication makes possible communities of individuals from all corners of the world. The most direct evolution toward an enlightened science is for these groups to just go about supporting each other in doing science free of disproved, official assumptions. Of course, an informed public is crucial. Already, however, that public is learning that the most dreaded words one can hear in modern life are "There is no credible scientific evidence that the substance in question is harmful to human beings." One just cringes and thinks, "How long before the data is released and the other shoe drops?" Individual survival based on free communication and individual decision making seems the slow but surer method for achieving both spiritual and scientific enlightenment. References Arp, H. (1999). Seeing red: Redshifts cosmology and academic science. Montreal, CN: Apeiron. Cribb, J. (1996). The white death. Australia: Angus & Robertson. Curtis, T. (1992, March 19). The origin ofAIDS. Rolling Stone, p. 54. Goldberg, S. (1999). Seduced by science: How American religion has lost its way. New York: New York University Press. Gould, S. J. (1999). Rocks of ages: Science and religion in the fullness of life. New York: Library of Contemporary Thought (Ballantine). Hahn, B., Shaw, G., De Cock, K., & Sharp, P. (2000). AIDS as a zoonosis: scientijc andpublic health implications. Science, 287,607. Hooper, E. (1990). A reporter's own story ofAIDS in East Africa. London: Bodley Head. Hooper, E. (1999). The river: A journey to the source of HIVand AIDS. New York: Little Brown. Sokal, A. (1996). Transgressing the boundaries: Toward a transformative hermeneutics of quantum gravity. Social Text, 46/47(14), 217-250. Sokal, A., & Bricmont, J. (1998). Intellectual impostures: Postmodern philosophers' abuse of science. London: Profile. Welsome, E. (1999). Theplutoniumjles. Dial, UK: Delacorte Press (USA). Journal of Scientific Exploration, Vol. 14, No. 3, pp. 455-459,2000 0892-33 10/00 0 2000 Society for Scientific Exploration LETTERS THE EDITOR Is the Activity of the Cerebral Cortex a "Brake" on the Functioning of the Brainstem? During the 1980s, experimental research in China indicated an increased excitability of the brainstem while the activity of the cerebral cortex was inhibited. Could the activity of the cerebral cortex serve as a "brake" on the functioning of the brainstem? Since the late 1970s, scientific researches on QiGong and extraordinary human capabilities have sprung up in China. Since 1982 at Beijing Traditional Chinese Medicine University, Professor Liu Guo-long has studied the central nervous system of long-term QiGong practitioners who were in a trance state. Here, I would like to summarize the results, which were published in Chinese (Cui and Liu, 1989; Liu et al., 1988). The cerebral cortex consists of five lobes: the temporal, occipital, parietal, frontal, and limbic. Changes of evoked potentials (EPs) of the brain waves were measured in several of these lobes as subjects entered a trance state. In the auditory cortex, in the temporal lobe, EPs decreased by an average of 72% ( P < .01) for 11 subjects entering trance in the process of moderate latency (a state of normal brain activity, 15-50 milliseconds). For 11 other subjects entering trance in the process of long latency (a state of stimulation, 50-300 milliseconds), the decrease was 43% (P < .01). Thus, a trance state inhibits the activity in the temporal lobe. The occipital lobe, the location of the visual center, was stimulated in 26 subjects by flashes and pictures. The potentials evoked by flashes and pictures decreased in the trance state by 32%-67% (P < .05) and 38%-53% (P < .05), respectively. Thus, a trance state inhibits the activity of the occipital lobe cortex as well. The parietal lobe, the location of the body's sensory center, was stimulated by somatosensory signals in 2 1 subjects. The EPs after entering into a trance state decreased by an average of 29% ( P < .05) of subjects. The evoked potentials after entering into a trance state decreased by an average of 29% (P < .05) in the sensory center and by 53% ( P < .01) in the round areas of the sensory center. Thus, a trance state also inhibits the activities of the parietal lobe and its round cortex. The frontal lobe, the locus of the high-level activities of consciousness, was studied in 14 subjects. The alpha rhythm of the brain in its normal state is chiefly found in the occipital lobe. However, after entering into a trance state, the brain showed a striking synchronization of alpha rhythm, particularly in the frontal lobes, and an increase in magnitude of the brain waves. The latter 456 Letters to the Editor also increased in the occipital lobe but not as markedly as in the frontal lobe. By contrast, normal consciousness and thinking were mainly inhibited after entering a trance state. No results were given for the limbic lobe, which regulates visceral activities. It lies inside the cerebral hemispheres and there are technical difficulties in directly studying its functioning. Changes in the brainstem were surveyed in 21 subjects entering into a trance state: The potentials of the diencephalon increased by an average of 42% (no P value provided in the original); of the midbrain by an average of 92% ( P < .001); of the pons by 140% ( P < .001); and of the medulla by 70% (P < .OI). The potentials of the nearby areas of the cerebral cortex decreased by an average of 9.5%. Thus, the trance state appears to increase excitation in the brainstem and to decrease activity in the cerebral cortex. To rule out the influence of psychological factors, the researchers surveyed 12 cats. They found that the brainstems of the cats showed the same facilitation (excitability increase) when the cats received the Qi sent by QiGong practitioners. The researchers suggested that a trance state decreases the control of brainstem function exercised by the cerebral cortex, leading to an increase in excitability of the brainstem. It is interesting that similar suggestions have been made in the Western literature about psychic abilities, namely that the cerebral cortex has an inhibitory effect on psi abilities and the brainstem has something to do with psi abilities. Bergson (19 14) proposed the "filter theory": If the brain developed biologically for reasons of survival, as an organ whose main function is to receive information, perhaps it also serves to screen out certain information in the service ofbiological survival. Thus, it would act as an organ of limitation, a filter. Bergson proposed that the cortex in particular plays this role. Eysenck (1967) suggested that a high state of arousal in the cortex mediated by the ascending reticular formation (the afferent or input side of the reticular formation) is antagonistic to psi. (The reticular formation lies in the brainstem. One of its functions is to mediate changes in degree of wakefulness.) Pribram (1973) and Luria (1973) have noted that the frontal and temporal cortices play an important part in the selective filtering of "afference" or input from outside. It is well known that cocaine and morphine have a marked inhibitory effect on the cerebral cortex but not on the brainstem, the "primary" nerve center. At the same time, drug takers often claim to experience something like religious ecstasy. Some psychotics, too, whose cerebral cortex is flawed to varying degrees, have sometimes been reported to experience psychic phenomena. Ganzfeld studies, in which sensory stimuli are blocked to enhance psychic capability, may also be pertinent. Myers (1903) introduced the notion of "subliminal self" in contrast to our normal waking consciousness, the "conscious self." He thought that the "sub- Letters to the Editor 457 liminal self" takes control in the case of automatic writing, trance, or other automatisms and can "reassert itself in its plentitude after the liberating change of death," although it seems to be blocked by conscious expression. Myers did not identify a part of the brain with the subliminal self; perhaps it is associated with the brainstem. Dixon (1971) and Beloff (1974) have explicitly suggested a close relationship between psi phenomena and subliminal perception on the one hand and reticular formation and the brainstem on the other. Ehrenwald (1977) also considered this possibility: "Certain elementary forms of ESP tend to bypass conscious awareness. They are not tied to processing in higher cortical centers and point to the subcortical, midbrain, or limbic region as their substrate." I believe that two conclusions can be inferred from the experimental research in China, as well as the research results reported in the Western literature. First, in normal waking consciousness, the functions of the brainstem are repressed by the activity of the cerebral cortex that deals with daily feeling and consciousness. The activity of the cerebral cortex is the "brake" on the functions of the brainstem. The brainstem cannot function unless the activity of the cerebral cortex is inhibited. Second, because the trance state often has a most important bearing on psi, the facilitation (excitability increase) of the brainstem in a trance state means that the brainstem's function, if it functions, should have direct relation to psi abilities, psi phenomena, and religious mystic experiences. Further work along the lines of Liu et al. (1988) and Cui and Liu (1989) might well be fruitful. Jesse Hong Xiong Department of Philosophy University of Toronto, 361 Euclid Avenue, Toronto, ON, Canada M6J2Kl e-mail: hongxiong @ yahoo.com References Beloff, J. (1974). ESP: The search for a physiological index. Journal of the Society for Psychical Research, 47, 403-420. Bergson, H. (1914). Presidential address to the Society for Psychical Research. Proceedings of the Society for Psychical Research, 27, 157-175 Cui, R., & Liu, G. (1989). Modern research on the mechanism of regulating mind by QiGong. Chinese QiGong, 1, 4-5. Dixon, N. F. (1971). Subliminalperceptions: The nature of a controversy. New York: McGraw-Hill. Ehrenwald, J. (1977). Psi phenomena and brain research. Handbook of parapsychology. New York: Van Nostrand Reinhold. Eysenck, H. J. (1967). Personality and extrasensory perception. Journal of the Society for Psychical Research, 44, 55-71. Liu, G., Cui, R., Niu, X., & Peng, X. (1988). Experimental research on the state of QiGong and Qi's effect upon nerve function. Chinese QiGong, 6, 3-5. Luria, A. R. (1973). The working brain. New York: Basic Books. Myers, F. H. W. (1903). Humanpersonality and its survival of death. London: Longmans. Pribram, K. (1973). Psychology of the frontal lobes. New York: Academic Press. 458 Letters to the Editor Understanding the Nature of Racial Prejudice (Re: "Toward an Understanding of the Nature of Racial Prejudice" by f Hoyle and Wickramasinghe, Journal o Scientific Exploration, 13(4)) I eagerly looked forward to reading this article not only because I am a person of color but also because I agree with the authors' view that such an "understanding could eradicate or ameliorate its worse manifestations." I must confess that after reviewing the basic geology, anthropology, and Darwinian biological material and the relationship of sunlight to Vitamin D production, as well as malignant melanoma known to every health food naturopathic aficionado, I found myself at the end of the brief article wondering if I had missed something. I reread it and discovered that I had not. "Race matters," as Derrick Bell elaborates in his book, Faces at the Bottom of the Well (New York: Basic Books, 1992), but even Professor Bell and countless others who have written about the "nature of racial prejudice" fail to fully explain the "genesis" of the problem. To the authors' credit, however, they acknowledge that racism is not entirely related to skin color alone. Aside from this small nugget of insight, I found their effort curiously unrewarding. Respectfully, Richard S. Hahn, M. D., RA. C.S. Alliance for Health, Inc. 428 Golden Gate Avenue Belvedere, CA 94920 Dear Dr. Hahn: The editors have changed at the Journal of Scientzfic Exploration. I , too, found the essay by Hoyle and Wickramasinghe disappointing-naive and simplistic. The points you make are well taken. However, they do little to help our readers toward a more sophisticated understanding.The Journal does publish extended commentary on previously published pieces, and we would happily entertain such a commentary, of up to several pages, from you on this issue. Please let us know whether you will do that or whether you would like your comments published in their present form. Yours sincerely, Henry H. Bauer Editor-in-Chief Dear Dr. Bauer: I have been involved with the attempts to accord Dr. Wen Ho Lee the same congressional courtesy and resolution that Mr. Deutsch, former chief of the Letters to the Editor 459 CIA, was given for exactly the same accusation that keeps Wen Ho Lee shackled in prison isolation while Deutsch enjoys the freedom of his retirement. I believe the issues of race, diversity, multiculturalism-i.e., basically "fairness"-transcend all specialized scientific interests. Having reread your letter of April 20, 2000, I would like to accept your offer to "publish comments in their present form." I would not object to publishing our correspondence including this letter. I do appreciate your offer to publish a more extended commentary, but I think it prudent not "to keep beating a dead horse." Cordially, Richard S. Hahn, M. D. FA. C.S. Emeritus Member, SSE Journal of Scientific Exploration, Vol. 14, No. 3, pp. 461-493,2000 0892-33 10/00 O 2000 Society for Scientific Exploration BOOK REVIEWS With the Tongues of Men and Angels: A Study of Channeling by Arthur Hastings. San Francisco, London, Montreal, et al.: Holt, Rinehart and Winston, 1991. 227 pp. Illustrations and photographs. References. Index. Appendix. Sometimes it is important to go back and reexamine a book that has been overlooked. This is a review of such a book: With the Tongues of Men and Angels. Arthur Hastings, its author, sets himself two tasks and does very well at addressing them: What is the nature of channeling? What is the significance of channeling for us? These are questions worth attention, not least because the phenomena of channeling has been around for thousands of years and defies all attempts to explain it away. It is so compelling that even the exposure of occasional frauds has not slowed the public's interest. Perhaps this is because, as Hastings states, fully 15% of the population-in the United States today that would be approximately 41 million people-report hearing voices and receiving guidance. I do not know where Hastings got that astounding-at least to me-figure, but it conjures an arresting reality. Hastings approaches his subject with an open mind-a surprisingly rare attitude, as anyone who takes the trouble to look at the literature of channeling quickly discovers. Whether it is called channeling, mediumship, automatic writing, guidance, prophecy, or any one of several other names, as a rule it excites more passion than insight. Hastings also writes with a light touch, which, for anyone doing research in this area, will be much appreciated. With the notable exception of Jon Klimo's recent and excellent Channeling, much of the modem literature on this subject is famously insubstantial, and the more serious studies, dating mostly from the late 19th and early 20th centuries, are so turgid and self-important, whether espousing or denigrating the subject, that reading them becomes an act of penance. Hastings begins his exploration where it should begin, with a discussion of just what we mean by channeling, and he makes the important point that this is a big umbrella that covers many different beasts. He settles on this as his definition of what channeling is. Channeling refers to a process in which a person transmits information or artistic expression that he or she receives mentally or physically, and which appears to come from a personality source outside the conscious mind. The message is directed toward an audience and is purposeful (p. 4). ' Klimo, Jon. (1998). Channeling: Investigations on receiving information frornparanomzal sources. Berkeley, CA: North Atlantic Books. 462 Book Reviews This is a more subtle definition than it may appear at first reading. Was Edgar Cayce a channel? With a few exceptions, he did not channel discarnates or dead people, claiming that his source was what he called the superconscious That is hardly the same thing as Frederick Bligh Bond, who reconstructed Glastonbury Abbey following guidance he received in automatic writing sessions from sources who claimed to be the monks who once inhabited the ruined ~ b b e But. both Cayce and Bond share one thing: They provided ~ ~ testable information that could be independently verified. This is quite different from the Ramtha persona channeled by J. Z. Knight. Like Bond's monks, this was presented as a discarnate source, but the Ramtha material is very short on testable information, running instead to long philosophical discourses. The key words in Hastings definition are "outside the conscious mind." In that sense, yes, Cayce was a channeler. Under this definition, the same could be said of Rudolf Steiner, but what about Wolfgang Amadeus ~ o z a r t~, r~a h m s , ~ and ~ o ~ e l a nall ,of whom reported hearing the music that they wrote down. d ~ That would take channeling into the waters of creativity and intuition, and Hastings admits as much. He decides to focus his book on channeling with a defined source. This helps the structure of his slender book, but I could not help wishing he had taken a broader rubric, since while critics may deride channeled "masters" such as Ramtha, they can hardly avoid the mastery of a Mozart sonata or a Housman poem, and there is clearly some kind of connection here. Having laid his groundwork, Hastings plunges into actual cases, some, such as Cayce, well known to millions and others, such as Pearl Curran, a St. Louis housewife who channeled the poetry, novels, and plays of Patience Worth-reputedly an English woman of the 17th century-are now virtually unknown except to scholars such as Stephen Braude, who has written extensively about this case. Phenomenon such as the CurranIWorth material, once widely popular and ranked equal to the work of Amy Lowell and Edgar Lee Masters, illustrate yet again the connection between creativity and channeling. Even more provocative is the mathematical genius Srinivasa Ramanujan, a largely self-educated East Indian whose work is one of the great mathematical accomplishments of the age, and who said some of his theorems came from Indian goddesses-a fact mostly ignored in the biographical canon about this remarkable man. 7 Edgar Cayce Reading 254-67. Bond, Frederick Bligh. (1918, 1921). The gates of remembrance: The story of the psychological experiment which resulted in the discovery of the Edgar Chapel at Glastonbury (2nd and 4th eds.) Revised. Oxford: Basil Blackwell. Delaney, Gail. (1984).Creativity in music. Proceedings of the Association for the Anthropological Study of Consciousness. Annual Meetings. Abell, A. M. (1964).Talks with the great composers (pp. 19-21). Garmisch-Parten-Kirchen: G . E. Schroeder-Verlag . Delaney. Loc Cit. Book Reviews 463 Hastings goes on to cite a long list of other musicians and artists who channeled (albeit not the type of channeling that meets his definition), and once again, I found myself wishing Hastings had addressed the issue of creativity in the context of channeling to a greater degree. Although, given his definition, this "open channeling," as he styles it, was not the focus of his book. Hastings does, however, make some very important points, both about channeling as he defines it and about open channeling. A study of this subject reveals first the "variety of skills demonstrated"; second, that the creation of this material is "effortless, immediate, and spontaneous"; third, that the skills demonstrated are not only above what the channeler can attain in the normal waking state, but often rise to the level of real genius; fourth, that the material often has a profoundly spiritual, if usually nondenominational character proposing the interconnectedness of all life; and fifth, that channelers are not psychotic, abnormal, or dysfunctional-or at least no more so than the general run of the population. This case history information, which makes up the bulk of the book, is fascinating, and most of it is little known. But more important than the case history material is the framework in which Hastings places it. He does not shy away from calling a fraud a fraud, and he gives the reader a clear path for evaluating what is going on. He also includes considerable material from the formal psi literature, which goes almost entirely unremarked in most books on channeling. My only complaint here is that I wish he had gone into greater detail on the relationship between remote viewing and channeling. Both, after all, are examples of anomalous cognition. I particularly appreciated the section on the relationship between channeling and religious prophets. This is very touchy stuff that most scientists avoid altogether, and some religious find offensive: If it is God talking to you, how dare you call it channeling. Yet, at their core, most religions are channeled, and if only because of the slaughters and mass suicides that are such prominent features in the landscape of religious channeling, this subject deserves the closest attention. Hastings' observations make a real contribution. Similarly, his section on various theories and models of channeling provide what is probably as comprehensive and cogent a take on this aspect as will be found anywhere. At the end of the day, Hastings achieves what many aspire to achieve and few attain in the field of the paranormal: a sensible well-grounded but accessible examination of a highly controversial subject. Stephan A. Schwartz Shepherdstown, WV Reaching for Reality-Seven Incredible True Stories of Alien Abduction by Constance Clear. San Antonio, TX: Consciousness Now, 1999. x + 230 pp. $16.95, paperback. P.O. Box 15994, San Antonio, TX 78212. 464 Book Reviews Constance Clear is a psychotherapist (no doctorate listed) who specializes in treating people who believe that they have been abducted by aliens from UFOs. (Any unidentified aerial phenomenon is, technically, a UFO, but references here will be to extraterrestrial spacecraft.) In this book, seven such cases are presented, partly in the words of the clients. There are numerous sketches by the clients and a few photographs of sores and the like. In approaching treatments of this topic, we must consider some background factors. Since Kenneth Arnold made his famed sighting in 1947, there has been no physical evidence such as part of a spacecraft. There have been no unambiguous photographs, no careful measurements of trajectories in flight, to say nothing of one of the aliens finally landing in the public square and demanding, "Take me to your leader." Further, to accept what is characteristically alleged in such cases, we must overlook laws of physics that are fairly well established. There is the time that it would take such craft to get here from outside our solar system, many times the plausible lifetime of any readily imaginable creature. There is the absence of any physical principle reconcilable with the way the UFOs are said to hover. One well-credentialed physicist consulted by the reviewer estimated the chance of the UFOs being real as one in a billion. Because of that one chance, he added, one should not make a categorical denial. Nevertheless, the burden of proof rests with those who assert these extraordinary phenomena. It is a burden that Ms. Clear does not meet and does not, in some ways, try to meet. One feels her strong empathy with her counselees, her painstaking efforts to report her clinical experiences accurately. Still, the author apparently accepts all that her clients tell her as fact, even though it involves their floating through the air, walking through walls, and so on. There are obvious questions that neither she nor they appear to have asked, such as if a spaceship landed in the front yard on dozens of occasions, why didn't the neighbors ever notice? In one instance, a man says that an alien tore a button off his pajamas and cut a swatch of cloth from them. When he awoke in his bed later, the button was in place and no cloth was missing or looked resewn. Conclusion: These aliens are clever! Even the book's title takes a stand for the objective reality of the clients' experiences. The subtitle includes "true" (worthy of belief), albeit immediately preceded by "incredible" (not worthy of belief). Ms. Clear organized her clients into a support group to share their abduction experiences with one another. No procedure seems likelier to encourage any misperceptions they had, nor less likely to encourage critical thinking. Further, she ignores ready-to-hand psychological explanations for the experiences. The author and the clients were no doubt presenting the truth as they saw it, but there are such things as dreams, hallucinations, false memories, fugue states, and dissociated states. Seasoned clinicians are presumably aware of them. Currently, at Fukushima University in Japan, research is being done on sleep paralysis, the result of a disconnect between brain and body as a person is on the verge of sleep. The disorder tends to produce symptoms markedly similar to those described in Clear's book. When witches were in style, vic- Book Reviews 465 tims of sleep paralysis often saw witches. Now they see UFOs, whose crews speak to them in their own language and local dialect or sometimes telepathically-a convenient mode, since the message can be whatever the recipient chooses it to be. Perhaps, in the future, they will see robotic supercomputers. UFOs and aliens who give medical examinations to earth people are possible in the broad sense in which anything that is not a contradiction in terms (e.g., a square circle) is possible. Nevertheless, the odds remain what they are. Perhaps the main contribution of this book to science will be one certainly not intended by its author: to illustrate how mass delusions can become acculturated and widespread and persist in the face of reason and contrary evidence. The witch trials in Salem and America in the McCarthy period come to mind. Robert Nordberg Marquette University 999 East Quarles Place, Fox Point, WI 53217 Project Mindshift: The Re-Education of the American Public Concerning Extraterrestrial Life, 1947 to the Present by Michael Mannion. M. Evans and Company, 1998. 304 pp. $22.95, hardcover. In this fascinating book, Michael Mannion goes far beyond mere historical rehash of the modern era of ufology to explore a new thesis: that the government is engaged in a subtle reeducation project to accustom the American public to the idea that extraterrestrials have contacted us and are among us. Throughout the book, Mannion shows a fine ability to look at human problems-such as UFOs and the possible alien agenda-and reflect them neatly through the lenses our culture employs, such as music, movies, the Internet, and science. He is neither sensational nor skeptical, which is refreshing. And he does not purport to have any final answers. This balance is desperately needed in ufology, in which ill-researched answers and grand conclusions are all too common. The idea of government activity in ufology is not new, of course. The government's (usually nefarious) involvement in shaping public opinion is a staple of conspiratorial ufology. But Mannion does not dwell on conspiracies, though he outlines the case for covert government involvement. He considers some high-ranking body similar to MJ- 12 to be likely. But to his credit, Mannion does not take sides in the acrimonious debate about whether the inner circle of powerful men forming the MJ-12 group actually existed. He only states that some such steering committee is likely, because the government had a demonstrated interest in UFOs in the modern era. The book's focus is more on the mindshift hypothesis. This is the notion that for a truly significant paradigm change to occur, there must be a mindshift in a society's way of looking at the world. The hypothesis Mannion is working 466 Book Reviews with is defined in the book as, "Our world has been and is now being visited by advanced intelligent entities from elsewhere, and this reality has been known to a limited number of people within the U.S. government since at least 1947 ." (pp. 82-83) Mannion cautions that the mindshift hypothesis is an assumption, not a conclusion. Working with it requires suspending deeply held beliefs, setting opinions aside, and delaying judgments until all the facts are in. Drawing on his hypothesis, Mannion speculates that if a small group of government officials knew about extraterrestrial life, and if they feared the public could not cope with this information, then what would they do? They would probably engineer some kind of program to slowly open the public to the idea so that people could gradually come to accept it. Central to the book's hypothesis are Mannion's interviews with the mindshifters. These are the luminaries known to all students of ufology, including Budd Hopkins, Raymond Fowler, John Mack, Stanton Friedman, Michael Lindemann, Edgar Mitchell, Don Berliner, and David Jacobs. Interestingly, most of these men remain skeptical about the government's conducting of a subtle but pervasive education campaign to effect a mindshift in public attitudes toward aliens. Most see either no evidence that the government has any kind of education program or shadowy evidence that the government has spent most of its energy hiding and disguising the truth about UFOs and extraterrestrial life. The only dissenter is Michael Lindemann, who thinks it is possible that the government is both hiding and revealing the truth in order to prepare the public for more direct revelations later. Following these interviews come exciting chapters in which Mannion looks for evidence of mindshift in print, radio, TV, movies, music, and the Internet. In all of these media, Mannion finds evidence of material on UFOs and extraterrestrial life, although it is questionable whether it adds up to a well-defined pattern. Generally, the book takes a chronological viewpoint, tracing the increase and sophistication of alien themes. The chapter on the Internet is the thinnest, mainly because the Internet is so new that it hasn't much chronological history. This chapter becomes almost a listing of Internet resources. On the other hand, the richest chapter in this section is the one on the movies. Mannion speculates that the science fiction genre, which has been full of aliens since the 1950s, may be the tool par excellence for a government reeducation program. He notes that several classic films, such as Invasion of the Body Snatchers, The Thing from Another World, and It Came from Outer Space all foreshadowed later UFO themes. He points to the possible evidence of covert government involvement in the movies by describing what happened with the 1950s film Invaders from Mars. In this movie, a boy watches a saucer burrow into the ground in his backyard. His father, who investigates, returns a changed man, with a black mark on his neck. Soon all the town's adults have the mark. Finally, the boy finds an astronomer who believes him, and the aliens are defeated. Interestingly, the European and American versions of the film were different. In Europe, the film portrayed the boy's story as real. But Book Reviews 467 I the American version depicted the boy's story as if it were all a dream. The original screenwriter was so incensed by this, says Mannion, that he had his name removed from the credits. Mannion buttresses the plausibility of government interference in movies by noting, "Movies and politics have always mixed well because Hollywood and Washington are both about money and power" (p. 176). The book builds a likely, though circumstantial, case for governmental interest and influence. A cornerstone of the book is the chapter Mindshift in Science. Mannion discusses the changes occurring in astronomy as a result of the new focus on finding extraterrestrial life in our own solar system. NASA itself has been excited by preliminary findings of microbial life in a meteor from Mars. Mannion also investigates the possibilities of zero-point energy, surveying the scientists who thought its use possible, from Andrei Sakharov to Nikola Tesla. Zero-point energy is the minimum energy possessed by a substance at absolute zero temperature. It is ubiquitous and fills all the fabric of space. It may be the energy that powers UFOs. Even NASA is interested in its technical applications. But the majority of the science chapter is devoted to the still-controversial theories of Wilhelm Reich, some of whose theories were similar to those of zero-point energy. The basis of Reich's science was the concept of "life energy," or "orgone energy," as he named it. He considered orgone energy to be tangible, measurable, and usable. He wrote descriptions of the energy that tally well with what physics knows today about energies. Reich became interested in UFOs in 1953. He felt that their observed characteristics fit perfectly with the behavior of orgone energy. He took photos of UFO tracks (three of these rare photographs-excellent examples of the scientific method at work-are reproduced in Mannion's book). Mannion feels that Reich's science and discoveries are key to understanding UFOs. He speculates that Reich's fate-he died while serving in a federal penitentiary-may have been the result of governmental pressure: There is evidence indicating that Reich's work was known to some at the highest levels of the U.S. government and taken seriously. His energy discoveries were a challenge to individual character structure, powerful economic forces, the emerging national security state-in other words, to the entire authoritarian social order itself (p. 283). Mannion paints a picture of humanity in a state of crisis now, but also holding the tools that can be used to overcome it. A serious scientific evaluation of Reich's work is essential at this critical juncture. Indeed, says Mannion, "The development of Reich's orgone energy research could bring about the greatest mindshift of all" (p. 285). 1 In his final chapter, Mannion draws the strands together into some novel conclusions. He begins with a summation of the possible alien agendas from all of the mindshifters he interviewed. None of their conclusions are particularly startling. But then Mannion discusses a fundamental question that aliens 468 Book Reviews might ask about us: What is wrong with human beings? Why have all the religious and philosophical attempts, from Plato to Judaism to Christianity, always failed to fundamentally restructure human life toward love and gentleness? Mannion concludes that aliens might consider humanity's fundamental flaw to be evident in the way we treat our newborn. Humans slap the newborn, cut the umbilical cord, wrap infants in swaddling clothes, feed them from bottles, circumcise them, and do many other harmful, if not cruel, things to the newborn. Mannion develops the idea that the infant, reaching out to the world, is greeted with harshness-and often abuse and neglect-and then withdraws. This withdrawal may be at the root of our deepseated social problems. He points out that humans interfere with the sexuality of the young in many ways. How is all of this related to UFOs? Through energy. Mannion reminds readers that Reich (as well as ancient philosophers) understood there to be one primordial, cosmic energy force that expresses itself through all living and nonliving things. It may be this primordial energy that aliens are studying in people. Mannion notes that many abductees describe sexual experiments by the aliens and wonders if perhaps the aliens are drawn to this because they know of human sexual dysfunction and its consequences. He says, "There is ample evidence to link sexual pathology with a wide range of destructive and self-destructive behavior. Perhaps ET interest in human sexuality has more to do with this biological tragedy than we know" (pp. 295-296). The book ends on a cautious note. Mannion neither puts forward answers nor charts a program for changing things to complete the mindshift. He simply points out that aliens may have come here for reasons that have nothing to do with humanity, much as the Europeans came to the Americas for reasons having nothing to do with the indigenous peoples. Aliens may not be (as Europeans were not) superior or godly, despite advanced technology. They may just be other life-forms, exploring creation in their own ways. Mannion's final bit of advice is not fleshed out. Whatever the aliens are up to, he says, humans have a great need and an old one: to reconnect with the cosmos through unconditional love. The problem with this is that it gives the book a slightly mystical aura that the rest of the contents do not support or discuss. Nevertheless, Mannion's book is different in that it seeks to locate patterns in all the activity around UFOs in the last half a century, and this makes the book valuable. Alexander lmich Anomalous Phenomena Research Center New York, NY Cryptozoology A to Z: The Encyclopedia of Loch Monsters, Sasquatch, Chupacabras, and Other Authentic Mysteries of Nature by Loren Coleman and Jerome Clark. New York: FiresidelSimon & Schuster, 1999. 270 pp. Book Reviews 469 $13.00, paperback. ISBN 0-684-85602-6. Illustrations and photographs. Bibliographical references. Loren Coleman, who lives in Portland, Maine, has an undergraduate degree in anthropology-zoology and a graduate degree in psychiatric social work. He has been an instructor, assistant, and associate professor, as well as a research associate in various academic university settings since 1980. He authored, coauthored, or edited eight books, including Suicide Clusters and The Field Guide to Bigfoot (with Patrick Huyghe, as well as numerous articles on elusive animals. His coauthor Jerome Clark, a resident of Canby, Minnesota, is the author of several popular encyclopedias on the history of natural phenomena. French scholar Bernard Heuvelmans, who is frequently referred to as the "father of Cryptozoology" (p. 105), has defined the field as "the scientific study of hidden animals, i.e., of still unknown animal forms about which only testimonial and circumstantial evidence is available, or material evidence considered insufficient ' (p. 75). The book contains over 120 entries on "cryptids" (as those enigmatic creatures are called) and includes more than 50 biographies of researchers, explorers, and authors as well as information on some organizations and periodicals in the field. Thus, the reader has a chance to learn about numerous purported mysterious animals-from the popular Abominable Snowman (also known as "Bigfoot" and "Yeti") to the obscure "Zuiyomaru Monster," which was reportedly found off the coast of New Zealand in 1977 but whose sighting disappointingly proved to be a decomposed shark. To summarize, the entries about the "cryptic" animals clearly fall into three categories: searched for and already discovered species (those that may be called "right" or "certain"), problematic and disputed ("undecided"), and those proven to be mistakes and hoaxes ("wrong" or erroneous). Some may doubt whether the field of cryptozoology actually exists. On one hand, critics of the field frequently refer to unreliability of witnesses due to faults of human attention and memory. They argue that sightings of unexpected events, such as unknown or rare animal species, present considerable challenges to accurate perception, interpretation, and recall (e.g., Hall, McFeaters, and Loftus, 1987), let alone deliberate hoaxes. On the other hand, some hitherto unknown species of animals have been discovered in this century, including the okapi (which is the centerpiece of the logo of International Society of Cryptozoology founded in 1982) in Congo, the giant panda in Tibet, and the latest saola ("forest goat" Pseudoryx ngbetinbensis) in Vietnam's Vu Quang Nature Reserve near the Laotian border. Thus, it is not surprising that the general public is more skeptical with regard to the disputed animals than trained professionals: While only 13% of average Americans believe in existence of Nessie (also known as the Loch Ness monster, a supposed inhabitant of Loch Ness lake in Scotland), nearly 40% of marine biologists give credence to the evidence that Nessies are real (Bauer, 1987, p. 70). It is also true that many animals have become extinct over the last few hundred years as a result of ecological disruption (e.g., deforestation) caused by 7 ~ 470 Book Reviews humans migrating to new environments or direct attempts to exterminate some species deemed to be "harmful." An example is the thylacine (pp. 238-239), also known as marsupial wolf or Tasmanian tiger (Thylacinus cynocepha1us)-a doglike carnivorous animal that was hunted to extinction in the early part of the 20th century. The last known living specimen died in captivity in 1936 in the Hobart Zoo in Australia. Persistent reports of sightings right down to the present offer hope that a few thylacines may still live in the wild. For example, a Park and Wildlife officer reported observing a thylacine in the Pyengana region of eastern Tasmania in January 1995. However, a follow-up search failed to find any trace of this animal. The book would benefit from a more holistic interdisciplinary approach essential for research into anomalies-a more extensive discussion of alternative, including nonbiological, explanations for the nature of mysterious creatures. One plausible point of view is a psychological explanation briefly mentioned in the book under the entry "Goblin Universe" (p. loo), which touches upon potential "paranormal" (in the authors' terminology) or rather archetypal origin of beliefs in legendary and mythological creatures. In this connection, one can remember Carl Jung's (1964, p. 70) example of a 10-yearold girl who dreamed of "the evil animalw-a snakelike monster with many horns who kills and devours all other animals. According to Jung, such "archaic remnants" or "primordial images," which are encountered in different cultures around the world, originate in the collective unconscious. Thus, the Piasa, a mythological creature found in legends of the Illinois-an Algonquian-speaking Indian tribe that concentrated along the Illinois River to the Mississippi River-and discussed in this volume (pp. 201-202) may fall into this category. Another example of a nonbiological explanation not addressed in this book is related to such persistent type of disputed creatures as sea snakes or lake monsters, the most famous of which is the Loch Ness monster, or Nessie. Back in 1987, anomalies researcher James Westman suggested that the Loch Ness monster may represent a form of energy rather than a living creature and should thus be studied by physicists rather than biologists. It is also conceivable that at least some of such sightings may actually be due to standing wave phenomena such as soliton waves or little-known weather phenomena such as miniwaterspouts. On the positive side, the book offers a multicultural perspective of the search for "cryptic" animals, including a discussion of Russian researcher Boris F. Porshnev's theories regarding yet undiscovered primates such as Almas or Almasty (Abominable Snowman-like creature) whose giant footprints were reportedly found in Tien Shan mountains and in the Caucasus in the south of Russia. Beginning with the 1950s, these "relict hominoids" were studied by a formal commission of the Academy of Sciences of the USSR. I still remember debates on the subject at meetings of the Moscow Society of Naturalists in the early 1970s. Book Reviews 47 1 I Although the field of cryptozoology is highly controversial, the book represents a useful and informative contribution to public knowledge of anomalies. A curious reader interested in the field will benefit from it-at least until a more scholarly encyclopedia on the subject is published. An easily readable style and extensive bibliography are particularly useful, even though there is no division between scholarly and popular publications, as are informative lists of museums, exhibitions, periodicals, and Web-sites. Larissa Vilenskaya 101 0 Harriet Street, Palo Alto, CA 94301 -3430 References Bauer, H. H. (1987). Society and scientific anomalies: Common knowledge about the Loch Ness monster. Journal of Scientific Exploration, 1(1), 5 1-74. Hall, D. F., McFeaters, S. J., & Loftus, E. F. (1987). Alterations in recollection of unusual and unexpected events. Journal of Scientific Exploration, 1(1), 3-10. Jung, C. G. (1964). Man and his symbols. Garden City, NY: Doubleday. Westman, J. (1987). Comments on bauer's "Loch Ness monster." Journal of Scientific Exploration, 1(2), 101. Understanding Tomorrow's Mind edited by Larry Vandervert (a special issue of the Journal of Mind and Behavior, volume 18). This special volume of the Journal of Mind and Behavior (spring and summer 1997 issues of volume 18) is devoted to papers relating chaos theory and quantum mechanics (QM) to psychology and the study of consciousness. The collection is titled "Understanding Tomorrow's Mind," and the implication is that further understanding of consciousness and the mind will require the use of tools such as the mathematics of chaos theory and the physics of QM. Although this is a fairly popular idea in some cognitive science circles (particularly in some corners of artificial intelligence and artificial life research), most work in psychology makes no connection with developments in the forefront of physics or mathematics. The claim that fundamental advances in these fields have anything to do with the study of behavior and consciousness is a bold one, because it is often thought that cognition arises at a much higher level and is controlled by principles far removed from the laws governing the bedrock of matter, within which mentality is embedded. This book presents an interesting sample of approaches that aim to make progress in cognitive research by using tools from fundamental physics and mathematics. The volume presents a sample of applications of nonlinear and quantum modeling approaches for the dynamics underlying psychological phenomena. It starts with a brief preface and a set of guidelines for using this book as a teaching tool. This is a very useful inclusion; however, it would have been even more useful if it had included some references and pointers to areas of work specifically not included in this collection (more on that below). 472 Book Reviews Chapter 1 is a tutorial on chaos by B. West-a very readable, easily understandable piece on the fundamentals of low-dimensional, deterministic, nonlinear, dynamical systems theory (i.e., chaos theory). The chapter has almost no formulas, but it does have enough graphs and a very lucid explanation of the fundamentals-what this theory is and what it is not. There is enough meat here to emphasize the relevant conceptual aspects of nonlinear sciences. Particularly useful is a table contrasting various aspects of scientific research as they are viewed from the classical versus the "nontraditional" paradigms of systems theory. The key concepts of chaos theory are given as: emergence, deterministic but unpredictable behavior, large effects caused by small inputs, discontinuous and nonlinear behavior, no characteristic scale (interesting detail appears at a wide range of magnifications along some parameter), and the necessity for qualitative understanding of the system. West covers examples in topology, physics, mechanics, embryogenesis, psychophysics, business/economics, and evolution. This is a very valuable paper in every way and should prompt every interested reader to delve more deeply into books such as West (1990), Barbi (1998), or Liebovitch (1998). The next chapter, by H. Stapp, is an introduction to QM. The introduction is a very good one; it begins quite abstractly, without the usual examples of basic physics experiments that demonstrate quantum "weirdness." Such examples are, I think, necessary to a good understanding of what QM is all about, and one can get a flavor for them in many good books understandable by the nonexpert. Examples include Herbert (1985) and Gribbin (1995). Stapp's approach is interestingly different from most expositions of QM and is quite relevant to the psychological theme of this collection. He discusses QM in relation to the work of William James, epistemology, and the notion of truth. He sums up the implications of QM as that "the complete description of nature at the atomic level was given by probability functions that referred not to underlying microscopic space-time realities but rather to macroscopic objects of sense experience." (p. 127) There are of course other intersections of QM with philosophy of mind, including issues of predictability (Dennett, 1984), dualism versus materialism versus idealism (Beloff, 1989; Smythies and Beloff, 1989), and the role and nature of consciousness in the material world (Wigner, 1979; Hodgson, 1991). The next 10 chapters are papers on aspects of cognitive science, which range from general metaphysical issues (such as "The Science of Consciousness and the Hard Problem") to very specific QM or chaos-mathematical models of neurobiological phenomena (such as "Phase Transitions in Learning"). The selections are generally quite interesting. There is not enough space to discuss them all, but a few specific comments can be made. An often-mentioned point (on which there is no consensus among the experts) is that QM, the ultimate theory of materialism (a rock-bottom theory of matter), seems to bring in dualism through the fundamental concepts of the observer and the wave function collapse (Wigner, 1979). In his chapter, Chris King makes an interesting analogy relevant to the issue of free will. QM has several properties that begin to sound as if they might provide a substratum to Book Reviews 473 free will, a concept that finds no comfortable place in pre-quantum science. Quantum processes are often nonlocal, unpredictable, and "ghostly" (Davies and Brown, 1986; Gribbin, 1995). A single particle in QM can occur anywhere within its wave function, despite a probability distribution strictly governing the aggregate. King proposes that free will in a single brain corresponds to quantum uncertainty in the ongoing brain state. He cites Eddington's 1935 estimate that "the uncertainty of position of a synaptic vesicle was as great as the width of the membrane, thus constituting a possible trigger for an unstable cascade." (p. 156) The approach of leveraging free will from unstable, nonlinear magnification cascades ( 2 la chaos theory) set off by quantum events has been much discussed. However, by itself it merely injects an element of randomness into behavior: Pure randomness coupled with deterministic brain algorithms is not what we really mean by "free will," although it may provide for one aspect of it-unpredictability. These ideas are analyzed by Dennett (1984), who argues that ultimately not only do we not know what we mean by free will, but also the popular concept of it that we hold to simply does not correspond to anything in the real world. A somewhat similar approach is used by John Eccles (Popper and Eccles, 1984), but there, the quantum uncertainty and chaotic amplification is a point of contact with a nonphysical consciousness; thus, an attempt is made to allow control by an immaterial mind while answering conservation of energy arguments against dualistic interactionism. G. Globus's paper (p. 195) contrasts the approaches of connectionism (the science of the properties of neural networks) with traditional representationalist information-flow-based A1 paradigms. He draws an analogy between optimization of the behavior of neural net with the minimization of a neural Lagrangian operator (a quantity associated with the velocity and acceleration of ionic currents in the brain). "I am nonlocal control and my meanings are cybernetic variables." (p. 195) His view is that consciousness is quantum fields "hoisted" by biosubstrates. It is not clear what this "hoisting" process is, and why only biological systems can accomplish it. There are many complex dynamical systems. What are the features of biology that make cognitive systems special? Globus then discusses the nature of the "I," or central controlling self; such a thing has not been found by neuroscience, and much discussion has taken place (Dennett, 1980; Hofstadter and Dennett, 1981) on whether this is an illusion and how (and why) this illusion is maintained. Globus's suggestion is that this issue is resolved by nonlocal control exhibited by quantum dynamical systems. He avoids category mistakes by identifying the "I" with being a brain under quantum control, not with the brain itself. A remaining problem of course is that for any physical mechanism (whether classical or quantum) proposed to account for the "I" or "self' is found, the natural response is still "yes, that is what explains the behavior of the system, but where is the first-person subjective experience, the unified self perspective we all share?" In light of this issue (which of course is viewed as a pseudoproblem by many), the identification of nonlocal controls of the brain's dynamics is more a contribution to the "easy" problem (i.e., basic neuroscience) than the hard problem (philosophy of mind). 474 Book Reviews Jibu and Yasue (1995) present some interesting ideas that have been influential in several of the contributions to this volume. They develop models based on quantum ordering of water in the brain and show derivations of mathematical properties of such systems as they relate to dynamics of brain states. Their ideas are more fully treated in their book. There are a couple of confusing statements in their contribution though. Molecular biology is a branch of biophysics developed phenomenologically by gathering fragmentary knowledge of molecule-based organic chemistry, quantum chemistry, and quantum statistical mechanics. Patching those fragments into a single system of knowledge is made by the longstanding and excellent intuitions of biologists. Nevertheless, molecular biology does not succeed in explaining the essential aspects of the brain functioning directly related to the highly advanced mental "objects" (p. 207). This is a mischaracterization of molecular biology, which is neither phenomenological nor fragmentary, nor does it focus on quantum chemistry. And of course, molecular biologists do not work toward explaining brain functions. Perhaps the authors are actually making a deeper claim: that the reductionist approaches of molecular neurobiology are insufficient to ever explain brain functioning. If so, I would have liked to see a discussion of their opinion of the theoretical shortcomings of the currently popular approaches based on the neurophysiology of real networks in the brain. Such models have worked fairly well in explaining aspects of visual processing and memory. I would also have liked to see some specific predictions of their model. W. Freeman presents a very interesting paper on the neurodynamics of intentionality (his inroad of choice toward the mysterious edifice of consciousness). The paper discusses brain dynamics of voluntary actions, as well as representations in the brain. He concludes that "there are no representations in brains, only meanings" (p. 301). These issues have been given good discussion by Libet's group (Libet, 1993). In summary, this is a very valuable book. Interestingly, it has no contributions by some prominent researchers in the area of quantum theory and cognitive science (such as Brian Josephson, Roger Penrose, Stuart Hameroff, etc.). On the one hand, this is good, because it left more room for contributions by ideas from very different schools of thought. Thus, this is a great complement to books such as Hameroff, 1987; Hameroff et al., 1996; Penrose, 1996. On the other hand, it would have been nice to have at least a couple works from those authors included for comparison. A concluding essay would also have been very welcome, if it had summarized the approaches and conclusions of the papers and compared and contrasted different approaches to the role of QM and chaos theory in cognitive science. An overview of future prospects and on-going research would also have been useful to outline possible areas for new and interested researchers. In general, this book is quite useful but is more for the expert in the field rather than for the required reading of all psychology students that the intro- Book Reviews 475 duction suggests. Most of the essays (except the excellent first two chapters of overviews) discuss topics at a level suitable for specialists but do not spell out the basics. Of course, this is impossible to do in a book with reasonable space constraints, given the complexity of all three of the subjects involved (QM, chaos theory, and cognitive science). Some of the papers (such as that by Sabelli et al.) are so densely packed with new ideas, models, and analogies that they are very hard to follow and read as a compressed, concise set of notes taken by someone who has had daily discussions with the author and understands the implications thoroughly. This book is best digested after familiarity with the contents of, for example, Churchland (1988), Gribbin (1995), Gleick (1988), Peitgen et al. (1992), etc. It is a great companion to Lockwood (1991), Searle (1985), Ramachandran and Josephson (1979), etc. Michael Levin Cell Biology Department Haward Medical School Boston, MA 02115 levin @ whiz.med. haward.edu References Barbi, M. (1998). Chaos & noise in biology and medicine. World Scientific Publishing. Beloff, J. (1989). The case for dualism. Churchland, P. (1988). Matter & consciousness. MIT Press. Davies, P., & Brown, J. (1986). The ghost in the atom: A discussion of the mysteries of quantum physics. Cambridge University Press. Dennett, D. (1980). Brainstorms: Philosophical essays on mind & psychology. MIT Press. Dennett, D. (1984). Elbow room: The varieties offree will worth wanting. MIT Press. Gleick, J. (1988). Chaos: Making a new science. Viking Penguin. Gribbin, J. (1995). Schrodinger's kittens & the search for reality. Little, Brown & Company. Hameroff, S. (1987). Ultimate computing: Biomolecular consciousness & nano technology. Elsevier Science. Hameroff, S., Kaszniak, A., & Scott, A. (1996). Toward a science of consciousness. MIT Press. Herbert, N. (1985). Quantum reality: Beyond the new physics. Doubleday. Hodgson, D. (1991). The mind matters: consciousness & choice in a quantum world. Oxford University Press. Hofstadter, D., & Dennett, D. (1981). The mind's I. Basic Books. Jibu, M., & Yasue, K. (1995). Quantum brain dynamics & consciousness: An introduction. John Benjamins North America. Libet, B. (1993). Neurophysiology of consciousness: Selected papers of Benjamin Libet. Birkhauser Boston. Liebovitch, L. (1998). Fractals & chaos simplified for the life sciences. Oxford University Press. Lockwood, M. (1991). Mind, brain & the quantum: The compound "I" . Blackwell. Peitgen, H., Jurgens, H., & Saupe, D. (1992). Chaos & fractals: New frontiers of science. Springer-Verlag. Penrose, R. (1996). Shadows of the mind: A search for the missing science of consciousness. Oxford University Press. Popper, K., & Eccles, J. (1984). The self & its brain: An argument for interactionism. Routledge. Ramachandran, V., & Josephson, B. (1979). Consciousness & the physical world. Elsevier Science. Searle, J. (1985). Minds, brains & science. Harvard University Press. Smythies, J., & Beloff, J. (1989). The case for dualism. University Press of Virginia. West, B. (1990). Fractal physiology & chaos in medicine. World Scientific Publishing. Wigner, E. (1979). Symmetries & rejections. Ox Bow Press. 476 Book Reviews The Abduction Enigma by Kevin D. Randle, Russ Estes, and William P. Cone. New York: Forge, 1999.4 16 pp. $25.95, hardcover. The Abduction Enigma is unlike many other works on the "the abduction phenomenon," because it takes a critical look and does so by coalescing the perspectives of three authors with varied backgrounds. That does not mean that each author is a hardened "skeptic" or "debunker." The first author is a well-respected ufologist who is perhaps best known for his work on the Roswell case. It is noteworthy that although Randle believes that Roswell involved the crash and retrieval of at least one extraterrestrial craft by the U.S. government (what might be called a "reverse" alien abduction!), he agrees with his coauthors-who have experienced anomalous events themselvesthat there is no scientific evidence for justifying the belief that the human race is being secretly (or not so secretly) abducted by aliens. The tone of the book gives the impression that the authors probably did not believe the extraterrestrial hypothesis was reasonable to begin with, but then again, many scientists hold apriori beliefs about topics they choose to investigate. The authors have given a nice background from the historical and sociological aspects of the abduction phenomenon to modem trappings of this area of research. There is also a nice section that reviews the major researchers currently involved in this field. However, the main value of this book in my opinion is their largely comprehensive review and integration of a vast amount of academic literature on this and related topics. A study of their bibliography revealed that most material was drawn from many articles in both mainstream and specialized journals, as opposed to the content of popular books and newsletters. The authors make clear arguments against the reality of alien abductions based on this collection of literature by showing how various evidence offered in support of the extraterrestrial hypothesis is weak to nonexistent. But the authors' valid arguments are also mixed with clear and sometimes superficial speculations on the validity of conventional explanations that are given as alternatives to the extraterrestrial hypothesis. Many readers might see this as a deficit, but I see it both as an asset and as an opportunity. In particular, many of their speculations on how the vast literature fits together to provide a conventional explanation for the abduction phenomenon can be falsified, so the authors have purposefully (or maybe unwittingly) given researchers new hypotheses to be tested. I was surprised that the well-articulated psychological theory proposed by Newman and Baumeister (1996) was neither specifically addressed at length nor even cited in this book. In fact, the entire issue of Psychological Inquiry (volume 7, 1996) included reactions to Newman and Baumeister's (1996) theory from some of the leading researchers in this area, as well as researchers in mainstream psychology. The reader might be asking why I would criticize the authors for this seemingly insignificant omission in light of the large number of articles they did consult. The Psychological Inquiry series of articles was in my opinion one of the most Book Reviews 477 important published contributions to this field of study (along with Pritchard et al., 1994) for bringing this phenomenon to the attention of serious scientists. Consequently, I would have liked for it to have been incorporated into a book of this scope. Let me emphasize here that aspects of Newman and Baumeister's (1996) arguments, such as the dangers of hypnotic regression and the likely psychological motivations that underpin abduction experiences, are discussed throughout the book, just not given the level of detail and amount of integration found in the original article. The production of the book is of very high quality and includes an excellent bibliography, as well as a combined subject and author index. The contents of the book are largely accessible to a general audience, but with only 17 illustrations, the general reader may feel this book lacks "enough pictures." Personally, I do not mind the lack of photographs because the clear and forceful writing style make the contents of the book extremely enjoyable to read and contemplate. I enthusiastically recommend it to anyone who wants an excellent introduction to this field of research or as a source of potentially new hypotheses for future study. Perhaps the greatest compliment I can give the authors is that the book actually inspired me to begin designing studies to test some of the ideas contained therein. Even if humankind is not being abducted by aliens, there would still remain a rich area of investigation on how and why sociological, biological, and psychological variables interact to produce abduction experiences. Thus, the abduction phenomenon has value outside the immediate context of ufology and may ultimately provide unique insights into the mechanisms that govern imagination, cognition, and personality. James Houran Department of Psychiatry SZU School of Medicine P. 0. 5201, Springfield, IL 62705 Box I ~ References Newman, & Baumeister. (1996). Toward an explanation of the UFO abduction phenomenon: Hypnotic elaboration, extraterrestrial sadomasochism, and spurious memories. Psychological Inquiry, 7, 99-1 26. Pritchard, A., Pritchard, D. E., Mack, J. E., Kasey, P., & Yapp, C . (Eds.). (1994). Alien discussions: Proceedings of the abduction study conference. Cambridge, MA: North Cambridge Press. ~ I Life on Other Worlds: The 20th-Century Extraterrestrial Life Debate by Steven J. Dick. New York: Cambridge University Press, 1998. 290 + xiii pp. $24.95 (c). ISBN 0-521-62012-0. Steven Dick is an astronomer and historian of science at the U.S. Naval Observatory. He has served as historian for the now-defunct National Aeronautics and Space Administration (NASA) Search for Extraterrestrial Intelligence 478 Book Reviews (SETI) program and was a member of a NASA workshop that examined the cultural aspects of success in SETI, including the short-term and long-term implications of contact with extraterrestrials. Unlike most others with such credentials, Dick also takes UFO reports seriously, is willing to say so publicly, and, if gently, chide establishment science for giving the subject short shrift. See more about this and its implications below. Life on Other Worlds is an abridged edition of Dick's Biological Universe: The Twentieth-Century Extraterrestrial Life Debate and the Limits of Science (New York: Cambridge University Press, 1996), undertaken, in his words, "in an attempt to reach a wider audience and.. .to integrate.. .new discoveries into the story." He has succeeded in giving us a marvelous volume of interest and value to expert and intelligent laymen alike. Discovery of extrasolar planets and the ensuing debate over what this means, the continuing controversy over the possible discovery of microfossils in Martian meteorites, and the recent identification of water in other meteoritic material are but the latest developments in a debate that is centuries old. Dick provides a brief but substantive review of the course of the debate and range of thought on the subject of extraterrestrial life from their earliest origins. Then, beginning with the Mars observations and theories of Percival Lowell in the late 19th century, he focuses on modern times. Lowell's claim to have discovered canals on the Red Planet, the for a time hope-raising results of the Viking spacecraft biological experiments, the Marsrock donnybrook, and SETI have titillated and at times consumed science and the public. Similarly, centrally related matters such as the search for planets orbiting stars other than our own, the quest for explanations for UFO reports, and the pursuit of understanding of the mechanisms giving rise to life continue to excite wide interest and controversy. The themes of life beyond our home planet, the discovery of such life, and, more often and usually more chillingly presented, its discovery of us have been fodder for science fiction authors from H. G. Wells and Kurd Lasswitz to Arthur C. Clarke and Robert Heinlein. Hollywood, too, has capitalized on broad public fascination with the subject, resulting in some of the most popular films of all time, among them E. T , Close Encounters of the Third Kind, and Independence Day. In Life on Other Worlds, Dick details in readable and nontechnical prose the history of the modern ET-life1ET-intelligence (ETI) debate in which these popular manifestations are rooted. This is not just a chronological recitation. Dick contributes an important dimension to the understanding of the questions and theories involved by placing them in historical perspective, showing how the concept of ETI arises from a "biophysical cosmology" that seeks confirmation no less than do competing physical cosmologies. In so doing, he very effectively demonstrates this is a subject at the very limits of the reach of science. Thus, the quest for confirmation is not only a pursuit of new scientific knowledge and a validation of ideas central to modern biology and other disci- Book Reviews 479 plines, but it also offers a dramatic "happening now" window on the nature and limitations of science itself. Dick's chapter "The UFO Controversy and the Extraterrestrial Hypothesis" provides a snapshot of the problems and promise arising from that nature and those limitations. In the opening paragraph, Dick notes, "Between public gullibility and scientific closed-mindedness, between perception and reality, lies an important chapter in the history of the extraterrestrial life debate and a story of the limits of science under the most trying circumstances." He then proceeds in a mere 32 pages to provide a remarkably meaty and well-balanced account of 50 years of ufological history, illuminating along the way the sources of scientific reluctance on the subject. Wrapping up his consideration of the UFO problem and its linkage to the ETI question, Dick writes, [Tlhe history of the debate finds abundant blame on all sides: The unwillingness of many scientists to actively engage the UFO phenomenon is certainly understandable in terms of peer pressure and career advancement, but the desire of some to squash the subject without considering the evidence flies in the face of scientific curiosity that supposedly led them to science to begin with. On the other side, the outrageous claims and hoaxes that presently flood the field are unworthy of scientific attention, and no one should be surprised if scientists fail to engage every will-of-the-wisp [sic] report. In the middle of these extremes may yet be a phenomenon that requires study, if only one can find it in the midst of the twin human failings of perception and deception (p. 167). So how can those of us who are engaged in serious study of UFOs clearly identify that phenomenon (or, more likely, phenomena) and give it scientific salience that is reasonably impossible to ignore? Taking this a step further, how can those of us who are of the opinion that some UFO reports are of observations of the handiwork of ETI and, in some cases, that ETI itself, bring ufology back into the scientific ET-life debate? The answer was forcefully suggested by physicist and pioneering SET1 advocate Philip Morrison, in his paper for the famous 1969 American Association for the Advancement of Science UFO symposium. As Dick summarizes, [Morrison,] with his penchant for zeroing in on the significant issues, argued that the debate came down to the nature of scientific evidence. "Reproducibility" was not enough, for one could not reproduce an aurora or eclipse, nor was "hard evidence" enough. The prime requirement for responsible evidence, he held, drawing a parallel with [that favorite of ufologists] the 19th-century acceptance of meteorites as extraterrestrial, was "independent and multiple chains of evidence, each capable of satisfying a link-by-link test of meaning" (pp. 158-1 59). Morrison also went on to contend, as Dick puts it, that "[nleither the extraterrestrial hypothesis nor any other explanation of UFOs had multiple chains of evidence or a link-by-link test." In this reviewer's opinion, this remains an open question, and it should be taken up by ufologists as their chal- 480 Book Reviews lenge and, ultimately, their challenge to science. To put it unscientifically, or perhaps protoscientifically, the shades of J. Allen Hynek and James McDonald are demanding it of us. Life on Other Worlds is, of course, about more than UFOs. However, the political, social, personal, institutional, and economic dynamics bearing on the UFO controversy within the precincts of science dramatically and sharply reflect those surrounding the broader issues of the extraterrestrial-life debate. In turn, these illuminate important down-to-earth truths that bear on the pursuit of other subjects on or just inside the fringe of scientific acceptability. The members of the Society for Scientific Exploration-who by definition are engaged in or at the very least are deeply interested in matters currently at the limits of and perhaps beyond the reach of science-can learn a good deal about such "unfortunate" realities from this book, a useful bonus of this enjoyable and well-crafted volume. Karl 7:Pflock P. 0. Box 93338, Albuquerque, NM 87199-3338 Ktperehwon @aol.com Shattering the Myths of Darwinism by Richard Milton. Rochester, VT: Park Street Press, 1997. xi + 308 pp. $24.95 (c). ISBN 0-89281-732-1. Richard Milton is an English journalist and design engineer. It is important to mention at the outset that he is not a creationist; neither is he a mainstream scientist. Rather he is consigned to that limbo between religious conviction and the certainties of doctrinaire science. That he is not a prisoner of paradigms can be seen in his earlier book, Alternative Science, in which he is concerned with parapsychology, cold fusion, and other subjects on the borderlands of the scientific endeavor. In Shattering the Myths of Darwinism, however, Milton aims directly at the heart of modern science: Evolution is the idea that things, especially biological things, change and develop. Usually these changes are from simple to complex, as from bacterium to human. Unquestionably, biological evolution is one of the most powerful and persuasive of all scientific paradigms. It is right up there with the big bang theory and plate tectonics. Despite the implications of his title, Milton does not shatter evolution per se; rather, he attacks the currently dominant idea of the mechanism of biological evolution-that is, random mutation and natural selection. Here are his own words: I accept that there is persuasive evidence for evolution, but I do not accept that there is any significant evidence that the mechanism driving evolution is the neo-Darwinian mechanism of chance mutation coupled with natural selection.. . I do not believe that the earth is only a few thousand years old. I present evidence that the currently accepted methods of dating are seriously flawed and are supported by Darwinists only because they provide the billions of years required by Darwinist theories (p. xi). Book Reviews 481 These words from the preface to the U.S. edition were written in response to the "hate" reviews that appeared when the book was published in England. Richard Dawkins, a luminary in the Darwinist camp, suggested that Milton needs "psychiatric help." This all means that the book is a wonderful repast for anomalists. In just 5 minutes, one can see why Dawkins and other evolutionists panicked. Milton systematically and objectively (in my view) reviews and evaluates the facts of geology, paleontology, and zoology. Conclusion: The Darwinist paradigm is untenable. Milton's specific doubts concerning the prevailing evolution paradigm generally parallel those that the creationists have been expressing for decades. The difference is that Milton avoids claims of supernatural, "intelligent" involvement with the machinery of evolution. He assumes the development of life is susceptible to rational scientific inquiry. Geology is an avocation for Milton, and he understandably zeroes in on the dating of strata. Most scientists are quite comfortable with the claim that radiometric techniques have once and for all accurately dated the earth and those onionlike layers of sediment that gird our planet. Milton informs us that this assumption is "entirely false." First of all, geologists do not radiometrically date strata directly. They can only date the igneous rocks, such as lavas and volcanic ashes, which are interbedded with the shales, sandstones, etc. Milton asserts that such dating by association is unreliable. Secondly, radiometric dating involves the measurements of by-products of radioactive decay. There is much room for error here. This can be seen in the world's two largest deposits of uranium ore in Canada and Zaire. These contain considerable quantities of lead-208 but little thorium-232. Such ratios are not consistent with the decay of uranium. Milton says that such salient anomalies cast doubt on all of radiometric dating. The gist of Milton's lengthy scrutiny of radiometric dating is that the evolutionists cannot attach reliable dates to the planet's sedimentary layers, or even to the planet as a whole. These uncertainties make it difficult to draw evolutionary family trees. Furthermore, these sedimentary layers, as Milton likes to point out, occur only in isolated patches here and there. Nowhere can one find a complete geological column such as those printed as facts in all the textbooks. Just as patchy is the fossil record imbedded in these sedimentary layers. This is illustrated in the well-known fossil record of the horse. Milton claims (as do some other scientists) that this fixture of the Darwinian argument is really concocted from incomplete skeletons assembled from widely separated localities. Milton worries about this as follows: It is, however, more than a little disturbing to learn that the descent of horses is being offered as the decisive evidence in favor of evolution on the strength of so little real physical evidence, and with so many gaps filled only by speculation (p. 104). 482 Book Reviews He is also concerned about the stubborn fixity of species; that is, the apparent limits met during breeding experiments conducted with many insects, birds, and mammals. This "genetic homeostasis" was actually recognized by Darwin himself and has been confirmed many times over, as in experiments with the fruit fly Drosophila. In other words, there are puzzling barriers that seem to thwart biological change. Many similar "problems" for the evolutionist can be found in Shattering the Myths of Darwinism. One more should suffice here: the famous finches of the Galapagos. Widely advertised as a proof of speciation 2 la the Darwinian mechanism of random mutation plus natural selection, ornithologists really find that these 13 "full" species are quite confusing. They vary morphologically within a species according to environmental pressures and frequently hybridize to produce vigorous offspring. Milton quotes the ornithologist David Lack as stating, "in no other birds are the differences between species so ill-defined." Apparently Darwin's finches are not clear-cut proof of Darwinism after all. All of Milton's examples are chosen to undermine one's faith in the current evolution paradigm. But even Milton is certain that species do change into new species; that is, they evolve. It is that elusive mechanism that is at issue in his mind. He asserts, for example, that natural selection is not the "engine" of evolution. Natural selection really reduces biodiversity as it screens out the unfit. In contrast, evolution actually increases diversity. Having cast doubts in our minds about the efficacy of small random mutations plus natural selection to produce speciation, Milton turns his attention to other potential mechanisms that might create diversity. Some are old ideas while some are new. Lamarck is first on his list and is not discarded, as is the usual custom. The "hopeful monsters" of Richard Goldschmidt are not far behind in Milton's mind. Also considered are the more recent "directed mutation" experiments of John Cairns. Even the morphogenetic fields of Rupert Sheldrake are brought into the discussion. Strangely, though, the work of Lynn Margulis on the potential role of endosymbiosis is not mentioned as a mechanism of evolutionary change. When one finishes the book, one is left with the shards of Darwinism and no obvious, convincing way of forging a replacement paradigm. But Milton's title promises only iconoclastics, so we should not penalize him for not producing a new synthesis. In truth, Shattering the Myths of Darwinism is well written and objective. After reading it, you will no longer meekly accept that we now know all there is to know about how, when, and why new branches sprouted on the tree of life. William Corliss Sourcebook Project RO. Box 107, Glen Arm, MD 21057 Book Reviews 483 Dark Matter, Missing Planets and New Comets (Paradoxes Resolved, Origins Illuminated), Revised Edition by Tom Van Flandern. Berkeley, CA: North Atlantic Books, 1999. 551 pp. $22.50, paperback. ISBN 1-55643-2682, tvf @ mindspring.com Suppression Stories by Brian Martin. Wollongong, Australia: Fund for Intellectual Dissent, 1997. 17 1 pp. Paperback. ISBN 0-646-30349-X, http://www. uow.edu. aulart s/sts/bmartin/dissent/documents/ Astronomy and its theoretical outgrowth astrophysics are not in the popular mind regarded as revolutionary sciences in the Kuhnian sense; yet we might recall that the observation of the Jovian moons four centuries ago, the extraterrestrial origin of meteorites two centuries ago, the confirmation of light bending near the sun early this century, and the recent impacts of comet Shoemaker-Levi 9 on Jupiter mark four major changes in paradigm over the relatively short course of science. Tom Van Flandern is a Yale-educated Celestial Mechanic and former U.S. naval observer. He notes, according to a previous reviewer (Keay, 1993), that current observations within the solar system suggest the speed of gravity is "at least 20 times the velocity of light." Does that enliven your sense of spacetime? Read on and discover that asteroids and comets are creatures of orbits (chapter 6) originating not from an unformed planet (the asteroid belt) and the almost inconceivably large Oort sphere (Van Flandern calculates it would hold all the stars in the Milky Way), but from the ejecta of a recent (about 3 million years ago) catastrophic event in or near the asteroid belt. As an anthropologist with a strong interest in astronomy, I cannot judge the ultimate merit of either the current or Van Flandern's alternative hypotheses. I can, however, observe that some of Van Flandern's hypotheses predict current reports of novel data. Astronomy is among the oldest human professions. Insofar as humans are political, the proclivity of power elites and bureaucracies to perpetuate themselves, and the corruptibility of humans in the face of and even more so in the defense of institutional power, are proverbial. Consider the coincidence between the Egyptian realignment of temples and the period of unrest between the old and new kingdoms (Tompkins, 197 1): a cost of the discovery of precession? Did our planetary system reconfigure itself, 3 million years ago, into our nine (eight?) planet system, with the moons of Venus and X-Mercury and Mars, respectively-migrating to solar orbits? An earlier reviewer (Castle, 1994) finds many of the issues raised by Van Flandern interesting but knows of "no.. .mechanism that would cause a planetary breakup." I believe the same objection was raised against Wegener s continental drift. Why is Iapetus dichotomously dark and light if not from some event within the solar system? Earth could have been on the far side to it. At 3 million years ago, no "death star" binary of the sun with a 26-million-year periodicity (of the "great dyings," the last roughly 13 million years ago) is implicated. Since the 7 484 Book Reviews 7 publication in early 1999 of Van Flandern s book, readers of Science have been treated to news of three additional members of our solar system with bisected terrain: Mars (May 28), Ganymede (October I), and Triton (October 15). Triton has "all the craters.. .on one side" of a surface "possibly less than 10 million years old," based on the reanalysis of 1989 data. "About 40% of Ganymede's surface is covered by dark, heavily cratered terrain, and the remainder.. . bright terrain." New Mars Orbital Surveyor (MOS) Mars topography shows six enormous shield volcanoes, three great circular impact basins (five of the former roughly antipodal to the greatest of the latter), the northern hemisphere with featureless, abyssal (?) lowlands and heavily cratered (moonlike) highlands to the south. If the shield of the five volcanoes were lowlands, the northlsouth topographic divide would be very nearly half and half. Van Flandern mentions dichotomous Martian terrain in support of the ex-planet X hypothesis, but what we see with new MOS data is not supportive of that hypothesis. Given featureless highlands, it might be another matter. And given the apparent falsification of the "face" on Mars (Pierri, 1999) by MOS photographic data, one wonders what other surprises are to be found in the not yet disclosed Cydonia photographic data. Why is an anthropologist reviewing this "revised" edition (of a book published 7 years ago, with four new chapters appended) rather than an astronomer? The luck of the draw? Perhaps, but let me report the puzzlement of the research librarian who assisted in my preparation for this review. "Van Flandern has written columns in several astronomy magazines," he told me. "None of which has reviewed his book. I found that rather strange." If the Journal of Scientific Exploration reader finds this reception of Van Flandern's book strange, let me recommend Brian Martin's Suppression Stories as a primer on the structure of intellectual change. Martin (see Martin, 1998), like the late Thomas Kuhn, is a physicist "gone soft." But while Kuhn turned his analytical talents toward the obscure early histories of physical sciences, Martin examines a far more accessible current intellectual landscape. Of course, he finds "patterns of suppression" to be among a host of patterns the whistle-blower or intellectual revolutionary may expect to encounter, to include the threat of defamation suits in Australia, the author's home. Because Martin has homogenized, by the sheer breadth of his clearly explicated sample, the individuals and their various ideas, a certain dispassion pervades the book. I suppose this is in part a response to continuing (after Kuhn s) criticism of the social and anthropological sciences relative to physical sciences, and a successful one. Those ignorant of recurring patterns in history and society very probably have no hope of shaping them. Martin knows this, I think, and offers the intellectual explorer a reliable, methodical reconnoiter of potential pitfalls, ambushes, and outright attacks or, as often, ringing silences. Ignoring Van Flandern's broadside against the current, accretionary model is the stuff of the social construction of science! Certainly the fit of new data, or its misfit, should drive our hypotheses; yet, I found his epistemology vague 7 Book Reviews 485 in some areas. Van Flandern, unlike Martin, raises more questions than he answers, which can be a virtue in a book and useful in the search for novel implications in the stream of data from continuing exploration of the solar system. Both books have references and indices. We know an order of magnitude, if not several, more about our solar system than we did a century ago. Have our hypotheses kept pace? Van Flandern argues they have not. If one supposes they have, then one may enjoy Mason & Dixon more than Dark Matter, Missing Planets and New Comets. Or one might read Suppression Stories first and enjoy all three. Gordon Strasenburgh 2680 Everett Street, North Bend, OR 97459 References Castle, Paul R. (1994). Review, February. Rejlector, the Astronomical League Newsletter. Keay, Colin. (1993). Review, September. Australian and New Zealand Physicist. Martin, Brian. (1998). Strategies for Dissenting Scientists. Journal of Scientific Exploration, 12,605. Pierri, David C. (1999). Geomorphology of Selected Massifs on the Plains of Cydonia, Mars. Journal of Scientific Exploration, 13,401. Tompkins, Peter. (1971). Secrets of the Great Pyramid. New York: HarperRow. The Powerful Placebo-From Ancient Priest to Modern Physician by Arthur K. Shapiro and Elaine Shapiro. Baltimore, MD, and London: Johns Hopkins University Press, 1997. The Placebo Effect-An Interdisciplinary Exploration Edited by Anne Harrington. Cambridge, MA: Harvard University Press, 1997. Paperback, 1999. Placebo is something of an oddity in modern Western culture. This is a culture self-consciously taking science as authoritative, congratulating itself that its medical practice is based on science and condescending toward beliefs for which there is no concrete evidence. Psychic phenomena are decried and pooh-poohed. Yet the placebo phenomenon is fully acknowledged even as it seems to establish that nonmaterial influences-psychological if not "psychic '-can effect powerful material results. This incongruity comes even more sharply into focus when scientific medicine accepts psychosomatic illness as entirely real while dismissing faith healing as superstition: those are surely but different consequences of the same sort of interaction between psyche and body. Incongruous too is that the term "placebo" carries the negative connotation of "sham treatment" at the same time as its effects are beneficial to the recipient or experiencer of the placebo (~236).' 9 ' Page numbers in Shapiro are given with an "S" and in Harrington with an "H." 486 Book Reviews Though the placebo phenomenon has long been acknowledged, it has remained until recently little studied, let alone understood; only in 1990 was "placebo effect" included as a reference in the computer files of the National Library of Medicine (S 136). Salient questions about it remain conundrums: What range of illnesses or symptoms can placebo cure or suppress? With symptoms that placebo can remove, what is its rate of success? Is placebo more effective under some circumstances than under others; and if so, what are they? Are some people more likely than others to benefit from placebo? That list is far from exhaustive. The placebo phenomenon poses a fundamental question that philosophy has tangled with for millennia and that science has tiptoed around, namely "the mind-body problem": what are the relations between on the one hand the human mind, consciousness, thought, and on the other hand the electrical and chemical happenings in the brain that, we believe, are somehow associated with thoughts and feelings? During the 1990s, people from various disciplines began to discuss placebo and its implications. Two books offer intriguing glimpses into what has been learned and what large mysteries remain. The Placebo Effect-An Interdisciplinary Exploration lives up to its title, as does The Power$ul Placebo-From Ancient Priest to Modern Physician. The first reports the proceedings of a conference but, unlike all too many proceedings, is so very well edited that it makes an excellent book. The second reviews a lifetime of learning about placebo by A. K. Shapiro, one of the pioneers in its study; it is a labor of love, unfortunately marred by repetitiveness and jargon but fully rewarding the reader's willingness to stay with it. Together, these two books-with their copious references-reveal an interdisciplinary research community in the process of forming around a topic whose progressive understanding will profoundly affect our lives and thoughts. However, the subject is a collection of noteworthy facts and of enigmas to research more than a coherent body of knowledge. "It is a mystery how a ubiquitous treatment used since antiquity was unknown, unnamed, and unidentified until recently," Shapiro begins. "It is even more remarkable because this is the only treatment common to all societies and cultures." That treatment, of course, is placebo. For us, "placebo" typically means a sugar pill, or the inactive ingredient used in clinical trials to test the effectiveness of a new drug. Shapiro points out that it can also be magic: "religious rituals (such as prayer and exorcism); incantation.. .and cleansing the mind by talking out the evil (as in psychoanalysis and psychotherapy)." Every interaction between someone who is regarded as a healer and someone who desires to be healed may summon a placebo response. Shapiro, himself a psychiatrist, reveals that within his profession, the question is sometimes openly argued whether all psychotherapy may not owe Book Reviews 487 its effectiveness to the power of placebo (S47). That "no one therapy is better than another" (S100) tends to support that suggestion (S103, S105, S120, S121, H229, H230). The very attempt to define "placebo" is enlightening; and once the phenomenon is fully understood, the term itself will become obsolete (S42). It is important to recognize that "placebo therapy may be used with or without knowledge that it is a placebo." Drugs given at physiologically ineffective dosages are thereby placebos. A sometimes useful alternative term is "nonspecific": placebo presumably summons a mechanism capable of ameliorating a wide variety of symptoms, which stands in contrast to drugs or surgical procedures aimed specifically at a particular disorder. Scientific cognizance of placebo dates only from the late 1930s, when Gold introduced the "blind test" to control for placebo effects and Houston suggested that the history of medicine is largely that of the placebo effect. Shapiro's historical survey in the first chapter of his book amply justifies that suggestion: Such ventures as homeopathy and Christian science may be quackery so far as "modern" medicine is concerned, but they were unexceptionable in earlier times. It is chastening to be reminded that controlled clinical trials, double-blind procedures, and statistical analysis were almost nonexistent before the 1950s (S34, S96, chap. 7). It is startling to discover that such testing is far from being routinely reliable: "Inert placebos frequently are inadequate controls in double-blind studies" because both patients and doctors are often able to differentiate the real drug from the placebo. The real drug typically has a variety of discernible effects (S35, S93, chap. 9). Apart from that, it is by no means universally agreed that controlled double-blind trials are even desirable, humane, or a proper part of physicians caring for their patients (S 168f, S 1780. Without properly controlled trials, however, one cannot know whether a treatment is directly effective physiologically or whether its success draws on the placebo response. Since even nowadays properly double-blind studies are the exception, all beliefs about the efSicacy of treatments used in earlier years are based on anecdotal evidence (S60ff). From that angle, the cupping and leeching used in Western medicine up to the 19th century is entirely comparable to the treatments used by shamans and witch doctors: they worked some of the time, and when they did, it was through stimulating the placebo response. On the other hand, one should not exaggerate the ancient wisdom based on long experience that, some say, identified useful herbal and other remedies. Nowadays we know how delicate and complex the process of extracting the physiologically active ingredients is, and how much raw material needs to be processed to obtain a useful amount of drug; it seems unlikely that there ever was much, if indeed any, properly standardized use of natural remedies. Most folk medicine probably owes its successes to the placebo effect (S7 lff). Facile description of placebo as the effect of suggestibility founders on the fact that "there is little correlation between tests of suggestibility in a laborato- 488 Book Reviews ry situation and the placebo effect in a clinical setting" (S38). Some people react favorably to drug placebo but not to psychotherapeutic placebo, and with others it is the reverse (S40). The ability to summon the placebo effect seems to be universal, however. A striking corollary is that the beneficial effect of any material treatment, drug, or surgery is the summation of (nonspecific) placebo and specific "scientific" treatment (S41): "Most drugs are useful only in a range in which psychological factors can occur" (S77). In this day of "managed care," it is vital to recognize that "bedside manner" is not an optional bonus to effective treatment: personable or charismatic physicians may actually cure more patients than brusque scientific technicians who operate on assembly-line principles (H241): "The placebo is powerless without the physician" (H39, H40, H52, H88). It may be, for instance, that the simple action of scheduling follow-up appointments reinforces therapeutic interventions (H84). The same line of thought tells us that some charming well-intentioned quacks may benefit people by stimulating their placebo response even as wellintentioned mainstream practitioners may unwittingly kill their patients because the orthodox treatment is actually harmful. In historical times, that is illustrated by the greater success homeopaths had in treating cholera than did mainstream physicians: the latter's brutal conventional treatments of bloodletting and the like harmed the patients, whereas the "ineffective" homeopathic remedies at least did no harm (S59, S60). Even deeper questions with important practical consequences are raised. Given that the placebo response is almost always available, and that it is in most instances at least a useful adjunct to surgical or physiological treatment, what influence is exerted by the increasingly mandated "informed consent" patients must tender before treatment? To what extent is a patient's confidence in a treatment reduced by the enumeration of all the risks and side effects? To what extent does obtaining informed consent hinder a beneficial placebo effect (H225, H237)? Placebo is powerful beyond common knowledge, though common knowledge does include the apparent ability of many people to delay their death until such a significant occasion as a birthday or religious holiday has passed (H231, H232). The placebo can be as effective as standard physical treatments, for example, in the treatment of peptic ulcer or high levels of lipids in the blood (H41, H42). Moreover certain surgical procedures, it turns out, were effective only through stimulating a placebo response: ligation of internal mammary arteries as treatment for angina pectoris, for example (S83), or prefrontal lobotomy (S91). Indeed, "surgical procedures are probably more likely to be placebos than is drug treatment [! !I, which is subject to more rigorous criteria of controlled study" (S84); moreover, "controlled trials in surgerywhich may involve the use of invasive procedures, such as sham surgery as a control for transplants-are more difficult to devise and ethically more difficult to justify" (S85). Book Reviews 489 "Psychotherapy, whether provided by psychiatrists, social workers, nurse therapists, clerics, or other professionals, is the treatment most subject to placebo effects. Even more placebo prone are the myriad treatments encompassed by alternative medicine, faith healing, and quackery" (S85). The main focus of Shapiro's chapter 5 is "the possibility that psychotherapy is the major placebo of the late 20th century" (S86). Even medical psychiatry may work largely through placebo: drugs treat depression with an effectiveness of between 45% and 80%, but placebo works in 30% to 50% of cases, making it at about two thirds as effective as the drugs (S93)-and typically more desirable because of the fewer harmful side effects. New Age or religious enthusiasts should not be tempted to cite the powerful placebo as proof of some spiritual or supernatural human essence, however: creatures other than humans exhibit the placebo response (H5, H6)-a finding that will seem surprising only to people who do not handle domesticated animals or who have never owned pets. Furthermore, the placebo effect is not invariably benign or desirable: placebo can induce side effects (S79f) and even be toxic (S154)-discussed in the chapter "The Nocebo Phenomenon" (H56-H76). In one study, misinformation about a drug's action reduced its effectiveness by half (H67); in another instance, a patient's rebellion against doctor and treatment made a drug much less effective (H228, H229). Some surgeons believe they should not operate on patients who are convinced they will die during the operation (H62); medical students, it is well known, often take on symptoms of the diseases about which they are learning (H61); there have been epidemics of hysterical illness. (H63-64) "Voodoo death" by witchcraft in Africa and Haiti has been reported, and by "bone pointing" among Australian aborigines (H65-66). How then are we to understand the placebo phenomenon? Not through medical science alone, not through psychology alone, not through any one intellectual discipline alone. But interdisciplinary thought is fraught with "methodological, epistemological, even metaphysical tensions involved in the attempt to arrive at an understanding greater than the sum of its parts," Harrington points out (H7, H8), particularly "when one attempts to move between the divide that separates the explanatory strategies of the so-called cultural or hermeneutic sciences (roughly concerned with 'meaning') and the so-called natural sciences (roughly committed to explanations in terms of 'mechanism')." The first chapter in the book edited by Harrington is by the Shapiros; it is in a sense a synopsis of their own book (which was published later). A salutary warning is given: "When a treatment is a placebo, it is very easy to counterfeit," and since (as the Shapiros show in detail) much conventional medical practice was the (unknowing) administration of placebo (though less so in more recent times), one can understand why quacks flourish and unconventional treatments can often claim comparable success rates with mainstream practice. Summarizing the few studies that have been made of the placebo / 490 Book Reviews phenomenon, the Shapiros raise an intriguing question: because placebo response helps us heal ourselves (or at least withstand illness and pain), has evolution selected in favor of those who can summon it most powerfully? Howard Spiro points out how little is known about placebo except that it works. Its use in clinical trials may mask important phenomena (H42, H43). The application of piece-meal information from single studies can cause clinical practice to swing pendulum-like, as in the prescription of milk for amelioration of ulcer symptoms, to its being regarded as bad for ulcer patients, to again being recommended-all without knowing whether it works as an antacid or as a placebo (H48). Robert Ader (chap. 7 in Harrington) considers evidence that the placebo response may be at least in part an instance of conditioned response. The variety of contributions in this edited book, from people versed in disparate disciplines, underscores what a challenge to our understanding is posed by the placebo phenomenon. It is not unlike the riddle of claimed psychic phenomena and of issues central to attempted "scientific" studies of human beings: one wants to measure independently instances and degrees of the phenomenon's occurrence on the one hand and on the other those factors one suspects to be causative of it. However, one cannot separately measure those suspected factors; in many instances, one has to rely on self-reporting as measures of both the various factors and the placebo effect itself. Placebo, it seems natural to believe, should be more effective for people who can summon high degrees of expectation, hope, belief in the efficacy of the treatment, and confidence in their physician or healer; but those can hardly be assessed other than by self-reporting. And moreover, when they are measured in some nonclinical situation, one cannot know whether they will be manifest in the same way in a clinical setting. Even further, studies of placebo often pose ethical issues (H237ff), all the more so with nocebo phenomena: would one want to design and carry out a controlled double-blind study of the effectiveness of bone pointing in killing Australian aborigines? Placebo or nocebo effects are inferred when-and only when-current medical knowledge offers no biological or physiological reason for what was observed to happen. But this is fallible because current knowledge is not final knowledge, and moreover, it presumes that placebo/nocebo effects do not occur with physiological ones, a most unlikely contingency (H68, H69). The problem is illustrated by the commonly made notional differentiation between placebo response and "spontaneous remission" (H167), which some might judge are actually the same thing. Thus, placebo brings into focus fundamental issues about human existence, in particular the Cartesian view of mind-body relations (H187). Peptides, it seems, are involved in placebo responses; and because the same peptides occur throughout the body as those in the brain, the possibility is raised that the mind "resides" throughout the body and not just in association with the brain (H85). Perhaps some such train of thought points to that reconciliation of Western and Book Reviews 491 Eastern medicine and thought that has so often been suggested (H233, H236). That and other salient issues raised by the placebo phenomenon-ethical, philosophical, and practical-are discussed in the very well edited "conversation" that forms the final chapter of Harrington's book. It ought to be required reading for everyone involved in medical science, clinical care, and the management side of the health business. Henry H. Bauer Virginia Polytechnic Institute and State University [email protected] Synchronicity, Science, and Soul-Making by Victor Mansfield. Chicago: Open Court Publishing, 1995. ISBN 0-8126-9304-3. Dr. Victor Mansfield's Synchronicity, Science, and Soul-Making is an attempt to unify Jungian psychology, modern physics, and Buddhist philosophy into a single construct. The book is split into three parts: an introduction to Jungian philosophy, a section exploring the connections with this philosophy to physics, and a comparison of Jungian philosophy to Buddhism, with the author's attempt to develop a philosophical model for synchronicity. It is an excellent book for people who are either looking for information on Jungian psychology or are already knowledgeable on the subject and would like to read an interesting discussion of uniting Jungian philosophy and Eastern philosophy. The first section does a very good job of discussing and explaining the rather difficult topic of Jungian philosophy and synchronicity. Synchronicity defined is an "acausal connection through meaning." Consider this example often used: A woman is psychologically inaccessible because of her belief in strict rational explanations of the world around her. One day as she is talking with Dr. Jung about a dream she had the night before about a scarab beetle, what should fly in the window at that exact time is a beetle whose color is quite close to that of a scarab. This punctured a hole in her rationalism and allowed her to continue her therapy. In this example, the woman talking about her dream and the appearance of the beetle would be considered acausal because they were completely separate events. What makes it an example of synchronicity is that the event had a special meaning to her and consequently allowed her to continue her spiritual progression. The author gives many other examples of synchronicity experiences throughout the book. This is quite useful, as synchronicity is a rather difficult topic to talk about without examples. After discussing which examples would involve synchronicity, he discusses which topics would not involve synchronicity. This is where the subject of parapsychology comes up and is also the first of several instances where he separates his interpretation of the nature of synchronicity from the strict Jungian conception. Dr. Mansfield asserts that 492 Book Reviews even though parapsychology deals with acausal events, such as ESP, it is the lack of meaning surrounding a particular occurrence that prevents it from being synchronicity. In fact, unlike parapsychology, the author asserts that it is virtually impossible to scientifically verify synchronicity because of the necessity of meaning in the occurrence. An observer could verify that an event occurred, but not the possible meaning to another observer. But what supplies this meaning? The author goes into detail about this, describing how the ego, superego, and the unconscious interrelate to create and interpret the meaning in a synchronicity experience. At this point, the author shifts gears and discusses how physics can be brought to bear on this subject to perhaps gain some insight. He discusses the history of science by comparing Dante with Galileo and contrasting their world views. For instance, in Dante's world, the world was permeated by God's intelligence, and thus acausal orderliness was the rule, not the exception. But once the Galilean mode of analysis begins to dominate scientific thought, the inherent reality of a subjective principle is denied. It is this "pendulum swing" that he asserts today has caused surge of the New Age movement as an unconscious compensation, or attempt of the unconscious to influence the ego through synchronicity to progress the process of individuation. The author asserts that synchronicity forces us to acknowledge the inherent interconnectedness of matter and psyche. To reinforce this argument, Dr. Mansfield turns to relativity, electromagnetism, and quantum mechanics to demonstrate how the observer is interconnected in a basic way with the world. As an example of this, he gives a detailed explanation of Bell's paradox and how it shows the nonlocality and acausal nature of physics. Although he did show physical examples in which nature is apparently acausal at times, his attempt to tie it back to synchronicity as an explanation was found a bit wanting. The last part of the book is a description of Buddhist thought (Middle Way Tibetan Buddhism to be specific) and its relationship to synchronicity and Jungian philosophy. This was one of the most intriguing parts of the book. Although his descriptions and explanations of Buddhist philosophy are excellent, his attempts to tie it back to synchronicity tend to be a bit problematic. The author gives a good explanation of the Doctrine of Emptiness (the doctrine that denies inherent existence or independent existence to all persons and things) and uses it to show support of synchronicity in Eastern thought through the belief that nothing is independent. Dr. Mansfield also gives an excellent overview of the Buddhist notion of samskara, the notion that everything we perceive is filtered by our mind state. This is another example of where the author differentiates himself from Jung, taking issue with Jung's dislike of meditation, his premise that the nature of mind is unknowable (in effect denying the existence of Buddhas), and that the entire notion of transcending the psyche is an archetypal theme and unattainable. But it is here that the author leaves the Buddhist philosophical arena and Book Reviews 493 ventures out on his own. The author asserts that because we can all agree on certain objective observations, the notion of a "World Soul," which transmits the world to us filtered by our perceptions must exist. He states that archetypes are in fact similar to an "unmeasured wave function" that provides a sort of psychic probability. But unfortunately, the author does not explore further how the existence of a World Soul can be reconciled with Buddhist thought (or synchronicity itself for that matter-the basis of synchronicity being that acausal events occur that guide the ego to its further individuation, an individuation whose existence is generally denied by Buddhist philosophy). Now faced with two apparently conflicting paths to self-knowledge, how does the author suggest the reader proceed? It is on page 217 that the reader gets the full story: Learning to concentrate the mind usually has a calming effect. However, there are inevitably periods when attempting to go deeper is like flushing skunks out of their dens.. .. My experience is that more meditation is not necessarily curative of the problems generated by meditation; however here [Jungian philosophy] can truly come to the rescue. Perhaps this is a valid reason for a Buddhist, or anyone else for that matter, to further study synchronicity. Craig Clayton Shaolin Gung Fu Institute I? 0. Box 95745, Seattle, WA 98145 craig.clayton @shaolin.com Journal of Scientific Exploration, Vol. 14, No. 3, p. 495,2000 0892-33 10/00 O 2000 Society for Scientific Exploration SSE 20th Annual Meeting Society for Scientific Exploration June 7-9,2001 La Jolla Radisson, La JollaISan Diego, California Program Chair: Peter Sturrock Program Invited Speakers TBA Events June 7-9,2001 Scientific Meeting at the La Jolla Radisson Conference Center, across the road from the University of San Diego Wednesday., June 6, 7:30p.m Welcome Reception, private party in the Outdoor Grove, La Jolla Radisson Friday, June 9, 2-6p.m. Field Trip, Private boat ride in San Diego Harbor (including bus ride to and from San Diego) Saturday, June 10,8p.m. Society Banquet, La Jolla Radisson SSE has reserved rooms at the La Jolla Radisson at $11 9lnight until May 4. There is a free shuttle from the San Diego airport to the hotel, which runs hourly. Society for Scientific Exploration Officers Prof. Peter A. Sturrock, President Varian 302 Stanford University Stanford, CA 94305-4060 Ms. Brenda Dunne Executive Vice President for Education C 131, School of Engineering &Applied Science, Princeton University, Princeton, NJ 08544-5263 Prof. Robert Jahn, Vice President D334, School of Engineering & Applied Science Princeton University Princeton, NJ 08544-5263 Prof. L. W. Fredrick, Secretary Department of Astronomy P. 0 . Box 3818 University of Virginia Charlottesville, VA 22903-08 18 Prof. Charles R. Tolbert, Treasurer Department of Astronomy P. 0 . Box 3818 University of Virginia Charlottesville, VA 22903-08 18 Council Dr. Marsha Adams 1100 Bear Gulch Rd. Woodside, CA 94062 Prof. Henry H. Bauer Chemistry, 234 Lane Hall VPI & su Blacksburg, VA 24061-0247 Dr. Stephan Baumann fMRI Project Director Psychology Software Tools, Inc. 2050 Ardmore Blvd. Pittsburgh, PA 15221 Prof. John Bockris Department of Chemistry Texas A&M University College Station, TX 77843 Dr. John S. Derr Albq. Seismology Center Albuquerque, NM 87 1 15 Dr. Roger D. Nelson C 131, Engineering Quad. Princeton University Princeton. NJ 08544 Dr. Harold E. Puthoff Institute for Advanced Studies-Austin 4030 W. Braker Ln., Suite 300 Austin, TX 78759-5329 Dr. Beverly Rubik Institute for Frontier Science 6 1 14 LaSalle Avenue, #605 Oakland, CA 9461 1 Dr. Robert M. Wood 1727 Candlestick Ln. Newport Beach, CA 92660 Journal of Scientific Exploration, Vol. 14, No. 4, p. 497,2000 0892-3310100 0 2000 Society for Scientific Exploration EDITORIAL Studies of scientific anomalies are commonly criticized for lack of efforts at reproducing results, for not following scientific method, and for making no progress. None of those apply to the report in this issue from the MindIMachine Interaction Consortium. Three independent laboratories undertook to replicate salient features of some of the data accumulated over twenty years by the Princeton Engineering Anomalies Research group. Analysis of the results meets the highest standards of incisive and sophisticated scientific work: Monte Carlo technique has been applied in a manner that much mainstream work could with advantage emulate. (Its rigor puts to shame, for example, a book I have just reviewed for a journal of higher education, a book by sociologists at good colleges and published by one of the most widely respected academic presses.) Nor is this without precedent: the best work in parapsychology has for some time been as rigorous methodologically as the best work in mainstream psychology. As to progress, this replication study is a landmark, quite irrespective of how its results are interpreted. "Skeptics" of the dogmatic ilk will happily, immediately, and superficially misconstrue its significance by claiming that it confirms the inherent irreproducibility of mindlmachine interactions. Seekers of truth, on the other hand, will recognize that the work points to and opens up new ways of understanding what research in this field must aim to do: find ways of getting conclusive results even though the phenomenon displays itself so capriciously that the variables to be identified and controlled must be other than the traditional, obvious ones. That marks real progress. John Wilson, the finest academic administrator I have known, occasionally commented on the reaction whenever he publicly praised a faculty member, department, or college: some of those not mentioned would ask him in hurt tone, did he not appreciate them too? Well of course he did, just as I appreciate every piece that our Journal publishes even if I am moved to comment only on some of them. What we publish, after all, is less than half of what is submitted, and has run the gauntlet of careful review by knowledgeable referees. Henry Bauer Journal of Scientgic Exploration, Vol. 14, N O . 4, pp. 499-555,2000 0892-3310100 O 2000 Society for Scientific Exploration MindIMachine Interaction Consortium: PortREG Replication Experiments R. JAHN, B. DUNNE, G. BRADISH, DOBYNS, A. LETTIERI, AND R. NELSON Y. Princeton Engineering Anomalies Research Princeton University, Princeton, NJ J. MISCHO, E. BOLLER, AND H. BOSCH Freiburg Anomalous Mind/Machine Interactions Znstitutfur Grenzgebiete der Psychologie und Psychohygiene e. V , Freiburg, Germany D. VAITL, J. HOUTKOOPER, AND B. WALTER Giessen Anomalies Research Project Justus-Liebig-UniversitatGiessen, Giessen, Germany Abstract-A consortium of research groups at Freiburg, Giessen, and Princeton was formed in 1996 to pursue multidisciplinary studies of mindlmachine interaction anomalies. The first collaborative project undertaken was an attempted replication of prior Princeton experiments that had demonstrated anomalous deviations of the outputs of electronic random event generators in correlation with prestated intentions of human operators. For this replication, each of the three participating laboratories collected data from 250 x 3000-trial x 200 binary-sample experimental sessions, generated by 227 human operators. Identical noise-source equipment was used throughout, and essentially similar protocols and data analysis procedures were followed. Data were binned in terms of operator intention to increase the mean of the 200-binary-sample distributions (HI); to decrease the mean (LO); or not to attempt any influence (BL). Contiguous unattended calibrations were carried forward throughout. The agreed upon primary criterion for the anomalous effect was the magnitude of the HI-LO data separation, but data also were collected on a number of secondary correlates. The primary result of this replication effort was that whereas the overall HI-LO mean separations proceeded in the intended direction at all three laboratories, the overall sizes of these deviations failed by an order of magnitude to attain that of the prior experiments, or to achieve any persuasive level of statistical significance. However, various portions of the data displayed a substantial number of interior structural anomalies in such features as a reduction in trial-level standard deviations; irregular series-position patterns; and differential dependencies on various secondary parameters, such as feedback type or experimental run length, to a composite extent well beyond chance expectation. The change from the systematic, intention-correlated mean shifts found in the prior studies, to this polyglot pattern of structural distortions, testifies to inadequate understanding of the basic phenomena involved and suggests a need for more sophisticated experiments and theoretical models for their further elucidation. Keywords: random event generator (REG) -random event experiments humanlmachine anomalies - mindlmachine interactions 500 R. Jahn et a1. Table of Contents Page ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499 ListofTables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .500 List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502 I. Context and Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502 A. History and Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 502 B. Prior PEAR Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503 1. Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503 2. Experimental Design and Results . . . . . . . . . . . . . . . . . . . . . . 504 I1. Consortium Replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 A. Experimental Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505 B. Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 5 0 6 1. Tabular Key and Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . 506 2. Primary Data Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508 3. Structural Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 508 I11. Structural Analyses and Their Interpretation . . . . . . . . . . . . . . . . . . . . 514 A. Primary Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 5 1 4 B. Structural Anomalies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520 1. Structural Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520 2. Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524 3. Series-Position Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532 4. Operator-Specific Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533 5. Standard Deviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .536 6. Counts of Successful Operators and Series . . . . . . . . . . . . . . . 537 IV. Summary Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 Appendix I: PortREG Equipment Calibrations . . . . . . . . . . . . . . . . . . . . . 546 Appendix 11: Stuctural Meta-Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 553 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .553 List of Tables MAIN TEXT Table 0: Table 00: Table F. 1: Table G. 1: Table P. 1: Prior PEAR Data (522 Series. 91 Operators) . . . . . . . . . . . . . 508 Concurrent Calibrations (1049 Series). . . . . . . . . . . . . . . . . . 508 All FAMMI Data (250 Series. 80 Operators) . . . . . . . . . . . . 508 All GARP Data (250 Series. 69 Operators) . . . . . . . . . . . . . . 509 All PEAR Data (250 Series. 78 Operators) . . . . . . . . . . . . . . 509 MindIMachine Interaction Consortium 501 Table C.1: Table F.2: Table G.2: Table P.2: Table C.2: Table F.3: Table G.3: Table P.3: Table C.3: Table F.4: Table G.4: Table P.4: Table C.4: Table F.5: Table G.5: Table P.5: Table C.5: Table F.6: Table G.6: Table P.6: Table C.6: Table F.7: Table G.7: Table G.8: Table M.l: Table M.2: Table C.7: Table C.8: Table C.9: Table P.7: Table C.10: Table C.11: Page Concatenation Across All Laboratories (750 Series. 227 Operators) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509 Gender Effects in FAMMI Data . . . . . . . . . . . . . . . . . . . . . . . 510 Gender Effects in GARP Data . . . . . . . . . . . . . . . . . . . . . . . . . 510 Gender Effects in PEAR Data . . . . . . . . . . . . . . . . . . . . . . . . . 511 Gender Differences in Concatenated Data . . . . . . . . . . . . . . . 511 Assignment Effects in FAMMI Data . . . . . . . . . . . . . . . . . . . 512 Assignment Effects in GARP Data . . . . . . . . . . . . . . . . . . . . . 512 Assignment Effects in PEAR Data . . . . . . . . . . . . . . . . . . . . . 513 Assignment Effects in Concatenated Data . . . . . . . . . . . . . . 513 Feedback Effects in FAMMI Data . . . . . . . . . . . . . . . . . . . . . 514 Feedback Effects in GARP Data . . . . . . . . . . . . . . . . . . . . . . . 514 Feedback Effects in PEAR Data . . . . . . . . . . . . . . . . . . . . . . . 515 Feedback Effects in Concatenated Data . . . . . . . . . . . . . . . . . 515 Runlength Effects in FAMMI Data . . . . . . . . . . . . . . . . . . . . 516 Runlength Effects in GARP Data . . . . . . . . . . . . . . . . . . . . . . 516 Runlength Effects in PEAR Data . . . . . . . . . . . . . . . . . . . . . . 517 Runlength Effects in Concatenated Data . . . . . . . . . . . . . . . . 517 Series-Position 2-Scores in FAMMI Data . . . . . . . . . . . . . . .518 Series-Position 2-Scores in GARP Data . . . . . . . . . . . . . . . . 518 Series-Position 2-Scores in PEAR Data . . . . . . . . . . . . . . . . . 518 Series-Position 2-Scores in Concatenated Data . . . . . . . . . . 518 Experimenter Effects in FAMMI Data . . . . . . . . . . . . . . . . . .519 Control Mode Effects in GARP Data . . . . . . . . . . . . . . . . . . .519 Effects by GARP Operator Types . . . . . . . . . . . . . . . . . . . . . . 520 Comparison of All Laboratory Data with 5000 Monte Carlo Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527 Most Prominent 2-Score Differences from Monte Carlo Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532 2-Scores in Secondary Parameter Cells, by Laboratory . . . . 533 Difference 2-Scores of Unconfounded Secondary Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - 5 3 7 x2Tests for Series-Position 2-Scores . . . . . . . . . . . . . . . . . . . 537 Consistency of Operators Between Prior PEAR and Replication Experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 Operator Performance Values (with Associated Probabilities) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539 2-Scores for Trial-Level Standard Deviations, by Laboratory and Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540 x2 APPENDIX I Table A1.F. FAMMI Concurrent Calibrations (852. 000 Trials) . . . . . . . . 549 Table A1. G. GARP Concurrent Calibrations (1.165. 000 Trials) . . . . . . . 549 Table A1. P. PEAR Concurrent Calibrations (1.130. 000 Trials) . . . . . . . . 549 502 R. Jahn et al. Page APPENDIX I1 Table A2.1: Summary of Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .552 List of Figures Figure 1: Figure 2: Figure 3: Figure 4: Figure 5: Figure 6: Figure 7: Figure 7a: Figure 8: Figure 9: Figure 10: Figure 1 1: Figure 12: FAMMI Cumulative Deviations . . . . . . . . . . . . . . . . . . . . . . . . . 5 2 1 GARP Cumulative Deviations ......................... .521 PEAR Cumulative Deviations . . . . . . . . . . . . . . . . . . . . . . . . . .522 Prior PEAR Cumulative Deviations . . . . . . . . . . . . . . . . . . . . . .522 Cumulative HI-LO Differences for All Three Labs . . . . . . . . .523 Mean-Shift Z-Scores versus Monte Carlo Populations . . . . . . .529 Difference Z-Scores versus Monte Carlo Populations . . . . . . .530 Composite Statistic for Difference Z versus Monte Carlo . . . . 531 FAMMI Data Split by Assignment (I,V), Feedback (G,N), and Run Length (H,T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .534 GARP Data Split by Assignment (I,V), Feedback (G,N), and Run Length (H,T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .534 PEAR Data Split by Assignment (I,V), Feedback (G,N), and Run Length (H,T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .535 All Data Split by Assignment (I,V), Feedback (G,N), and Run Length (H,T) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .535 Prior PEAR Cumulative Deviations in Three Epochs . . . . . . .545 I. Context and Background A. History and Organization Electronic random event generators (REGS) have long been used in a wide range of laboratory experiments designed to test the hypothesis that human consciousness may interact directly with random physical systems (Radin & Nelson, 1989; Schmidt, 1970). The results have provided strong statistical evidence that the mean outputs of these devices can deviate from chance expectation in direct correlation with prestated intentions of the participants and that aberrations in various other features of the output count distributions may reflect subtler aspects of the humanlmachine interactions. Over the past two decades, the Princeton Engineering Anomalies Research Laboratory (PEAR) has produced very large databases in REG experiments of this class (Nelson, Bradish, & Dobyns, 1989), which have further confirmed the existence of these types of humanlmachine anomalies, and have indicated some of their physical and psychological characteristics (Jahn, Dobyns, & Dunne, 1991; Jahn et al., 1997). While these PEAR experiments have constituted extensive conceptual replications of earlier work elsewhere (Bierrnan & Houtkooper, 1975; Radin & Nelson, 1989; Rhine & Humphrey, 1944; Schmidt, Morris, & Rudolph, 1986) and also have included many internal replications within themselves, it was MindIMachine Interaction Consortium 503 felt that more might be learned from further, more broadly based studies of similar character and comparable controls, conducted in collaboration with other researchers having complementary professional interests and experience. For this purpose, a consortium of laboratories was assembled in 1996, comprising the Freiburg Anomalous MindIMachine Interactions group (FAMMT) at the Institut fiir Grenzgebiete der Psychologie und Psychohygiene (IGPP) in Freiburg, the Giessen Anomalies Research Project (GARP) in the Center for Psychobiology and Behavioral Medicine at Justus-Liebig-Universitat Giessen, and the PEAR Laboratory at Princeton University. The primary agenda of this "MindIMachine Interaction Consortium" was a program of professional interaction and shared technology that would broaden and deepen our collective understanding of these consciousness-related anomalous phenomena. As an initial effort to establish sound and effective strategies for long-term collaboration, it was agreed that the first project to be addressed would be an extensive, commensurate repetition of prior PEAR REG experiments, conducted contemporaneously in all three locations. The first phase of this project was to be as strict a replication as feasible, given the essential differences of structure and style of the three laboratories. At the same time, it was to provide a platform for developing and deploying effective shared technologies, protocols, database acquisition and management techniques, and interlaboratory and interpersonal communications that would enable productive longer-term collaborations. A second phase of the project also was planned that would accommodate the three laboratories' specialized interests and capabilities in psychological, psychophysiological, and engineering investigations, respectively, but this article shall deal only with Phase I. B. Prior PEAR Experience 1. Equipment. Over its many years of mindlmachine experimentation, the PEAR program has developed several versions of electronic random event generators, utilizing different primary sources of noise but maintaining important common features of design. An original "benchmark" experiment employed a commercial random source sold by Elgenco, Inc. The core of this module is proprietary, but Elgenco's engineering staff describe it as "solid state junctions with precision preamplifiers," implying processes that rely on quantum tunneling to produce unpredictable, broad-spectrum noise in the form of low-amplitude voltage fluctuations. A much simpler and more compact REG, termed "PortREG," was developed subsequently, based on thermal noise in resistors, which also produces a well-behaved, broad-spectrum fluctuation. A yet later-generation device, called "MicroREG," uses a field effect transistor for the primary noise source, again relying on quantum tunneling to provide uncorrelated fundamental events that compound to an unpredictable voltage fluctuation. In all cases, the electronic process begins with a white-noise frequency dis- 504 R. Jahn et al. tribution. For example, the benchmark REG, on which most of the prior data were acquired, presents a flat spectrum, +I- 1 dB, from 50 Hz to 20 kHz. A subsequent 1000-Hz low-end cutoff attenuates frequencies below the data-sampling rate. This filtering, followed by appropriate amplification and clipping, produces an approximately rectangular wave train with unpredictable temporal spacing. Gated sampling, typically at 1-kHz, then yields a regularly spaced sequence of randomly alternating +I- bits, suitable for rapid counting. To eliminate biases from such environmental stresses as temperature change or component aging, "exclusive or" (XOR) masks are applied to the digital data streams in regularly alternating +I- patterns. In the experiments, output data are presented and recorded in "trials" that are the sum of N samples (typically 200 bits) from the primary sequence, thus mitigating any residual short-lag autocorrelations. The final output of the benchmark REG thus is a sequence of conditioned bits and, in the later devices, of bytes, presented to the computer's serial port, which then are collected into a sequence of trials, usually presented at approximately one trial per second. Calibrations of all of the devices conform to statistical chance expectations for the mean, standard deviation, skewness, and kurtosis of the accumulated trial-score distributions, and for time-series of independent events (cf. Appendix I). 2. Experimental design and results. The basic experimental designs embody further protocol-level protections against artifacts. Using a "tripolar" protocol, participants generate data under three conditions of prespecified intention, namely to achieve high (HI) or low (LO) output distribution mean values, or to generate baseline (BL) data. With the exception of these expressed intentions, which are immutably prerecorded in the experiments computer files, all other potentially influential protocol variables are maintained constant within an experimental session. In addition to the primary variable of tripolar intention, a number of secondary parameters are available as options that can be explored in separate sessions and assessed as factors that may contribute to the experimental outcomes. These include human variables, such as the identities of the individual operators, their gender, the number co-operating in the effort, and whether they are "prolific," i.e., have accumulated sufficient data to permit robust internal comparisons of their results; technical variables, such as the different noise sources, including not only the physical random sources described but also various hardwired and algorithmic pseudorandom generators, designated as nondeterministic and deterministic sources, respectively; operational variables, including information density (bits per second); the number of trials in automatically sequenced "runs;" the instruction mode (volitional or instructed); the type of feedback provided to the operator, etc.; and physical variables, including the spatial separation of the operator from the machine (up to thousands of miles) and temporal separations between operator attempts and actual operation of the devices (up to several hours or even a few days). For the purposes of the replication studies reported here, we shall refer mainly to that segment of previous PEAR data provided by individual opera7 MindIMachine Interaction Consortium 505 tors adjacent to "benchmark REG equipment. These "local, single-operator" experiments, contributed over 12 years by 9 1 participants, constituted 522 replications at the "series" or "session" level, comprising nearly two and a half million, 200-bit trials. The primary results of this segment are summarized as "Prior PEAR Data" in Table 0. This database also was subjected to a broad range of subordinate analytical tests, including specific searches for indicative structural details and broad-based analyses of variance, all of which have been extensively reported in the archival literature and supporting technical reports (Jahn et al., 1997; Dunne, 1991 ; Dunne et al., 1994; Nelson et al., 2000) and will be reviewed as appropriate in the following text. In passing, it might be noted that these particular experiments were complemented by an array of studies that used many other forms of random generator equipment and protocols (Jahn, Dunne, & Nelson, 1987), including the much more compact "PortREG" devices chosen for the replication program to follow, several macroscopic mechanical analogs (Dunne, Nelson, & Jahn, 1988; Nelson et al., 1994), various pseudorandom devices (Jahn et al., 1997), "remote" and "off-time" protocols (Dunne & Jahn, 1992), and nonintentional "FieldREG" experiments (Nelson et al., 1996, 1998). Results of these studies were generally consistent and collectively extended the statistical significance of the entire program by several orders of magnitude (Jahn et al., 1997). Many humanlmachine experiments of this sort have been conducted at other laboratories, and most of these have yielded commensurate anomalous results (Radin, 1997). Related studies have also demonstrated responses from biological substances or living organisms employed as the random targets of the operators' intentions (Braud, 1993; Braud & Dennis, 1989; Grad, 1963). In some cases, the role of the operators has been played by other than human species, e.g., by chicks, rabbits, and mice, many of whom seem capable of eliciting anomalous correlations of machine behavior with their biological or emotional needs (Peoc'h, 1995). From this array of empirical studies, it appears that operator desire is capable of establishing observable relationships to the outputs of such random physical systems, by some unknown means that is largely independent of the nature of the device and also independent of the intervening distance and time. The ubiquitous character of these anomalies bespeaks broad potential importance to contemporary scientific understanding and to individual and cultural welfare. I 11. Consortium Replication A. Experimental Design At a planning meeting held shortly after the inception of the MindIMachine Consortium, the members decided to undertake yet another replication of this class of REG experiments. Second-generation PortREG technology was selected for the random source because of its simplicity, portability, and relatively low cost, with confidence of its efficacy based on various indications from 506 R. Jahn et al. preceding PEAR research and that of others that these anomalous effects are independent of the source of randomness (Jahn et al., 1997; Schmidt & Pantas, 1972). All three laboratories would employ identical protocols and data-processing techniques, to the extent feasible given the differing languages, disciplinary backgrounds, and skills. Although the primary hypothesis to be tested was confirmation of the earlier PEAR results on a simple HI-LO mean-shift criterion, secondary investigations were to provide structural data on the characteristics and correlates of the phenomena. Specifically, it was agreed that each laboratory would use large pools of operators to accumulate 250 experimental "sessions" or "series," each series consisting of 1000 x 200-sample trials in each of the HI, LO, and BL intentions and, in addition, would extract whatever structural aspects of the data befit its capabilities, such as separate HI, LO, and BL performances, gender effects, serial position effects, standard deviations, feedback correlations, experimenter effects, etc. B. Experimental Results This section presents all of the pertinent data generated by the three laboratories in as commensurate and complete a format as possible here. We begin with a table key and brief explanatory text regarding the tabular formats. Then follows a sequence of tables that summarize the overall results of each laboratory with respect to the primary HI-LO mean-shift hypothesis, followed by a concatenation of all three databases. For comparison, these tables are preceded by similar representations of the earlier PEAR data and of the contemporaneous calibration data described in more detail in Appendix I. Following these summary tables, we then display a large array of explorations into data distribution structures and secondary parameter correlations attempted by each laboratory, both individually and collectively. 1 . Tabular key and comments. All of the following tables use a common statistical notation: Ap = o = Z = Shift in empirical trial-level mean from chance expectation of 100 (also called "effect size") Empirical trial-level standard deviation Standardized 2-score of the mean-shift, calculated as: Zdg = 2= where N , = Number of experimental trials o,, Theoretical chance standard deviation for 200-sample trials = (7.07 1) Z-score for differences of any two indicated data subsets (see text below) Chi-squared statistic to test for goodness-of-fit of empirical data values to a comparison standard MindIMachine Interaction Consortium 507 The first set of tables presents the overall results of the entire PortREGreplication database for all three intentions and for the HI-LO, or A, criterion which is regarded as the primary variable. The mean shift for the A column is calculated by inverting the LO data with respect to the mean and concatenating them with the HI, so it is the average, rather than the sum, of the mean shifts in the intended directions. This is intended to make statistical comparisons easier by preventing intrinsic differences of scale between the separate intentions and the A values. This representation makes the A data effectively a single large pool of trials which also have a theoretically expected mean of 100 and standard deviation of (50)'12 = 7.07 1, and in which a positive mean shift corresponds to success in the direction of intention. For all of the replication series, N, comprises 1000 trials per intention. (The earliest prior PEAR data were taken in larger series of 5000 trials per intention; the series size subsequently was reduced in stages to a standard of 1000 trials per intention, which allowed the experiment to be completed in a single session. Thus, the total number of prior PEAR trials is considerably greater than one would infer from the counts of series listed in the table; e.g., the prior PEAR database is approximately equivalent to 834 PortREG series.) For the sequence of structural tables that follow, Zdgrefers to the differences in mean shifts, computed as follows: Given two populations, N, and N,, having 2-scores 2, and Z,, we may compute a normalized effect size for each as ei= z,~(N,)"~,which is related to Ap by a multiplicative constant, e.g., ei = ~p~l(50)'~~. The uncertainty associated with a 2-score is always 1 by construction, so the E, have measurement uncertainties oi= 1 / ( ~ , ) ' /The standard nor~. mal deviate, or 2-score, for a difference between sets 1 and 2 is therefore the , difference E, - E, divided by the uncertainty of this difference, o ,which is 2 simply the sum in quadrature of the individual uncertainties: oj = o : %. This can be reduced to an expression in the original Ns and 2s: + For most of these presentations and the associated discussions, we have chosen to use only 2-scores without associated tail-probability (p) values, on the grounds that the former are completely unambiguous, depending only on the statistical character of the data used, whereas p-values require subjective and occasionally contentious decisions regarding the appropriateness of one- or two-tailed statistics, primary or secondary analyses, Bonferroni corrections for multiple or prospective analyses, and so forth. The direct correspondence of the Z values to particular "tail-probabilities" is, of course, well tabulated, e.g.: 508 R. Jahn et al. 2. Primary data summary. The mean-shift results and their associated standard deviations, obtained by each of the laboratories in each direction of intention, are summarized in Tables 0,00, F. 1, G.1, P.1, and C.1. It is immediately clear from these summary data that although the mean HI-LO separations found by each of the laboratories all proceed in the intended direction, they fail by an order of magnitude to reach the level of the prior PEAR data or any persuasive level of statistical significance. The implications of this result are discussed at length in Section 111. Nonetheless, subtler structural anomalies, such as the almost universal depression of the trial-level standard deviations below the theoretical and calibration values, are already evident in the summary tables above, and invite more detailed searches for other secondary correlates. The following tables display the results of such examinations (cf. section 1II.B.5). 3. Structural data. ( a ) Gender efSects. In the following sequence of tables, we display breakdowns of the laboratory data in terms of various subordinate secondary parameters that proved instructive in the prior PEAR studies. As a first example, the distinctions between male and female operator performance that were studied extensively in the early work (Dunne, 1998) are broken down here by laboratory and intention (Tables F.2, G.2, P.2, and C.2). Other than the differTABLE 0 Prior PEAR Laboratory Data (522 Series, 91 Operators) Measure BL LO HI A TABLE 00 Concurrent Calibrations (1049 Series) Measure AP CT Theory 0.000000 7.0711 0.0000 FAMMI -0.000901 7.0753 -0.1 175 GARP 0.000166 7.0691 0.0253 PEAR -0.000207 7.0697 -0.0305 Z TABLE F. 1 All FAMMI Data (250 Series, 80 Operators) Measure AP 0 BL -0.002308 7.0550 -0.1632 LO -0.006496 7.0642 -0.4593 HI 0.006336 7.07 13 0.4480 A 0.006416 7.0678 0.6416 Z MindIMachine Interaction Consortium TABLE G.1 All GARP Data (250 Series, 69 Operators) Measure BL LO 509 HI A TABLE P.1 All PEAR Data (250 Series, 78 Operators) Measure BL LO HI A TABLE C.1 Concatenation Across All Laboratories (750 Series, 227 Operators) Measure BL LO HI A ence between single-operator and dual-operator performance, which was explored only by the PEAR group, the only remarkable gender differences evident in the concatenated data are in the baseline results. Most of this effect is contributed by the FAMMI and GARP operators, with little assistance from PEAR, despite the prominence of such a disparity, albeit with opposite sign, in the prior PEAR experience (Dunne, 1998). Also possibly worth noting are the almost uniformly higher standard deviations of the female operators. (b) Assignment efSects. As one element in a broad search for subjective or psychological correlates, the data have been divided into those trials wherein the directional intention of the operator was assigned by an auxiliary random process of some sort (Instructed), and those for which the operator selected the direction (Volitional), within the constraints of balanced numbers of HI, LO, and BL trials (Tables F.3, G.3, P.3, and C.3). Here, one subset of data emerges as disparate; the GARP experiments in the Instructed mode show very significant anticorrelation with intention, in contrast to the corresponding Volitional data, which correlate positively. The difference 2-score is highly significant by any reasonable criterion. However, similar effects are not found in the FAMMI or PEAR data, leaving the concatenated data less impressive in this distinction. ( c ) Feedback efSects. The feedback presented to the operator is another subjective correlate previously examined at PEAR. The alternatives here are (a) a R. Jahn et al. TABLE F.2 Gender Effects in FAMMI Data Measure BL LO HI 0.002567 7.0673 0.1406 0.011990 7.0775 0.5362 0.3264 A 0.000623 7.0644 0.0483 0.015105 7.0727 0.9553 0.7095 Male operators (150 series, 40 operators) AP 0.029607 0.001320 CT 7.0542 7.0616 Z 1.6216 0.0723 Female operators (100 series, 40 operators) AY -0.050180 -0.018220 CT 7.0559 7.0680 Z -2.2441 -0.8148 Differences Zd8(F - M)P -2.7639 -0.6769 TABLE G.2 Gender Effects in GARP Data Measure BL LO HI -0.005306 7.0776 -0.2643 -0.010810 7.0651 -0.5426 -0.1946 A -0.001036 7.0595 -0.0730 0.005500 7.0536 0.3905 -0.3268 Male operators (124 series, 35 operators) 0.023645 -0.003234 AP 7.0493 7.0414 0 Z 1.1775 -0.1610 Female operators (126 series, 34 operators) AP -0.015103 -0.021810 0 7.0624 7.0422 Z -0.7582 -1.0948 Differences Zdm(F- M)P -1.3699 -0.6567 graphic display showing the cumulative deviation; (b) a digital display with large numbers showing the current trial and running mean; and (c) no feedback at all, with results reported only at the end of the experimental run (Tables F.4, G.4, P.4, and C.4). Here we find several noteworthy entries in the GARP data, two in the digital subset ( A and BL) and one in the no-feedback subset (HI), and two in the FAMMI data, in the HI-digital and HI no-feedback subsets, all of which feed through to their difference values, and are sufficient to drive several significant excursions in the concatenated data. (d) Runlength ejfects. Since the duration of the experimental runs that require steady attention of the operators conceivably might introduce subjective factors such as boredom, distraction, and anxiety, alternatives of 100-trial (1.5-minute) and 1000-trial (15-minute) runs were admitted into the protocols (Tables F.5, G.5, P.5, and C.5). Noteworthy here are the differences between the two run lengths in the LO-intention GARP data, which, supported modestly by the corresponding PEAR data, feed through to a marginally interesting concatenation value. MindIMachine Interaction Consortium TABLE P.2 Gender Effects in PEAR Data Measure BL LO HI A Male operators (126 series, 36 operators) AP 0.005063 0.017333 CT 7.0591 7.0578 Z 0.2542 0.8701 Female operators (76 series, 22 operators) AP -0.013539 0.006382 CT 7.0680 7.0673 Z -0.5279 0.2488 Multiple operators (48series, 20 operators) AP 0.014479 -0.030417 CT 7.0584 7.0581 -0.9424 Z 0.4486 Differences Zd8(F-M)P Zdif/ - 2 ) ~ (1 -0.5728 -0.4572 -0.3372 1.2151 Note: The Gender parameter at PEAR is treated as three-valued rather than two-valued, since operator pairs also contributed to the replication database. Rather than doing three Zd8 comparisons, one set of comparisons between males and females, and separate comparisons between combined results of individual operators and the multi-operator database, are presented. TABLE C.2 Gender Differences in Concatenated Data - Measure BL LO HI 0.005020 7.0662 0.4490 -0.01042 1 7.07 13 -0.8099 -0.9058 A Male operators (400 series, 111 operators) AP 0.020028 0.004953 CT 7.0542 7.0542 Z 1.7913 0.4430 Female operators (302 series, 96 operators) AP -0.026325 -0.01 3526 CT 7.0617 7.0570 Z -2.0459 -1.0512 Differences Zdg(F-M)p -2.7192 -1.0841 0.000034 7.0602 0.0043 0.001553 7.0642 0.1707 0.1260 ( e ) Series-position effects. Since subjective issues of boredom, anxiety, overconfidence, and learning also might manifest in the operator's performance over more major blocks of experimental effort, data also have been processed on a series-by-series basis, in a search for some definitive series-position pattern, such as that found in the prior PEAR studies (Dunne et al., 1994). In the following tables, the column labeled N lists the number of operators completing that number of series, and the notation 5+ denotes the combined results of all series numbered 5 and higher. Those PEAR and GARP operators who had previously performed five or more series or their equivalent on any similar REG experiments were regarded as contributing replication se- 512 R. Jahn et al. TABLE F.3 Assignment Effects in FAMMI Data Measure BL LO -0.015397 7.0656 -0.5244 -0.003807 7.0637 -0.2359 -0.3459 HI 0.027017 7.1087 0.9202 0.000089 7.0600 0.0055 0.8038 A 0.021207 7.0872 1.0215 0.001948 7.0619 0.1707 0.8129 Instructed (58 series, 23 operators) AP 0.027466 0 7.0534 Z 0.9354 Volitional (192 series, 80 operators) AP -0.011302 CT 7.0554 Z -0.7004 Differences z d , (I - v ) ~ 1.1571 TABLE G.3 Assignment Effects in GARP Data Measure BL LO 0.082192 7.0870 1.8743 -0.023598 7.0365 -1.5795 2.2835 HI -0.100192 7.0559 -2.2847 0.002612 7.0730 0.1748 -2.2190 A -0.09 1192 7.0714 -2.9409 0.013105 7.0548 1.2405 -3.1838 Instructed (26 series, 17 operators) AP -0.024269 CT 6.9907 Z -0.5534 Volitional (224 series, 69 operators) AP 0.007411 CT 7.0635 Z 0.4960 Differences Zdm(I-V)P -0.6838 ries only in the 5+ category (Tables F.6, G.6, P.6, and C.6). Interpretation of this disparate array of results is deferred until section 111.3. ( f ) Individual laboratory explorations. Some parameter or protocol options were explored by only one of the three laboratories having a particular interest in that factor, leaving no possibilities of interlaboratory concatenations. For example, Table F.7 lists the FAMMI data acquired under supervision of various experimenters. Numbers 1, 2, and 3 refer to three particular individuals; Group 4 subsumes several incidental experimenters. No remarkable individual scores appear, and a x2statistic computed by summing the squares of the Zscores in each subset shows no evidence of significant differences in behavior. In fact, the for the HI intention is so small as to suggest anomalous consistency ( p = 0.980). In Table G.7 are listed the results of a GARP investigation of the importance of the control of the REG trials by an automatic sequencer vs. allowing the operator to initiate each trial ad libidurn. No sensitivity to this option appears in these data. x2 MindIMachine Interaction Consortium TABLE P.3 Assignment Effects in PEAR Data Measure BL 513 LO HI A Instructed (133 series, 45 operators) AP 0.007241 CT 7.0617 Z 0.3734 Volitional (1 17 series, 52 operators) AP -0.005632 CT 7.0617 Z -0.2725 Differences Z d f l (1 - VIP 0.008346 7.0586 0.4304 0.000846 7.0633 0.0409 0.2646 -0.002594 7.0614 -0.1338 0.020359 7.0632 0.9848 -0.8098 -0.005470 7.0600 -0.3990 0.009756 7.0632 0.6674 -0.7598 0.4542 TABLE C.3 Assignment Effects in Concatenated Data Measure BL Instructed (217 series, 85 operators) AP 0.008871 CT 7.0510 Z 0.5844 Volitional (533 series, 201 operators) AP -0.002193 CT 7.0602 Z -0.2264 Differences z d f l (1 - VIP 0.6145 Finally, in Table G.8 are presented GARP results for four classes of operators: those selected and processed in a formal fashion; members of the research staff; students in the laboratory; and casual visitors. Here the only striking disparity is contributed by the visitor category in the BL intention, leading to a slightly elevated x2indicator for that condition. (g) Temporal evolution of effect sizes. As an alternative representation of the full replication databases, Figures 1 through 3 present sets of cumulative deviation graphs that summarize the historical evolution of each laboratory's compounding results for the mean shifts under HI, LO, and BL intentions. For comparison, Figure 4 shows similar plots of the prior PEAR results. Figure 5 compares cumulative deviations of the HI-LO separations for each of the three laboratories. In all of these figures, the dotted parabolic envelopes are the loci of cumulative deviations corresponding to one-tailed chance probabilities of .05 at the given abscissa. R. Jahn et al. TABLE F.4 Feedback Effects in FAMMI Data Measure BL LO HI A Digital (7 series, 3 operators) AP 0.055571 CT 7.0474 Z 0.6575 Graphic (229 series, 80 operators) AP 0.000642 CT 7.0531 Z 0.0434 None (14 series, 8 operators) AP -0.079500 CT 7.0900 -1.3303 Z Differences Zd,ff - G),u (D Zd, (D - N)p Zd,ff(G-N)p 0.6402 1.3049 1.3018 TABLE G.4 Feedback Effects in GARP Data Measure BL LO HI A Digital (50 series, 37 operators) AP 0.058980 CT 7.0518 Z 1.865 1 Graphic (189 series, 69 operators) AP -0.003709 CT 7.0572 Z -0.2280 None (1 1 series, 10 operators) AP -0.110818 CT 7.0518 -1.6437 Z Differences Zd, (D - G),M Zd,(D-N)p z,, (G - N)p 1.7629 2.2802 1 .5444 111. Structural Analyses and Their Interpretation A. Primary Results The formal hypothesis with which this ensemble of mindlmachine experiments was undertaken was that the prior PEAR database, as represented in Table 0 and Figure 4, would be statistically replicated in scale and character. From the summary Tables F. 1, G.1, P.1, and C.1 and from the cumulative devi- MindtMachine Interaction Consortium TABLE P.4 Feedback Effects in PEAR data Measure BL LO 515 HI A Digital (37 series, 12 operators) 0.018811 CT 7.0885 Z 0.5117 Graphic (195 series, 7 1 operators) AP 0.001908 CT 7.0582 Z 0.1191 None (18 series, 8 operators) AP -0.042444 0 7.0442 Z -0.8053 Differences Zd,(D-G)p Zd,(D - N)p Zd8(G- N),u 0.4216 0.9533 0.8052 TABLE C.4 Feedback Effects in Concatenated Data Measure BL LO HI A Digital (94 series, 52 operators) AP 0.0429 15 0 7.0659 Z 1.8607 Graphic (613 series, 220 operators) AP -0.000297 0 7.0560 -0.0329 Z None (43 series, 26 operators) AP -0.072000 0 7.0610 Z -2.1115 Differences Zd, (D - G)p Zd,(D - N)p Zd,(G-N)p 1.7446 2.79 14 2.0327 ation graphs of Figures 1 , 2 , 3 , and 5, we conclude that this hypothesis has not been confirmed. Although the agreed upon primary indicators of effect, the HI-LO ( A ) mean shifts and their corresponding 2-scores, progress in the intended directions in all three laboratory results and in their cross-laboratory combinations, the effect size is essentially one order of magnitude smaller than for the prior data (.0034 versus .0208) and thus falls well below any credible statistical significance ( 2 = 0.596 versus 3.809). Alternatively stated, if R. Jahn et al. TABLE F.5 Runlength Effects in FAMMI Data Measure BL LO HI A 0.003202 7.0626 0.2850 0.018654 7.0873 0.8507 -0.6271 100-trial runs (198 series, 80 operators) AP -0.006313 0.001283 (T 7.05 15 7.0593 -0.3973 0.0807 Z 1,000-trial runs (52 series, 22 operators) AP 0.012942 -0.036115 0 7.0682 7.0825 -1.1647 Z 0.4174 Differences Zd8(H-T)~ -0.5527 1.0733 0.007687 7.0659 0.4837 0.001192 7.0921 0.0385 0.1864 TABLE G.5 Run-Length Effects in GARP Data Measure BL LO HI -0.008850 7.0736 -0.5206 -0.006351 7.0661 -0.2492 -0.0816 A 0.013098 7.0620 1.0896 -0.022097 7.0443 -1.2264 1.6249 100-trial runs ( 1 73 series, 68 operators) AP 0.005590 -0.035046 (T 7.0566 7.0503 -2.0615 Z 0.3288 1,000-trial runs (77 series, 33 operators) AP 0.000805 0.037844 (T 7.0545 7.0224 Z 0.0316 1.4851 Differences Zdfl(H-T)~ 0.1562 -2.3795 the prior PEAR results are used as the standard of replication, this prediction is refuted at a Z = -2.87 level. Given the sophistication and scope of the experimental and analytical procedures followed in both these contemporary studies and in the prior PEAR work, and given the many examples of both "successful" and "unsuccessful" high-quality research performed elsewhere over the past several decades (Radin & Nelson, 1989), this stark failure to replicate reaffirms an enduring and ubiquitous "reproducibility problem" that has long characterized mind/ machine interaction experiments of this class (Bierman & Houtkooper, 1981; Shapin & Coly, 1985). Some resolution of this replication paradox would seem to be essential to sustained progress in this field. To this purpose, various categorical possibilities need to be acknowledged and assessed: 1. Some physical or technical conditions, essential to generation of the anomalies, were not properly recognized and/or incorporated in the replication program. The primary and secondary parameters so far in- MindIMachine Interaction Consortium TABLE P.5 Runlength Effects in PEAR Data Measure BL LO 100-trial runs ( 1 39 series, 49 operators) AP 0.009540 -0.01 3626 0 7.0661 7.0690 Z 0.5030 -0.7 184 1,000-trial runs ( I 1I series, 42 operators) AP -0.009207 0.027955 0 7.0561 7.0504 Z -0.4338 1.3172 Differences Zd,(H-T)~ 0.6586 -1.4609 TABLE C.5 Runlength Effects in Concatenated Data Measure BL LO HI A 100-trial runs ( 510 series, 197 operators) AP 0.002045 -0.015104 0 7.0572 7.0589 Z 0.2065 -1.5254 1000-Trial Runs (240 series, 97 operators) AP -0.001196 0.017246 0 7.0582 7.0484 Z -0.0828 1.1948 Differences Zd, (H - T)P 0.1852 -1.8482 2. 3. 4. 5. vestigated are not crucial to these phenomena and thus yield marginal results contaminated by artifact and obscured by random flux. Certain subjective psychological conditions, essential to generation of the anomalies, were not properly recognized and/or incorporated in the replication. The statistical analyses and/or their theoretical foundations deployed to distinguish anomalous and normal behavior are inadequate for the task. The basic assumptions underlying the conceptual framework within which these experiments were designed are incorrect or inadequate to encompass the phenomena involved. The phenomena underlying the anomalies are intrinsically irreplicable and unpredictable, even on a statistical basis and even with all objective and subjective parameters closely controlled, and thus are inaccessible to definitive scientific study. The last, most radical possibility surely should be deferred until all other op- 518 R. Jahn et al. TABLE F.6 Series-Position Z-Scores in FAMMI Data Series no. N BL LO HI A TABLE G.6 Series-Position 2-Scores in GARP Data Series no. N BL LO HI A TABLE P.6 Series-Position 2-Scores in PEAR Data Series no. N BL LO HI A TABLE C.6 Series-Position 2-Scores in Concatenated Data Series no. N BL LO HI A tions are exhausted. Selection among the remaining categories may possibly be informed by the internal structure of the experimental databases, e.g., from the secondary parameter breakdowns of the previous section, the higher moments of the distributions, or the sequential correlations in the data streams. In the prior PEAR studies, such attention to structural details of the data distribu- MindIMachine Interaction Consortium TABLE F.7 Experimenter Effects in FAMMI Data Measure BL LO Experimenter number 1 (37 series, 8 operators) AP 0.025351 0.044757 CT 7.0567 7.0840 Z 0.6896 1.2175 Experimenter number 2 (109 series, 15 operators) AP -0.010220 -0.023670 CT 7.0673 7.0583 Z -0.4772 -1.1052 Experimenter number 3 (80 series, 50 operators) AP 0.015900 -0.008825 CT 7.0316 7.0780 Z 0.6360 -0.3530 Experimenter group 4 (24 series, 46 operators) AP -0.069708 0.000250 CT 7.0741 7.0138 Z -1.5272 0.0055 Chi-squared on Zs with 4 df (90% CE: 0.71-9.49) 3.4402 2.8283 x ' Note: df = degrees of freedom. TABLE G.7 Control Mode Effects in GARP Data Measure BL LO HI A Auto (193 series, 68 operators) AP 0.011927 CT 7.0529 Z 0.7410 Manual (57 series, 25 operators) AP -0.022333 CT 7.0661 Z -0.7541 Differences Zdg(A-M)p 1.0164 -0.015824 7.0363 -0.9831 -0.001667 7.0603 -0.0563 -0.4200 -0.012176 7.0685 -0.7565 0.005789 7.0808 0.1955 -0.5330 0.001824 7.0524 0.1602 0.003728 7.0705 0.1780 -0.0799 databases, which in that case contained strong primary results. Indeed, most of the salient features of these prior results devolved from such structural assessments, and much of our admittedly tentative and incomplete understanding of the basic nature of the phenomena is based on them. It behooves us, therefore, to establish whether the contemporary replication database, despite its minimal primary yield, nonetheless also embodies internal structural aspects that depart significantly from chance expectation. If so, these could uncover some other form and degree of anomalous effect, or indicate flaws in the experimental design that reduced the overall yield. R. Jahn et al. TABLE G.8 Effects by GARP Operator Types Measure BL LO HI A Formal operators (169 series, 41 operators) AP -0.005089 -0.028266 CT 7.0621 7.0330 Z -0.2958 -1.6433 Staff operators (30 series, 6 operators) AP 0.053967 CT 7.0115 Z 1.3219 0.008167 7.0383 0.2000 Student operators (41 series, 17 operators) AP -0.033732 0.044415 CT 7.0448 7.0622 Z -0.9659 1.2718 Visitor operators (10 series, 5 operators) AP 0.165300 -0.043800 CT 7.1276 7.1166 Z 2.3377 -0.61 94 Chi-squared on Zs with 4 df (90% CE: 0.71-9.49) 8.2328 4.7418 x ' B. Structural Anomalies 1. Structural parameters. The data tables presented in section II.B.3 summarize our attempt to collate the results of the three laboratories, individually and collectively, with various experimental parameters, in the hope that any . significantly deviant subsets or disparities between alternative modalities might illuminate the most important objective or subjective correlates. Specifically studied, to varying degrees, have been the following structural cells: Operator Parameters Gender: Male; Female; Multiple Types: Formal; Staff; Student; Visitor Protocol Parameters Assignment of intention: Instructed; Volitional Feedback modalities: Digital; Graphic; None Machine control: Automatic; Manual Run lengths: 100 trials; 1000 trials Sequential Effects Series-position Experimenter Effects Individuals by code number As already noted, a substantial number of suggestive disparities have indeed appeared in the data subsets. However, because of the number of cases exam- MindIMachine Interaction Consortium 521 0 50 100 150 200 250 Thousands of Trials Fig. 1. FAMMI cumulative deviations. 0 50 100 150 Thousands of Trials Fig. 2. GARP cumulative deviations. 200 250 522 R. Jahn et al. 0 50 1 00 1 50 200 250 Thousands o Trials f Fig. 3. PEAR Laboratory cumulative deviations. 0 200 400 Thousands o Trials f 600 800 Fig. 4. Prior PEAR cumulative deviations. MindIMachine Interaction Consortium I 523 I 0 1 00 200 300 400 500 Thousands of Trials Fig. 5. Cumulative HI-LO differences for all three labs. ined, some seemingly meaningful distinctions may appear by chance, so we cannot interpret the several large Z-scores in the structural tables until we have somehow corrected for the multiplicity of tests, to learn whether these are indeed larger or more numerous than would be expected by chance for the number of analyses that have been generated. The discussion of sequential and experimenter effects will be deferred to a later section. For the moment, we will consider only the operator and protocol parameters, as they are broken down in Tables F.2 through F.5, G.2 through G.5, and P.2 through P.5. These tables report a total of 124 mean-shift 2-scores for the various intentional condition subsets. More importantly, 76 Zdifs scores for differences between parameter conditions are presented. Since any structural anomalies in these parameters would appear as differences of performance between different parameter conditions, the 76 Zdflscores are obviously the crucial population to test. We may also check the population of mean-shift 2-scores, but this test is less central to the examination of structure, first because the statistical resolution is relatively weak since each Z involves only one half of a parameter comparison, and second because the absence of an overall intentional effect makes significant mean shifts in these full subsets much less likely. We might naively suppose that we can perform the requisite multiple-tests correction simply by comparing the large population of Zd8 scores to the theoretical 2 distribution. For example, since the subset comparisons are not di- 524 R. Jahn et al. rected, i.e., we do not have a prior hypothesis regarding the sign of any Zdfl, the presence of structure might be expected to inflate the absolute magnitude of some ZdiY. scores, and therefore the standard deviation of the Zdlff score distribution. And, indeed, when we examine the standard deviations in these populations, we find that the 76 Zdflvalues have a standard deviation of 1.258, rather than the theoretically expected value of 1, a result unlikely with p = 0.00098. At face value, this might seem strong evidence for structure in the Zd,ffpopulation. The flaw in such a conclusion is that the analysis presupposes that the scores comprising the population are mutually independent, which they are not. To begin with, each score in the A column of the data tables is strongly correlated with the scores in the HI and LO columns. The breakdown in the feedback parameter, having as it does three levels, produces a set of three parameter differences, each strongly correlated with the other two. Worse, there are additional correlations between 2-scores in different parameter comparisons, because the populations are not in uniform proportion. For example, the fraction of instructed-assignment series generated by females is not necessarily the same as the fraction of volitional-assignment series generated by females, because of the freedom of operators to choose secondary parameters. When these proportions are not equal, Zdlff - V) will acquire an intrinsic cor(I relation, positive or negative, with Zd8 (F - M). Similar considerations apply among almost all of their parameter sets. The presence of these correlations, of variable magnitude and sign between different Zdg scores, complicates the comparison with theory immensely, so much so that the attempt was abandoned. Instead, it was decided to determine the theoretical values of the population-summary parameters empirically through a Monte Carlo procedure, the details of which are given in the next section. 2. Monte Carlo simulations. ( a ) General treatment. We wish to determine whether the populations of Zscores, especially the population of 76 Zdfl scores, emerging from Tables F.2 through F.5, G.2 through G.5, and P.2 through P.5, depart from the expected chance distribution for this array of tests when applied to random data. To determine this chance distribution, we employ a Monte Carlo procedure which in essence involves repeatedly performing the analysis on data that are guaranteed to be random. The analysis programs that were used to process the empirical data for the above tables take, as input, the indicial information describing the parameters for each series, and the actual data generated in the series. For the Monte Carlo process, we submit to those programs exactly the same indicial information, along with ersatz data constructed with a numerical pseudorandom algorithm to match the null-hypothesis distribution for these experiments. The fact that we are using the indicial information from the actual experiments guarantees that we reproduce the correct correlation structure in the output Z population. Mind/Machine Interaction Consortium 525 (We use simulated data rather than simply reordering the actual data, because if structure does exist in the actual data, the statistics of the raw data must necessarily be distorted to some extent. Randomly reordering the raw data, as is often done in Monte Carlo applications, does not serve our purpose in the current case. A random reordering breaks the connection between the data and the indicia1 information but leaves intact-merely relocated-the shifted values that constitute the structural anomaly and therefore does not give a reliable measure of the null-hypothesis distribution.) Thus, each iteration of the Monte Carlo process produces its own population of 76 Zd8scores. (It also produces a population of 124 mean-shift 2-scores, which are also analyzed and reported for the sake of completeness.) This process is then repeated a total of 5000 times to ensure that the distribution parameters are well estimated. Any measure-e.g., the standard deviation described above-that characterizes the population of Zdg scores produced by the actual data thus can be compared with 5000 samples from its null-hypothesis distribution produced by the Monte Carlo procedure. Table M. 1 presents the results of this comparison with the Monte Carlo populations for several such summary measures. Each of these measures is a slightly different quantification of the qualitative hypothesis that the population of Zdgscores in the actual data has larger absolute values than predicted under the null hypothesis. The measures presented are the standard deviation, discussed above; the largest absolute value of any ZdlfS the population; and in the number of ZdHscores in the population exceeding each of three thresholds. The "population" referred to here is always the population of 76 ZdlfSvalues (or in Table M.la, 124 mean-shift 2-scores) produced by a single instance of the analysis, real or simulated (not the population of 5000 simulated instances). The columns of Table M. 1 present, first, the value of the named measure in the actual data; next, the mean and standard deviation of the named measure across the 5000 Monte Carlo iterations; and next, the number of Monte Carlo iterations where the value of this measure exceeds the value in the actual data. (The number in this column, when divided by 5000, is a form of empirical upper-tailp-value describing the position of the actual data in the Monte Carlo distribution.) A final column presents measure values obtained when the actual data are replaced, not by simulated data but by calibration data from the experimental apparatus. This is included as a precaution against the possibility that differences between real and simulated data might derive from properties of the physical data source, rather than from an experimental effect. The actual calibrations from Freiburg, Giessen, and Princeton were used to replace the experimental data for their respective laboratories, in this calculation. From Table M . l , we note that, as expected, the population of 124 meanshift 2-scores is indistinguishable from the null hypothesis distribution as constructed by the Monte Carlo process. The ZdlfS Table M. 1b, however, is much more interesting. For example, the standard deviation of the ZdHpopulation now yields a p-value of .014, quite different from the erroneous calculation 526 R. Jahn et al. mentioned above, but clearly indicative of anomalous structure. Although conceptually the standard deviation increase is the primary indicator of a modified Zdg distribution, the other measures can provide additional information about the nature of the modification. However, the introduction of these different measures might suggest that the question of multiple analysis has appeared yet again, requiring some form of Bonferroni correction. This multiplicity is an unavoidable consequence of the initial exploratory decision to examine several specific ways in which the actual population of Zdflscores might depart from the null hypothesis prediction. It is possible, however, to render irrelevant all issues of multiple testing by calculating a single summary statistic encompassing all five measures presented in Table M. 1. As the table shows, each of the five measures has a mean and standard deviation determined from the Monte Carlo population. A normalized score can be calculated for each parameter relative to this distribution by subtracting the distribution mean from the observed value and dividing the difference by the standard deviation. (We do not call this normalized score a 2-score because some of the measures are not normally distributed.) The sum of these normalized scores is a single statistic that weights equally the departure from Monte Carlo norms in each of the five measures. This sum can be calculated not only for the actual data but also for each individual iteration of the Monte C a r l ~ simulation. Comparing this combined-measures summary statistic in the real data with the distribution of values in the 5000 Monte Carlo iterations gives us a single, definitive p-value for the degree to which the real data stand out from the null hypothesis: There are 109 iterations that exceed the real data in the summary statistic, and 0 exact ties, leading to ap-value of .022. Since this is a single-test result requiring no correction, we may safely conclude that the population of ZM scores in the PortREG database can be distinguished from the null hypothesis at a p = .022 level. Thus the apparent structural anomalies noted in Tables F.2 through F.5, G.2 through G.5, and P.2 through P.5 are, to this same level of confidence, real differences rather than statistical artifacts. Figures 6, 7, and 7a represent these results in an instructive graphical form. Figure 6 shows the positions of the full subset empirical data Z-scores on the Monte Carlo calculated distributions. As expected, there is little departure from chance behavior here, save a slight positive shift of the largest Z-value. In Figure 7, however, substantial displacements of the empirical Zdg values with respect to the Monte Carlo background are clear by each of the five criteria, reaffirming the numerical values mentioned above. Figure 7a, shows similar major displacement of experimental value of the composite statistic just described, with respect to the Monte Carlo distribution. While this analysis cannot guarantee that any particular subcells are aberrant, it can identify a hierarchy of such disparities that are most likely to represent legitimate structural anomalies. For example, Table M.2 lists the ten most prominent departures of the subcell difference Z-score from their corresponding Monte Carlo simulations, indexed by direction of intention and laboratory. MindlMachine Interaction Consortium 527 The secondary parameters are given in the order that makes the ZdlfS positive; thus, the first entry lists "V - I," denoting that the volitional data have a larger A-effect than the instructed. From Table M. lb, we know that the number of ZdlfJ, the range above 2.0 to be affected by chance is about 3.5; hence, it is in likely that some six or seven of the entries in Table M.2 correspond to real, nonrandom differences in operator achievements. ( b ) Most favorable cells. While such Monte Carlo treatments provide no guarantees that any given one of these categories in fact entails anomalous results, they can provide guidelines for the most profitable cells to study more directly, leading to identification of the more important secondary parameters, and hence possibly to superior further experiments. As just one example, the data subset comprising all of the trials performed at the GARP laboratory using volitional assignment of direction of intention, nongraphic feedback, automatic machine control, and 100-trial runs shows a significant yield in the HI-LO separation of p = 0.488 + 0.0241 (2= 2.02), whereas the subset of all data delineated by instructed assignment, graphic feedback, automatic control, and 1000-trial runs shows a strong negative yield of p = -0.2308 + 0.0913 (2= -2.53). The source of this disparity may be further localized by noting that the combination of all GARP instructed, graphic subsets yields p = 0.1010 + 0.0323 (2= -3.13), suggesting that the subjective parameters of volitionallinstructed assignment and graphiclnongraphic feedback were particularly pertinent to GARP operator performance. Such observations then prompt examination of the corresponding subsets in the FAMMI and PEAR databases to see if such effects appear in these venues, as well. To facilitate such interlaboratory cell comparisons, it is necessary to devise a standard procedure for dividing all of the PortREG databases into commenTABLE M.1 Comparison of All Laboratory Data with 5000 Monte Carlo Simulations Measure Data 5000 Monte Carlos No. M. C. > data Calib. data (a) Distributions of 124 mean-shift Z-scores 0.980+ 0.129 SD of Z 0.961 2.941 2.691 + 0.437 Largest lZl 16 16.702 + 6.932 No. (of 124): lZl> 1.5 5 5.725 + 4.037 No. (of 124): IZI > 2.0 1 1.572 + 1.950 No. (of 124): lZ)> 2.5 (b) Distributions of 76 Zdifsscores SD of Z,, 1.258 3.184 Largest IZ,& > 1.5 19 No. (of 76): Iz,& No. (of 76): IZd&> 2.0 10 2 No. (of 76): IZd&> 2.5 0.995 + 0.114 2.597 + 0.432 10.206 + 3.834 3.540 + 2.299 1.003 + 1.189 " Since these parameters are discrete, an exact match can occur between the value in the actual data and the value in a Monte Carlo iteration. Therefore, the number reported here is the number of Monte Carlo values strictly greater than the data plus one half the number of exact matches; this is a standard approach to calculating tail populations with discrete data. 528 R. Jahn et al. surate subsets that control for various possible confounds. As already noted, many of the subset parameters are mutually confounded due to unequal subset sizes. For example, the GARP data appear to show differences between intentional assignment modes and also between feedback modes. Since the proportions of a given assignment mode are not guaranteed to be the same in all feedback modes, when we only dissect the data according to one parameter at a time we cannot know whether (a) a real difference between assignment modes drives an apparent difference between feedback types, (b) a real difference between feedback types drives an apparent difference between assignment modes, (c) both parameters are independently important, or (d) both parameters are interdependently important; i.e., that the difference in performance might not be associated with either parameter in isolation but only appears when they jointly take on appropriate values. To distinguish these cases, we need to decompose the data according to several parameters at once, creating "cells" that are consistent according to several secondary parameters. This has two benefits. First, we can distinguish among cases (a) through (c), by making unconfounded tests for each parameter. Second, we can identify case (d) if the differences between cells contain information not explicable in terms of the unconfounded effects of isolated parameters. Ideally, one should break down the data according to all secondary parameters. Unfortunately, there are so many of these that to make such a complete subdivision would result in very small data subsets with correspondingly poor statistical resolution. Moreover, there is a significant risk that some cells in such a complete breakdown would be entirely empty, appreciably complicating the interpretation. As a balance between rigor and practicality, the following compromises are made: 1. Only "optional" parameters subject to operator choice are considered. Gender, fixed for each operator, is ignored. Series position, also not optional, and in any case showing hard-to-interpret variations, also is ignored. 2. Only parameters for which all three laboratories examined the parameter are considered. This reduces the selection to assignment mode, run length, and feedback. 3. Because each laboratory has a huge majority of its data in the graphic feedback condition, the other two modes are collapsed into a single "nongraphic" feedback category. The result of these compromises is the eight-cell (2 x 2 x 2) breakdown used in Table C.7 and Figures 8-11. In these, a three-letter code is used to indicate the parameter values: the first letter, I or V, refers to instructed or volitional assignment; the second, G or N, to graphic or non-graphic feedback; the third, H or T, to 100-trial or 1000-trial runs. The values plotted on the figures are absolute mean shifts in direction of intention for HI, LO, BL, and A. The 2-scores tabulated in Table C.7 are based only on the A-effect, but all four intentional conditions are plotted in the figures. MindIMachine Interaction Consortium Data 529 Y- ii O 3- E .cn 2. 10:6 0.8 :O c 1 .O 1.2 1.4 1.6 Std Dev of Z Largest Z No. z > 1.5 Data No. Z > 2.0 500. 000 500. Data - n n - - O 0 5 10 No. Z > 2.5 15 Fig. 6. Mean-shift Z-scores vs. Monte Carlo populations. Returning to our particular example, the comparisons of performance under the volitional, nongraphic, 100-trial protocol (VNH), and the instructed, graphic, 1000-trial protocol (IGT), are seen to be particularly inconsistent across the three laboratories. This has encouraged further, ad hoc experimen- 530 R. Jahn et al. Data 3- 1.4 Std Dev of Z Largest Z 1 Data I " 0 5 10 No.z > 1.5 15 20 25 No.Z > 2.0 # 2000y 1000. 8 1500. O ' - Data m z o 500- 0 2 4 No. Z > 2.5 6 8 Fig. 7. Difference 2-scores vs. Monte Carlo populations. tation, which is now in progress, and has prompted some new initiatives in theoretical modeling, which cannot be detailed here. Similar structural exercises can be attempted in terms of other discriminators suggested by the Monte Carlo "most prominent" list above, such as opera- MindIMachine Interaction Consortium 531 Sum of Normalized Differences Fig. 7a. Composite statistic for difference Z vs. Monte Carlo. I tor gender, or single vs. multiple operators, both of which revealed striking disparities in the prior PEAR studies. In the replication studies, however, these effects are not so clearly evident. With reference to Tables F.2, G.2, and P.2, the only suggestive disparities appear in the PEAR data alone, and here most prominently in the single- vs. multiple-operator comparison, which was not explored by the other laboratories. Nonetheless, Table C.7 also presents a set of rudimentary correlation coefficients that indicate a much closer correspondence of the cell-by-cell result patterns between GARP and PEAR than between FAMMI and either other laboratory. Obviously, it would be most desirable if it were possible by some means to extract from these structural cell results a completely unconfounded set of correlations with individual secondary parameters. Some form of analysis of variance (ANOVA) suggests itself, and indeed such has been employed twice in analyzing the prior PEAR data (Nelson et al., 1991, 2000), but even with the much higher overall yield of that database, the insights gained thereby did not vastly exceed those acquired from more directed ad hoc analyses. Nonetheless, once one has the cell scores, it is straightforward, although tedious, to construct the unconfounded secondary parameter effects. For example, to assess the effect of assignment mode, one must first compare the four pairs of cells that differ only in this parameter, i.e., IGH vs. VGH, IGT vs. VGT, INH vs. VNH, and INT vs. VNT. Each of these comparisons can be reduced to a difference 2-score using the formula at the end of section 11.1. The four 2-scores I I I 532 R. Jahn et al. TABLE M.2 Most Prominent Z-Score Differences from Monte Carlo Comparisons Parameter ASG V-I GEND M-F RUNL T-H FDB D-N ASG I-V FDB D-N ASG V-I FDB D-G MULT 2-1 FDB D-G Intention Lab GARP FAMMI GARP GARP GARP GARP GARP FAMMI PEAR GARP &if so produced then can be combined into a single Z giving the overall effect of that parameter, according to the composition rule: where Z, denotes the composite Z for a set of scores, Zi, all measuring the same effect on databases of sizes n,, i = 1,... ,N. In this manner it is possible to extract unconfounded correlations with certain specific secondary parameters, with the results displayed in Table C.8. Particular further examples could be cited, but the broader point at issue is that the combination of the Monte Carlo simulations of the cellular data subsets with subsequent specific analyses of the most suggestive cells may help to localize the most pertinent objective and subjective parameters, and to refine future experiments to optimize these factors. We feel that only through such a detailed and disciplined process, tedious as it may be, is there hope for more effective and replicable experimentation, leading to better understanding of the phenomena. 3. Series-position effects. One possible structural indicator not explicitly explored in the Monte Carlo comparisons but readily accessed within the various laboratory databases, commonly termed "series-position effects," relates to the evolution of operator performance as a function of the number of experimental series performed. The prior PEAR data displayed a remarkably ubiquitous and consistent trend for scores to be highest for the first series attempted, then to deteriorate for the next two series, then to return to higher performance on the fourth, fifth, and subsequent series (Dunne et al., 1994). With reference to Tables F.6, G.6, and P.6, some such serial oscillations of performance are apparent, particularly in the GARP and PEAR data, but these are far from consistent across the three laboratories. Nonetheless, the composite data (Table C.6) also show some series-position pattern, but quite different from that of the prior PEAR results. MindIMachine Interaction Consortium 533 As a supplementary indicator, standard x2tests applied to these patterns, computed relative to chance expectation and relative to their respective empirical mean values, are displayed in Table C.9, along with their corresponding probabilities of chance occurrence. The last line presents the same analysis of the prior PEAR data. Clearly, only the GARP data exhibit a credible series-position pattern, albeit quite different in form than the prior PEAR results. Namely, the highest scoring in that replication is occurring in the second series, rather than in the first, and the lowest scoring in the fourth, rather than the third. In other words, the series pattern has shifted by one series. 4. Operator-specific features. Another structural anomaly identified in the prior PEAR data was the persistence of individual operator accomplishment features or "signatures," apparent over several series of effort, or over entire databases. Since few of the operators involved in the replication studies produced sufficient data for us to pursue this tendency solely in that context, we have modified the question to query whether those five operators who have appreciable databases in both the prior PEAR experiments and the replication study show similarities of performance between the two applications. For each of these operators, we calculate a 2-score for the difference in their HI-LO performance between the old and new experiments, using the Zdgformula in Equation I. We use the same formula to calculate differences between their performances in the three individual intentions, HI, LO, and BL. The sum of the squares of those Zm, becomes, for each operator, a with 3 d measurf ing the overall change in performance across all three intentions between the original experiment and the replication. The results, along with the associated chance probabilities, are presented in Table P.7. Two potentially instructive features are apparent. On the one hand, the first four operators, both individu- 2 TABLE C.7 2-Scores in Secondary Parameter Cells, by Laboratory Parametera IGH IGT INH INT VGH VGT VNH VNT Correlation coefficients of these response patterns FAMMI-PEAR FAMMI-GARP P -0.0061 -0.4500 -1.1188 Z(P> -0.0143 GARP-PEAR 0.6501 1.7452 FAMMI GARP PEAR All 3 labs Note: p = correlation coefficient; p = 1 = perfect correlation; p = -1 = perfect anticorrelation; Z(p) = standard normal deviate corresponding to value of p. a I = instructed protocol; V = volitional protocol; G = graphic feedback; N = no feedback; H = 100trial runs: T = 1000-trial runs. IGH Fig. 9. IGT INH INT VGH Data Category VGT VNH VNT GARP data split by assignment (I,V), feedback (G,N), and run length (H,T). MindIMachine Interaction Consortium 535 IGH IGT INH INT VGH Data Category VGT VNH VNT Fig. 10. PEAR data split by assignment (I,V), feedback (G,N), and run length (H,T). Data Category Fig. 11. All data split by assignment (I,V), feedback (G,N), and run length (H,T). 536 R. Jahn et al. ally and collectively, performed remarkably similarly on the two experiments. On the other hand, Operator E displays a stark difference in performance between the prior PEAR and replication efforts that is virtually an inversion or "antireplication" of the prior "signature." (It may be worth noting that this operator repeatedly expressed strong resistance to being asked to validate a prior achievement through replication.) Clearly, these contradictory results cannot be resolved further without considerably more operator-specific data, but the subjective issue raised could ultimately prove important. Other aspects of operator-specific structural anomalies have also been explored by similar techniques. For example, the possibility that a mixture of strong performances in the intended directions and in the directions opposite to intentions among the individual operators may cancel one another in the overall yield, thus obscuring the operator-level effects in the database, can be checked by a x2calculation encompassing all operators at the three laboratories, under all intentions. Specifically, by squaring the individual operator Zscores (thus obtaining a sign-independent quantity) and adding these across all operators, we construct a with degrees of freedom equal to the number of operators. Table C.10 presents such results for the three laboratories along with their associated chance probabilities (in parentheses). Because the prior PEAR experiments indicated a gender difference in the tendency toward idiosyncratic performance, the databases are subdivided by gender as well as by laboratory. Since the tests on the three individual intentions are mutually independent, they can be collected in a combined value indicative of the overall departure from chance behavior in all three intentions (last column). No elevated values that would suggest idiosyncratic operator performance appear. To the contrary, the PEAR female operators show a strikingly depressed x2,especially in the LO intention, that compounds to an extraordinarily diminished value across all three intentions (39.33 on 66 dfi p = .996). Considered as an improbably small x2, this corresponds t o p = .004, which we must immediately correct to .008 since we are willing to consider both unusually large and unusually small x2. Bonferroni adjustment of this value for the seven independent subsets (two genders each at GARP and FAMMI, three at PEAR) still leaves a suggestive p = .051. Thus, there are moderate grounds for suspecting that this particular operator population is somehow producing performances that cluster too tightly about zero yield. Such calculations have been repeated on a series-by-series basis. Again, only the PEAR females show significant anomalies that survive the multiple-testing adjustments. It also may be worth noting that the data collected on series-position effects (Tables F.6, G.6, P.6, and C.9) and on operator-specific features (Tables P.7 and C. 10) show a polyglot nature of above-chance occurrences similar to those covered in the Monte Carlo treatment (Table C.7). 5. Standard deviations. A different form of structural irregularity that may have indicative value can be detected in the individual laboratory and compos- x2 x2 x2 MindIMachine Interaction Consortium TABLE C.8 Difference Z-Scores of Unconfounded Secondary Parameters Parameter test Assignment (I-V) Feedback (G-N) Runlength (H-T) FAMMI GARP PEAR All 3 labs 537 0.4224 -0.7343 -0.4275 -2.9204 -1.0871 1.3792 -0.9689 -0.4485 0.5599 -1.2611 -1.7823 0.9331 ite databases. Even cursory examination of the tables of section 1I.B reveals many instances where the trial-level standard deviations are less than the theoretical value of 7.07 1. This, of course, might be an artifactual result of a flaw in the random noise sources, so these standard deviation figures should be compared not with the theoretical value, but with an empirical value derived from the concurrent calibrations of the instruments (cf. Appendix I). Since the three calibration datasets have consistent means and standard deviations, a pooled estimate of the latter may be constructed, yielding o = 7.07 10 with an empirical uncertainty of a0.0028. Table C.11 reports Z-scores for the difference between the trial-level standard deviations of the active experimental data and this calibration estimate. This method of comparison to an empirical standard technically makes them Student's t-scores rather than 2-scores. However, since there are well over 10,000 degrees of freedom in even the smallest datasets examined, the difference between the Z and t distributions safely may be neglected. By either standard, we find a statistically robust difference between the active experimental data and the calibration data in the composite across all three laboratories that is driven by substantial depressions in the LO and BL conditions. The prior PEAR finding of significantly higher experimental standard deviations for female operators compared to males (Dunne, 1998) is not sustained in magnitude by the replication data, although virtually all of the individual laboratory results show slight separations in this direction. 6. Counts of successful operators and series. In addition to the trial-score distribution criteria on which all of the preceding tabulations and discussions are predicated, the data also have been examined in terms of the fraction of experimental series and the fraction of operators, whose results conform to any extent with the direction of intention. Although those perspectives had proven Laboratory FAMMI GARP PEAR All 3 labs Prior PEAR 2Tests for Series-Position Z-Scores x2(VS. theory); 5 df 2 (VS. empirical mean); 4 df p of 2 a TABLE C.9 p of x2 1.9316 13.5799 1.1688 5.2207 27.3385 .859 .019 .948 .390 .00005 1.7431 13.5699 1.1680 5.0880 18.2453 .783 .009 .883 .278 .001 538 R. Jahn et al. instructive in some of the prior work, they clearly are not independent of the mean-shift values and in this replication study have added little new insight. Nonetheless, full tabulations of these quantities are available on request. IV. Summary Comments As described in the introductory section, this coordinated replication study was the first collaborative research project attempted by the Freiburg, Giessen, and Princeton laboratories, as much to test the viability of the consortium concept, structure, management, and operations strategy as to create a major new database in mindlmachine anomalies. By the former criterion, the project has been undeniably successful in that methods for provision of common experimental equipment, acquisition and reduction of experimental data, and analysis and interpretation of results have been well established and are available for deployment in subsequent research endeavors. Visitation and exchange of personnel among the laboratories at both the staff and management levels occur frequently, and the electronic communication channels that enable sharing of data and ideas function on a regular basis. In short, this first project has demonstrated that this ambitious consortium can function productively on such collaborative research enterprises. As far as the replication results themselves are concerned, we are left with an empirical paradox. Whereas the prior PEAR experiments clearly displayed anomalous secular trends in REG output distribution means in correlation with operator intention, the three-laboratory replications, which employed essentially similar equipment and protocols, failed by an order of magnitude to replicate the primary correlations. Yet, these replication studies presented instead a substantial pattern of structural anomalies related to various secondary parameters, to a degree well beyond chance expectation and totally absent from the calibration data. To borrow a fluid mechanical metaphor, it is as if the influence of operator intention now was manifesting itself as a structural "turbulence" in the output data of the replication, rather than in a more orderly displacement of the data streams as was found in the prior PEAR studies. With the various ad hoc examinations of these structural details described in sections I1 and I11 in hand, our search for some understanding of this substantial change in the character of the anomalous responses of the machines to operator TABLE P.7 Consistency of Operators Between Prior PEAR and Replication Experiments Operator x <PI ' Z 0, [2-tail]) MindIMachine Interaction Consortium 539 intention may be aided by systematic reconsideration of certain explicit and implicit assumptions with which the replication studies were undertaken: 1 . Source independence: The anomalous eflects would manifest in the same form and scale on the PortREG sources as they had on the original PEAR benchmark machine. The prior PEAR data reported in Table 0 had been generated using a far more expensive and complex REG device that was replete with an array of failsafe controls, interior checkpoints, and other protections against short- and long-term deviations from strictly random behavior, that would unequivocally guarantee the integrity of the experimental results. The shift to the much simpler, less expensive, and more portable PortREG equipment seemed justified on the basis of its earlier successful deployments in other PEAR-based experiments, most notably our FieldREG studies (Nelson et al., 1996, 1998), and an extensive body of past evidence that comparable anomalous results could be obtained utilizing categorically different random physical sources (Jahn et al., 1997; Schmidt & Pantas, 1972). Yet, since that time certain other applications of PortREG equipment also have failed to produce results comparable with the prior benchmark findings, raising some questions about its consistency of sensitivity to operator intention (Jahn et al., 2000). It has been suggested by one of PEAR'S long-term operators that this reduction in effect may not be attributable to physical differences in the noise sources, per se, but to the shift of the REG unit from its original central focus in the experimental configuration to one where it appears to play only a peripheral supporting role to the computer that now dominates the operator's attention. Specifically, in the prior PEAR experiments digital feedback was presented as an LED display on the face of the REG device itself, with the TABLE C.10 Operator Performance Dataset FAMMI Female Male All GARP Female Male All PEAR Female Male Co-operator All a 2 Values (with Associated Probabilities) LO HI dy 40 40 80 34 35 69 22 36 20 78 BL A Combined a The degrees of freedom for the "Combined" column,which sums up the mutually independent contributions of BL, LO, and HI, are triple the number listed in the "df" column. 540 R. Jahn et al. computer playing a more passive data-recording role, and the redundant archival data hardcopy was produced contemporaneously with the generation of the experimental data, rather than in a deferred printout. In the PortREG experiments, however, the noise source is housed in a small, unobtrusive gray box that is a far less evident component of the experimental system. Operator feedback, both digital and graphic, is produced on a computer display, rather than on the noise unit itself, and data printout is under computer control on a separate printer facility that operates only at the end of the run. Thus, the subjective experience of an operator generating data differs appreciably between the two experiments, so that while it is possible that the PortREG devices are still inherently sensitive to operator intention, their less prominent role in the experimental configuration may compromise their patterns of response. Another operator has suggested that the vast proliferation of interactive, visually engaging computer displays into public and personal applications over the past decade may have eroded much of the novelty of this format of humanlmachine interaction, rendering the experimental task less challenging and enjoyable. In either case, the role of feedback, rather than the noise source itself, may be the more pertinent concern, as further discussed in items 3 and 4 below. 2 . Operator pool equivalence: The overall pe$ormance of the pool of operators pe$orming the replication experiments would be similar to that of the pool of operators that produced the prior PEAR results. This presumption seemed soundly based on extensive earlier results that these anomalous effects invariably appeared as broadly distributed, marginal shifts over the full operator population, rather than being dominated by a few exceptional operators (Jahn et al., 1997). The fact that PEAR, continuing its policy of using only uncompensated, anonymous volunteers, many of whom had participated in the prior experiments, achieved no better replication than TABLE C.11 Z-Scores for Trial-Level Standard Deviations, by Laboratory and Gender Data All FAMMI Male Female All GARP Male Female All PEAR Male Female co-op Composite BL LO HI A I Note: Z-scores calculated from normal approximation to the distribution of standard deviations, which is accurate for these large datasets. MindIMachine Interaction Consortium 541 GARP or FAMMI, who followed more structured handling of operators, continues to suggest that the composition of the operator pool, per se, is not likely to be a major factor. Yet, some of the structural evidence from this present study, as discussed in items 4 and 7, may indicate otherwise. 3. Insensitivity to secondary parameters: The overall results would be insensitive to minor alterations in the secondary experimental parameters. The prior PEAR data generated with digital feedback or no feedback were statistically indistinguishable from the graphic-feedback data, leading to the assumption that feedback was a matter of indifference or at most of individual operator aesthetic preference. Both of the ANOVA studies of the prior PEAR data also failed to uncover any overall feedback sensitivity. Yet, the differences in replication results related to this parameter indicate that it may have been a mistake to choose graphic feedback as the introductory default, even though it seemed to be the most popular choice of the operators. Similar considerations apply to the run-length option. Indeed, the breakdown by secondary parameter cells in Table C.7 indicates that data generated solely in the most conducive secondary conditions had effect sizes comparable to those seen in the prior PEAR experiments. While none of this explains why the relative insensitivity to these parameters observed previously should have changed, this presumption also now must be questioned. 4 . Insensitivity to operator attitudes: Various psychological or subjective parameters pertinent to operators' attitudes in addressing the experimental task, such as their prevailing emotional state, their sense of purpose or enjoyment, the laboratory ambience, the experimenter's expectations, and other environmental factors, would be adequately preserved in the aggregate b y the operator selection and handling procedures exercised in the replication. Prior PEAR experience (Jahn & Dunne, 1988, 1997), supplemented by extant psychological and parapsychological literature (Rosenthal, 1963; Schlitz, 1986), suggested that certain aspects of the experimental ambience may be conducive to generation of anomalous effects. Examples include a friendly, relaxed, even playful atmosphere; a supportive attitude summarized as "permission to succeed;" a lack of pressure or urgency for success; an "unfocused" or "long-wavelength" state of thought and attention; etc. Given the nonreplication, however, it now appears that either these psychosocial factors are not so important or we failed to instill a propitious balance of them into our operators' experiences. Possibly supportive of the importance and difficulty of maintaining these attitudinal factors is some mild evidence for an "epochal" segmentation of the chronological results from each laboratory. For example, with reference to the cumulative deviation graphs of Figure 5, we can identify in each laboratory's 542 R. Jahn et al. full record long spans of HI-LO yield (FAMMI: trials 60,000-195,000; GARP: trials 245,000-345,000; PEAR: trials 195,000-350,000) that were quite comparable to those of the prior PEAR studies. The reality of such bimodal inhomogeneities in these databases, vis-g-vis chance excursions of binary random walks, cannot be confirmed statistically for this amount of data, but it is interesting to recall that the larger body of prior PEAR results also displayed a bimodal epochal character that took a statistically more convincing form. Specifically, there we found three virtually equal-length epochs, having strong performance over the first, chance performance over the second, and strong performance over the last (cf. Figure 12). While it is difficult to establish a Bonferroni-type correction factor for this sort of retrospective reexamination of an extant database, taken at face value the distinction between the three epochs is quite significant = 7.566 on 2 df, p = .0228). The second epoch is a "nonreplication" of the first quite as stark as the overall PortREG nonreplication and is of a comparable scale. It was earlier noted that taking the overall prior PEAR database as a standard, the replication effort refuted the prediction at a level of Z = -2.87. Yet, Figure 12 shows us that when PEAR itself, employing a known, productive experiment with the same protocols and operator pools, generated an REG database of the scale of PortREG three times in succession, it failed to show anomalous yield one time in three. In this view, the joint failure of three laboratories to replicate is an event with p = .037, rather than thep = .004 one would infer from the above 2-score. In both the prior PEAR and replication cases, the strong epochal results are diluted by the remainder of their respective databases. Nevertheless the presence of extended segments of high yield, and of negligible yield, in both the prior PEAR and in all three replication databases, raise valid questions concerning what subjective factors bearing on the operators or, for that matter, on the experimenters, prevailed during these lengthy periods of apparently successful replications, and did not in the other, nonproductive major segments. (x2 5. Intention as primary correlate: The specification and control o operaf tor "intention" is adequate to designate this property as the primary correlate of the anomalous eflects. While there is no doubt that the stipulation of an operator intention as BL, HI, or LO, irrevocably specified and recorded prior to initiation of an experimental run, qualifies as an objective index for the subsequent data, it is equally clear that the processes by which the operator assumes and deploys that intention are inherently subjective in character, and hence potentially vulnerable to any influences that alter that subjectivity. We need look no further than the substantial aberrations in baseline behaviors, or the ubiquitous constrictions of trial-level standard deviations, or the epochal successions just mentioned, to infer that subtle subconscious as well as conscious mental and emotional processes may be at work in conditioning the operator's expression of intention. How these processes react to the perceived "success" or "failure" of an MindIMachine Interaction Consortium 543 ongoing experimental run or of a previously completed series; to the operator's sense of "resonance" with the experiment; to the sense of importance of the achievement; or to the temporal variations in the operator's mood or state of health are not really illuminated by these experiments, and remain far from our grasp. What does emerge, however, is a legitimate question as to whether intention is the best primary correlate for such anomalies or, as suggested by the FieldREG experiments (Nelson et al., 1996, 1998), some subtler criterion for the requisite mindlmachine "resonance" would be more fundamental, or at least complementary to it. 6 . Replication criterion: Successful replication validates the phenornenon; failure to replicate disqualifies it. The concept of objective replication or falsification is crucial to the exact sciences. Yet examples abound where varying degrees of compromise with rigorous replicability have been tolerated out of pragmatic necessity. For example, the essential indeterminancy of quantum events forced physicists to acknowledge that for some experimental configurations, no degree of control over the apparatus will allow the exact prediction of a single observation. Instead, exact prediction and measurement are reserved for ensembles and distributions, rather than for individual events, i.e., the definition of "replication" has been subtly changed to accommodate the intrinsic indeterminancy. Similar modifications are routinely applied in the study of dynamical chaos and complex systems, e.g., in fluid mechanical turbulence, granular media, fracture and fatigue processes, etc. Indeed, in any systems sufficiently complex that the validity of statistical limit theorems must be questioned, the concept of empirical replication may need to be modified. In our case, the potential indeterrninancy of various physical outcomes is overlaid with a plethora of potentially relevant biological and psychological variables associated with the human operators and experimenters that may exceed our ability to specify, measure, or detect, let alone to control. To expect that these hypercomplex systems will submit to classical expectations of causality, determinism, and replicability may be overly presumptive. Many attempts to address such mindlmatter replication problems have been advanced in the recent literature. One of the authors (J.H.) previously proposed that failures to replicate frequently occur if a sequence of experiments is interrupted by an overall analysis of the results up to that point. He has termed this the "Meta-Analysis Demolition Effect" and has discussed its psychological and pragmatic implications (Houtkooper, 1994). Others have suggested that better understanding of the limitations on the dynamical replicability of unstable physical systems could benefit mindfmatter interaction research, as well (Atmanspacher, 1997; Atmanspacher & Scheingraber, 2000). It has also been proposed that the lack of dependable reproducibility might be intrinsically related to the appearance of the anomalies, and thus constitutive of our understanding of them (Atmanspacher et al., 1999). Yet another approach has 544 R. Jahn et al. treated all mindlmatter interactions as inherently quantum mechanical in character, and thus prone to the intrinsic quantum uncertainties (Jahn & Dunne, 1986). Many other rigorous and speculative propositions could be cited, but the replication problem remains a central conundrum in this class of research. 7 . Anomaly indicators: Composite "bottom-line" mean shifts in directions of intention would be the primary indicators of anomalous eflect; any structural anomalies would simply be embellishment thereon. While the overall mean-shift criterion is undoubtedly the simplest to specify, evaluate, and promulgate, it is not a particularly informative source for comprehension of any subtle psychophysical processes underlying the phenomenon. In prior work, whether "successful" by the overall mean-shift criterion or not, much more has been inferred from the structural details of the databases, than from their gross characteristics (Dunne, 1998; Dunne et al., 1994; Jahn, Dobyns, & Dunne, 1991 ; Jahn et al., 1997; Nelson et al., 2000). In this PortREG replication program as well, having acknowledged the bemusing failure to replicate the prior scale of "bottom-line" results, we are presented with an impressively deep reservoir of structural features that in their striking internal disparities may testify equally emphatically to a broad variety of operator influences. Just as those studies in human behavior that encompass many heterogeneous groups of people rarely yield results that are universally valid for all participating population subsets, so the broad range of personal characteristics of the operators of these experiments, if relevant at all, could be expected to express themselves in less-than-consistent, variably incoherent forms. In this view, the polyglot nature of the results is not so much paradoxical as it is consistent with, and even supportive of, the hypothesis that some human behavioral characteristic is indeed interacting with the machines. Nor should we ignore the magnitude of this constellation of structural anomalies. Recall that those components encompassed by the Monte Carlo treatment stood out from chance at about the p = .02 level. But the other structural features uncovered in the data, which necessarily required alternative evaluations, contribute further to an overall chance unlikelihood that extends well beyond that. Specifically, Appendix I1 outlines conservative meta-analytical computations that place the composite structural anomalies at a level of chance expectation in the range of 0.001 to 0.002 (two-tailed). This approaches the level of significance that would have been achieved had the overall mean-shift replication been successful. That is, if the average prior PEAR H-LO mean shift, A, had been sustained over the replication database, the corresponding 2-score would have been about 3.60 ( p = .0002, one-tailed). In comparison, the equivalent 2-scores for the structural anomalies in the replication database range from 3.10 to 3.30, depending on the particular analysis base employed (cf. Appendix 11). While these reexaminations of presumptions and retrospective arguments clearly do not resolve our replication paradox, in some respects they may help MindIMachine Interaction Consortium - -- -- - - 545 H (a) First Epoch 10................................. , , \ .-. .-- .- ..,..--' .. B ---.--*---4 . -5 10- ------, ---i-~.------...--.. , ---. --2---------.....---'"\,-/'-/ --C - - '-..-- - - - - - - - _ _ _ _ _ ---------______ \ \ - ',X/-,~-L I I I I 1 - (b)Second Epoch A n -6 - 1 I I I I (c) Third Epoch 5- -5 - ---- -------.--, r(---'----rr-t \ f 50 100 150 Thousands of Trials 200 250 ,---,L &/ 0 Fig. 12. Prior PEAR cumulative deviations in three epochs. 546 R. Jahn et al. to focus suggestions for future research. The change from systematic, intention-correlated deviations to a comparably anomalous, albeit less orderly pattern of structural distortions testifies to our incomplete understanding of the basic phenomena, and warns that future empirical and conceptual efforts must proceed at a more sophisticated level. The next round of experiments and analyses will need to identify and address the implicit as well as the explicit assumptions, both in the initial designs and in the assessment of empirical results, and delve more deeply into the relationship between the anomalous manifestations and the underlying psychological and physical sources from which they emerge. No simpler conceptual route seems likely to prevail, but vigorous and insightful pursuit of this more difficult one not only may ultimately illuminate the particular mindlmachine anomalies under study here but also may provide a much broader view of the relationship of the human mind to all physical reality. Appendix I: PortREG Equipment Calibrations The protocol for the PortREG replication specified that concurrent calibrations be generated at each laboratory to correspond to each experimental session, using the same acquisition software but modified to run automatically. Beyond these, many other ad hoc calibration efforts were undertaken to establish that the REG devices were performing according to specifications and to characterize their performance in finer detail. Typically, the concurrent calibrations were generated following one or more experimental sessions, in blocks consisting of 3000 x 200-bit trials. Most were taken as 1000-trial runs, but some also were collected in 100-trial runs. Each of the laboratories collected more than the specified number of concurrent calibration trials from their respective REG sources. Specifically, GARP and PEAR generated over one million trials and FAMMI more than 850,000 trials. The results are displayed in Tables A1.F, A1.G, and A1.P. The first column of the tables lists the parameters computed in the standard suite of statistical tests for calibrations. Included are the first four moments of the statistical distribution, i.e., the Mean, SD (standard deviation), Skewness, and Kurtosis. The distribution of trial outcomes is compared with theoretical expectation by the standard calculations (x2 Bins), and the standard deviation is calculated for blocks of 100 and 1000 trials (100-tr Sigma and 1000-tr Sigma, respectively). The distribution of runs of consecutive trials scoring greater than 100, and trials scoring 100 or less is compared with theoretical expectation (x2 Runs), and a similar comparison against theory is made of the proportion of runs of length 50 remaining on one side of the origin (Arcsine). Finally, two autocorrelation functions are computed, for the raw trial sequences and for blocks of 50 trials (Autocorr Raw and Autocorr 50). The probability values are computed from the appropriate statistical indicators (2-scores, F values, and x2s). In general, the consistency of the data and the deviations of parameter estimates are in accord with theoretical expectations for independent random bits x2 MindIMachine Interaction Consortium 547 having binary probability of precisely .5, and hence these calibrations confirm the nominal statistical distribution of the overall data. However, a few specific departures from the theoretical distribution, and their implications for analysis of the experimental data, should be noted: 1. One of the most consistent structural departures from expectation in the experimental data occurs in the trial-level standard deviations shown in Table C.11. Thus, it is particularly important to examine the corresponding behavior of the calibration data. None of the three calibration databases shows a significant deviation from the nominal trial-level standard deviation of the appropriate theoretical binomial distribution. Specifically, there is only a slight increase ( p = .215) in the FAMMI calibrations, and a slight decrease in the GARP and PEAR calibrations (p = .66 a n d p = .61, respectively). Therefore, it is valid to pool these values to an empirical standard of comparison for the experimental data, as described in the main text, section 1II.B.5. 2. The FAMMI calibrations show a marginally significant elevation in the trial-level goodness-of-fit x2 test (p = .045), even though all four parameters of the trial-level distribution are nominal. Of greater concern is the fact that the standard deviations of both 100-trial and 1000-trial blocks are significantly elevated (p = .001, p = .012). Since the triallevel standard deviation is nominal, this indicates a nonindependence between trials, which produces increased average deviations at the block lengths used in the actual experiments. Taken at face value, this would suggest that the mean-shift 2-scores emerging from the FAMMI data are exaggerated by as much as 5.5%. (This is obtained by comparing the observed standard deviation of 1000-trial blocks, 235.871, to the theoretical value of 223.607; the ratio, 1.0548, is the factor by which 2-scores would be inflated by this departure from theoretical standard deviation.) The presence of intertrial dependence is confirmed by a significant autocorrelation (p = .005) at the trial level, driven by a succession of large, positive correlations at various lags, especially lags 5, 6, 10, and 12. A breakdown of the FAMMI calibrations into four roughly chronological sections shows that the amplified standard deviation of blocks is primarily in the first half of the data, particularly in the second quarter (series 50 to 99), which show a standard deviation increase as severe as 11.4% in the worst case. [The FAMMI team observed these deviant early calibrations and replaced the original device with a new one. No deeper examination was made, but the difference between the first and second half of the FAMMI calibrations suggests the source of the problem was some subtle malfunction of their REG device.] By any reasonable criterion, these aberrations should have no consequential impact on the primary or secondary FAMMI data, or the interpretation thereof. 3. The GARP calibrations fail of perfection only in being too good, with a for the deviation of the trial distribution so small that 97.6% of ran- x2 548 R. Jahn et al. dom samples would be expected to show greater departures from the theoretical populations. , 4. The PEAR calibrations show an elevated skewness, y = 0.0057, corresponding to Z = 2.48, in the trial distribution. The reasons for this are obscure; a chronological breakdown into 10 segments shows marginally significant positive skewness in Blocks 4, 6, and 7, with an overall bias toward positive skewness. The distribution among the blocks suggests that a small positive skewness is present throughout, with the increased population of significant outliers being a consequence of normal variation about this shifted mean. Since trial-level skewness is a departure from normality, which will be suppressed rapidly in calculations involving large numbers of independent trials, this is not considered a damaging aberration so long as the trials are independent. All of the PEAR results relating to intertrial structure are nominal, suggesting that the trials are indeed mutually independent, despite their distributional oddity. 5. The chronological breakdown of PEAR calibration data suggests the existence of a brief epoch (May through June 1998) during which triallevel standard deviation may have been suppressed. (In this segment, o= 7.0302 and p = .9970. This result remains a p = .03 suppression even after Bonferroni correction for the examination of 10 subsets. Whether further correction for the many other parameters under scrutiny is appropriate here may be left to the individual analyst.) Since this epoch, even if it represents a genuine local suppression of standard deviation, corresponds to a concomitantly small proportion of the experimental data, and since the overall trial-level standard deviation of the calibrations is nominal, the previous remarks and conclusions concerning the 2-scores of Table C. 11 do not need revision. As a supplement to the concurrent trial-level calibrations, GARP also collected bit-level calibration data, to examine the behavior of the REG source at this finer scale. In contrast to the "quality control" approach of the concurrent calibrations, the GARP procedure is a "device properties" approach (Houtkooper, 1998), which examines short-term dependencies as characterized by Markov-chain transition probabilities. These are in straightforward relationship to traditional parameter-based tests, but this alternative allows more specific deviations from randomness to be scrutinized and permits calculation of standard deviations between sections of data and, hence, sensitive detection of episodic deviations from ideal randomness. These bit-level data reveal an expected effect, namely a slight excess of the bit sequences 01 and 10 over 00 and 11. The source of the effect is the design of the REG, which includes an XOR alternating template to eliminate actual physical bias in the threshold setting of the comparator that defines voltage levels as bits. The size of this excess of alternations is on the order of a few parts in 10,000 and is detectable if data sets are accumulated over a few days. (The tests require on the order of 100 million bits.) Of course, the standard de- MindIMachine Interaction Consortium TABLE A1.F FAMMI Concurrent Calibrations (852,000 Trials) Parameter Mean Std.Dev. Skewness Kurtosis Bins 100-tr Sigma 1000-tr Sigma Runs Arcsine Autocorr Raw Autocorr 50 Theory Actual Probability x2 x2 TABLE A1 .G CARP Concurrent Calibrations ( 1,165,000 Trials) Parameter Mean Std.Dev. Skewness Kurtosis Bins 100-tr Sigma 1000-tr Sigma Runs Arcsine Autocorr Raw Autocorr 50 Theory Actual Probability x2 x2 TABLE A1.P PEAR Concurrent Calibrations (1,130,000 Trials) Parameter Mean Std.Dev. Skewness Kurtosis x2Bins 100-tr Sigma 1000-tr Sigma Runs Arcsine Autocorr Raw Autocorr 50 Theory Actual Probability x2 viation of 200-bit trials is affected by interbit structural behavior on scales up to 199-bit sequence length and cannot be predicted reliably from this alternation excess. It is for this reason that the empirical standard deviation estimate 550 R. Jahn et al. from the calibrations, including the empirical uncertainty thereof, was used as the standard of comparison for the statistical measures in Table C.11. Appendix 11: Structural Meta-Analysis The main text introduces, analyzes, and discusses many different structural features of the database, some of which prove to be individually significant, others not. The question to be addressed here is how to compound all such structural evidence into an overall statistical figure of merit. Specifically, the general problem of evaluating a number of distinct analyses on a collective basis is addressed by a meta-analytic technique. It should be noted at the outset that not all of the participating structural analyses enter on an equal footing. Some of them are consequences of other analyses, i.e., they are re-examinations or more detailed investigations of effects that have already been evaluated in the other formats. Also, certain analyses were preplanned while others were retrospective. Moreover, while most of the analyses are based on the entire three-laboratory database, others are restricted to only single-laboratory data. The following numbered list introduces each of the structural analyses in the order they are encountered in the main text, describing its status in terms of the foregoing factors and providing any additional information required to specify how the conclusion of that specific analysis is reached. A probability value (p) is quoted for each analysis, to facilitate meta-analytic combination via the method of adding logarithms (Rosenthal, 1984). 1. The breakdowns by secondary parameters presented in Tables F.2 through F.5, G.2 through G.5, and P.2 through P.5 comprise a preplanned structural analysis, i.e., examination of these parameters was part of the original experimental design. While this is a complex calculation with many subparts, it has been collectively evaluated against the null hypothesis by the Monte Carlo analysis of section III.B.2, resulting in p = .022. 2. The series-position results, presented in Tables F.6, G.6, and P.6, constitute another preplanned analysis. The summaries in Table C.9 result in p = .026 after Bonferroni correction for including separate results from each of the three laboratories. (Only the rightmost column of Table C.9 is relevant, since the raw would respond to overall mean shifts, if any.) 3. Table F.7 reports a preplanned exploration of experimenter effects conducted only at FAMMI; combining the independent values results in p = .887. 4. Table G.7 reports a preplanned examination of control mode conducted only at GARP; constructing a from the independent 2-scores yields p = .684. x2 x2 x2 x2 MindIMachine Interaction Consortium 551 5. Table G.8 reports a preplanned examination of operator types conducted only at GARP; the composite corresponds t o p = .241. 6. Following Table G.8, a few summary figures, and the discussion of the Monte Carlo analysis noted in Item 1 above, the next analysis in the text is the discussion of "favored cells" in section III.B.2.b. This retrospective analysis examines internal features of the structural qualities which have already been evaluated against the null hypothesis in Item 1. Although ap-value of .274 can be computed for this (by applying a Bonferroni correction to the most striking Z-score reported), it cannot properly be included in the meta-analysis. 7. In contrast, the correlation coefficients reported in the second half of Table C.7 are a retrospective examination of a different phenomenon. Such correlations between laboratories are independent of the Monte Carlo evaluation. After Bonferroni correction this yields p = .243. 8. The retrospective examination of unconfounded secondary parameters (Table C.8), like Item 6, is a direct consequence of the structural elements analyzed in Item 1. It produces p = .031 after Bonferroni correction but cannot properly be included in the meta-analysis. 9. Table P.7, presenting the evaluation of individual operator consistency between experiments, is an independent retrospective analysis, albeit one limited to a single laboratory (PEAR). This yields p = .011 after Bonferroni correction. 10. The summary of PEAR operator-specific performances presented in Table C. 10 also qualifies as a preplanned analysis requested by certain of the authors. It yields p = .05 1 after Bonferroni correction. 11. The discussion following Table C. 10 mentions, but does not report, a similar analysis based on a calculation for the series-level, rather than operator-level, data. This was a retrospective analysis that detects the same structural properties as the operator-specific analysis and must therefore be regarded as a derivative of Item 10; its p-value of .021 must therefore be excluded. 12. The trial-level standard deviation results in Table C. 11 follow from a retrospective analysis that is independent of all previous analyses, with p = .049 after Bonferroni correction. 13. The counts of successful operators and series, mentioned in the last subsection of section 111, are consequent to and dependent on the mean shifts. An earlier version of the Monte Carlo analysis incorporated these along with the mean-shifts and proved statistically consistent with the results of Item 1; thus, we may quotep = .022 for this but must consider it a consequent analysis and exclude it from the meta-analytic combination. 14. The previous 13 items cover all of the analyses presented in sections I1 and 111, but for completeness, we must note one other independent retrospective analysis that was not included in the text. From earlier PEAR experience, it was speculated that the trial-level variance might be re- x2 x2 552 R. Jahn et al. duced in runs that were successful in the direction of intention, relative to its value in those runs contrary to intention. The calculatedp = .234. Table A2.1 summarizes these 14 analyses, now organized by category. The index numbers in the left margin of the table refer to the itemized list above. To compound the results of a set of analyses individually reported as p-values, we may take advantage of the fact that under the null hypothesis p is uniforrnly distributed between 0 and 1, whence -2 lo&) is distributed as a with 2 degrees of freedom (dB. The addition properties of then guarantee that a sum of n such values is a with 2 2 df (Rosenthal, 1984). 1 Considering first only the preplanned analyses that incorporate the entire database, we have pi = {.022, .026, .051 ). This results in = 20.885 on 6 df, yielding a composite meta-analytic p = .0019. Adding the three retrospective analyses that cover the entire database increases this to 32.65 1, now on 12 df so the meta-analysis reaches p = .0011. Finally, including the four analyses to 45.538 and df to 20, based on single-laboratory contributions increases yielding p = .0009. Thus while the various analyses, which might be considered questionable due to retrospective status or limitation to a single laboratory, increase the statistical significance, they do so only by a factor of 2 from the initial figure for preplanned, whole-database analyses. Including retrospective analyses raises the issue of the "file-drawer effect," where the visible results might spuriously overestimate an effect by overlooking an unreported background of null results. The standard measure for considering the possible impact of unreported studies is the number of such studies, x2 x2 x2 x2 x2 x2 TABLE A2.1 Summary of Analyses Item Preplanned; using all data 1. Secondary parameters 2. Series position 10. Operator performance Retrospective; using all data 7. Interlab correlation 12. Trial-level o 14. Success-based o Preplanned; single-laboratory data 3. Experimenter effects 4. Control mode 5. Operator type Retrospective; single-laboratory data 9. Operator consistency Reanalysis of effects already analyzed 6. Favored cells 8. Unconfounded parameters 11. series x2 13. Operator and series counts Form of analysis Monte Carlo Independent Independent Independent Independent Independent Independent; FAMMI only Independent; GARP only Independent; GARP only Independent; PEAR only Consequence of (1) Consequence of (1) Consequence of (10) Consequence of (1) p-value MindIMachine Interaction Consortium 553 with null outcomes, that would need to be added to the reported database in order to reduce the overall result to nonsignificance. For the current result, this file-drawer number is 14. Given the difficulty of finding any other new and substantive analyses that are not in some way reexaminations of structural aspects already considered, and given that this file-drawer number is equal to the total number of analyses already reviewed, including several such "duplicates" (6,8, 11, and 13), it would seem that there is little risk of file-drawer dilution of this survey statistic. In conclusion, the aggregate interpretation for the PortREG analyses with all multiple-testing and redundancy concerns taken into account is p = .0009 against the null hypothesis that the data contain no anomalous structures, or p = .0019 if only preplanned complete-data analyses are included (which has the virtue of rendering file-drawer considerations completely moot). Acknowledgments The authors acknowledge with deep gratitude the financial support of the Institut fiir Grenzgebiete der Psychologie und Psychohygiene, which allowed this Consortium to be formed and this research project to be accomplished. In the interpretation of the experimental data and its dialogue with theoretical models, and in critical editing of this report, our consultation with Dr. Harald Atmanspacher and his colleague Dr. Werner Ehm have been invaluable. We also express our thanks to the many operators who generated the experimental data and to the many staff persons who helped in implementing these studies and this report, most especially Ms. Lisa Langelier-Marks and Ms. Elissa Hoeger. References Atmanspacher, H. (1997). Dynamical entropy in dynamical systems. In Atmanspacher, H., & Ruhnau, E. (Eds.), Time, temporality, now (pp. 327-346). Berlin: Springer. Atmanspacher, H., Bijsch, H., Boller, E., Nelson, R. D., & Scheingraber, H. (1999). Deviations from physical randomness due to human agent intention? Chaos, Solitons, and Fractals, 10(6), 935-952. Atrnanspacher, H., & Scheingraber, H. (2000). Investigating deviations from dynamical randomness with scaling indices. Journal of Scientific Exploration, 14(2), 1-18. Bierman, D. J., & Houtkooper, J. M. (1975). Exploratory PK tests with a programmable highspeed random number generator. European Journal of Parapsychology, 1(1), 3-14. Bierman, D. J., & Houtkooper, J. M. (1981). The potential observer effect or the mystery of irreproduceability. European Journal of Parapsychology, 3(4), 345. Braud, W. G. (1993). On the use of living target systems in distant mental influence research. In Coly, L. (Ed.), Psi research methodology: A re-examination. New York: Parapsychology Foundation. Braud, W. G . , & Dennis, S. P. (1989). Geophysical variables and behavior: LVIII. Autonomic activity, hemolysis, and biological psychokinesis: Possible relationships with geomagnetic field activity. Perceptual and Motor Skills, 68, 1243-1 254. Dunne, B. J. (1991). Co-operator Experiments with an REG Device (PEAR Technical Report No. 91005). Princeton, NJ: Princeton Engineering Anomalies Research, Princeton University, School of EngineeringIApplied Science. 554 R. Jahn et al. Dunne, B. J. (1998). Gender differences in humantmachine anomalies. Journal of Scientific Exploration, 12(1), 3-55. Dunne, B. J., Dobyns, Y. H., Jahn, R. G., & Nelson, R. D. (1994). Series position effects in random event generator experiments, with appendix by Angela Thompson. Journal of Scientific Exploration, 8(2), 197-21 5. Dunne, B. J., & Jahn, R. G. (1992). Experiments in remote humantmachine interaction. Journal of Scientific Exploration, 6(4), 311-332. Dunne, B. J., Nelson, R. D., & Jahn, R. G. (1988). Operator-related anomalies in a random mechanical cascade. Journal of Scientific Exploration, 2(2), 155-179. Grad, B. (1963). A telekinetic effect on plant growth. International Journal of Parapsychology, 5(2), 117-133. Houtkooper, J. M. (1994). Does a meta-analysis demolition effect exist? Abstracts of 18th International Conference of the Society for Psychical Research, September 2-4 (pp. 14-15). Bournemouth, UK: Society for Psychical Research. Houtkooper, J. M. (1998). IGPP Mind-Machine Interaction Consortium: Giessen Anomalies Research Program, MMItPortREG Replication Phase 1. GARP Technical Report, Draft 19981001. Giessen, Germany: Center for Psychobiology and Behavioral Medicine, JustusLiebig-Universitat Giessen. Jahn, R. G., & Dunne, B. J. (1986). On the quantum mechanics of consciousness, with application to anomalous phenomena. Foundations of Physics, 16(8), 721-772. Jahn, R. G., Dobyns, Y. H., & Dunne, B. J. (1991). Count population profiles in engineering anomalies experiments. Journal of Scientific Exploration, 5(2), 205-232. Jahn, R. G., & Dunne, B. J. (1988). Margins of reality: The role of consciousness in the physical world. New York: Harper Brace Jovanovich. Jahn, R. G., & Dunne, B. J. (1997). Science of the subjective. Journal of Scientific Exploration, 11(2), 201-224. Jahn, R. G., Dunne, B. J., Dobyns, Y. H., Nelson, R. D., & Bradish, G. J. (2000). ArtREG: A random event experiment utilizing picture-preference feedback. Journal of Scientific Exploration, 14(3), 383-409. Jahn, R. G., Dunne, B. J., &Nelson, R. D. (1987). Engineering anomalies research. Journal of Scientific Exploration, 1(1), 21-50. Jahn, R. G., Dunne, B. J., Nelson, R. D., Dobyns, Y. H., & Bradish, G. J. (1997). Correlations of random binary sequences with pre-stated operator intention: A review of a 12-year program. Journal of Scientific Exploration, 11(3), 345-367. Nelson, R. D., Bradish, G. J., & Dobyns, Y. H. (1989). Random event generator qualification, calibration and analysis (PEAR Technical Report No. 89001). Princeton, NJ: Princeton Engineering Anomalies Research, Princeton University, School of EngineeringIApplied Science. Nelson, R. D., Bradish, G. J., Dobyns, Y. H., Dunne, B. J., & Jahn, R. G. (1996). FieldREG anomalies in group situations. Journal of Scientific Exploration, 10(1), 1 11-141. Nelson, R. D., Bradish, G. J., Jahn, R. G., & Dunne, B. J. (1994). A linear pendulum experiment: Operator effects on damping rate. Journal of Scientific Exploration, 8(4), 47 1-489. Nelson, R. D., Dobyns, Y. H., Dunne, B. J., & Jahn, R. G. (199 1). Analysis of variance of REG experiments: Operator intention, secondary parameters, database structure (PEAR Technical Report No. 91004). Princeton, NJ: Princeton Engineering Anomalies Research, Princeton University, School of EngineeringtApplied Science. Nelson, R. D., Jahn, R. G., Dobyns, Y. H., & Dunne, B. J. (2000). Contributions to variance in REG experiments: ANOVA models and specialized subsidiary analyses. Journal of Scientific Exploration, 14(1), 73-89. Nelson, R. D., Jahn, R. G., Dunne, B. J., Dobyns, Y. H., & Bradish, G. J. (1998). FieldREG 11: Consciousness field effects: Replications and explorations. Journal of Scientific Exploration, 12(3), 425-454. Peoc'h, R. (1995). Psychokinetic action of young chicks on the path of an illuminated source. Journal of Scientific Exploration, 9(2), 223-229. Radin, D. I. (1997). The conscious universe: The scientific truth ofpsychic phenomena. San Francisco, CA: HarperEdge. Radin, D. I., & Nelson, R. D. (1989). Consciousness-related effects in random physical systems. Foundations of Physics, 19(12), 1499-15 14. Rhine, J. B., & Humphrey, B. M. (1944). The PK effect: Special evidence from hit patterns. 11. Quarter distributions of the set. Journal of Parapsychology, 8, 287-303. MindIMachine Interaction Consortium 555 Rosenthal, R. (1963). Experimenter attributes as determinants of subjects' responses. Journal of Projective Technique and Personality Assessment, XXVIZ, 324-33 1. Rosenthal, R. (1984). Meta-analytic procedures for social research. Beverly Hills, CA: SAGE Publications. Schlitz, M. J. (1986). An ethnographic approach to the study of psi: Methodology and preliminary data. Proceedings of presented papers, The Parapsychological Association 29th Annual Convention. Rohnert Park, CA: The Parapsychological Association, 187-204. Schmidt, H., & Pantas, L. (1972). Psi tests with internally different machines. Journal of Parapsychology, 36, 222-232. Schmidt, H. A. (1970). A quantum mechanical random number generator for psi tests. Journal of Parapsychology, 34, 219-224. Schmidt, H. A., Morris, R., & Rudolph, L. (1986). Channeling evidence for a PK effect to independent observers. Journal of Parapsychology, .50(1), 1-16. Shapin, B., & Coly, L. (Eds.) (1985). The repeatability problem in parapsychology. Proceedings of the 32nd International Conference of the Parapsychology Foundation, held in San Antonio, Texas, 1983. New York: Parapsychology Foundation. Journal of Scientijic Exploration, Vol. 14, No. 4, pp. 557-570,2000 0892-3310100 O 2000 Society for Scientific Exploration Unusual Play in Young Children Who Claim to Remember Previous Lives Division of Personality Studies University of Virginia Health System P 0. 8001 52 Box Charlottesville, VA 22908-0152 Abstract- Among 278 cases of children who claim to remember previous lives, 66 (23.7%) engaged in play that was unusual for their families and had no model in family members or other obvious normal stimulus. This paper reports 25 examples of such atypical play. The play accorded with claimed memories of previous lives expressed by the children when they could speak. The child's unusual play sometimes gave its parents the first indication they had that the child was possibly remembering a previous life. In 22 cases the child's statements were found to match events in the life of a specific deceased person. In such cases the play was also found to correspond to some aspects of that deceased person's life, such as his or her vocation, avocation, or mode of death. Keywords: Play - previous lives - behavioral memories - reincarnation - memories of trauma Introduction The concept of play has received much attention from psychologists and psychiatrists, several of whom have offered general explanatory theories of play. In the late 19th century, Lazarus (1883) wrote that play resulted from the need of human beings for constant activity, which he described as a primary impulse. For him, "Activity is life ... Its opposite is nothingness, emptiness" (p. 45; my translation). According to this view, if we have no other activity, we invent one and call it play. Freud (192011961) considered play an attempt to master a stressful situation, and the play of children following a significantly traumatic event may have this motive: the reenactment in play supposedly reduces the negative emotional accompaniment of memories of the trauma. Later theorists of play have emphasized its value in developing physical capacities and physical and cognitive skills (Bornstein et al., 1996; Cotton, 1984; Lillard, 1993; Millar, 1968; Vygotsky, 1978). The kitten that chases a ball learns and practices the skill needed to chase a mouse; similarly, a child given a toy automobile can learn with it the rudiments of driving a real automobile. I think it is fair to say that most investigators of play in children have sought 558 Stevenson for some general principle explanatory of all play. Few have concerned themselves with the question of why a child selects one theme for its play rather than another. This, however, is an old question. Commenting on Lazarus's assertion, mentioned earlier, that play derived from an ineluctable need for activity, William James wrote "No doubt this is true, but why the particular forms of sham occupation" (James, 1890, vol. 2, p. 429; my emphasis). Later investigators of play have ignored James's pertinent question, with three exceptions. First, there are familiar instances in which a child in its play imitates a parent or an older sibling; a common example is a female child's play at "keeping house" in imitation of her mother. Second, boys between 1 and 2 years old show different preferences for the objects of their play than girls of the same age (Fagot, 1974; Jacklin, Maccoby, & Dick, 1973). Also, children with gender identity disorder usually show preferences for the play that is more common with members of the opposite sex (Doering et al., 1989; Rekers & Morey, 1990). Third, children who have undergone some severe trauma frequently seem to reenact the stressful experience in their play (Saylor, Swenson, & Powell, 1992; Terr, 1981, 1988, 1991). This article contributes to the understanding of why a child's play develops around one theme instead of another. It reports a variety of unusual types of play in young children: that corresponding to the statements about previous lives made by some young children, usually between the ages of 2 and 5. Children who claim to remember previous lives can be found in almost all countries, including European countries (Stevenson, 1987) and the United States (Stevenson, 1983a), although they are more readily discerned in some parts of the world, such as South Asia, than in others. Such children, on average, begin to speak about a previous life at the age of about 2 and continue until the age of about 5 (Cook et al., 1983; Stevenson, 1987). Early investigators of these cases gave attention almost exclusively to the questions of, first, whether the child's statements referred to a particular deceased person and no other person, and, second, whether in at least some cases the child and its family had no prior knowledge of the concerned deceased person (Stevenson, 196611974, 1987). (For convenience only, I sometimes refer to this person as "the previous personality.") In recent years, my colleagues and I have extended the studies of these children to other features of their personalities compared with those of their peers who do not claim to remember a previous life (Haraldsson, 1995, 1997) and to a wide range of behavior shown by the children that is unusual in the child's family, but appropriate for the person whose life the child claims to have lived (Stevenson, 2000). Such behavior includes a variety of likes and dislikes for particular persons, foods, clothing, climates, and places. In an earlier article I described and discussed the phobias among these children; among a group of 387 children 141 (36%) showed a phobia, nearly always corresponding to the mode of death in the claimed previous life (Stevenson, 1990). In the present article I describe, with examples, another type of unusual be- Play and Previous Lives 559 havior that these children often show: play that seems to have no contemporary model in the child's life or other normal explanation. I am not suggesting that the play of these children adds importantly to the evidence that reincarnation is the best interpretation of their cases. Their play is, however, consistent with that interpretation. Methods Selection of Cases for This Report In order to obtain an estimate of the frequency of play in the cases, 278 cases, which have been published (by myself), were examined. Of these, 225 were reported in Stevenson (1997) and the remainder in earlier works, mostly of detailed case reports (Stevenson 196611974, 1975, 1977, 1980, 1983b, and 1987). The total of 278 provided a denominator of all investigated cases studied and previously reported by myself. I did not count as play instances in which the child practiced a style of worship that was wrong for its family, but correct for the previous personality. For example, I did not include cases in which a Hindu child in India who claimed to remember a previous life as a Moslem practiced namaz, the Moslem style of worship. If this kind of unusual behavior occurred in a Western family, it might well be considered play. Also excluded were instances in which the play, although appropriate for the claimed previous life, had a known model in the child's family or near neighbors. The importance of this exclusion arose in connection with play as a soldier and with kite-flying, both of which types of play children enjoy in many parts of the world. Nevertheless, the details and circumstances of some of the play at being a soldier deserve mention. The cases of the pool were investigated principally through interviews with multiple firsthand informants, first on the side of the child and then on that of the deceased person, if one has been found whose life corresponded to the child's statements. Whenever attainable, printed documents, such as birth and death certificates, identity cards, and medical records, were consulted and copied. The cases were analyzed individually for competing interpretations, such as fraud, ordinary social knowledge, and paranormal cognition on the part of the child. Series of cases have also been analyzed for recurrent features and variations in the features of the cases in different cultures (Cook et al., 1983; Stevenson, 1986). I have given fuller details of methods elsewhere (Stevenson, 1966/1974, 1975,1997). In this report I do not give information about whether some of the children showed knowledge about a previous personality that they could not have obtained normally. Readers interested in that aspect of the cases may find details bearing on it in the references to more detailed reports that I cite. Here I wish only to draw attention to the close correspondence in many cases between the child's statements about a previous life and its unusual play. Accordingly, I 560 Stevenson will state, for simplification only, that a child "remembers a previous life," not that it "claims to remember" one. At the same time, readers should understand that the play I shall describe occurred in the context of a case with many other features, including evidence in the verified cases that the child showed substantial knowledge about a particular deceased person, which knowledge, in many instances, it could not have obtained normally. My colleagues and I have adopted the convention of describing cases in which the child's statements were verified as correct for a particular deceased person as "solved" (S); and we describe cases where this has not been possible as "unsolved" (US). Asserting that a case is solved does not exclude the possibility that the child could have acquired its correct information normally; that could have happened when the child and the person about whom it speaks belong to the same family or even the same community. There are, however, numerous cases in which we can exclude such normal transmission of information with reasonable certainty (Haraldsson, 1991; Mills, Haraldsson, and Keil, 1994; Stevenson, 196611974, 1975; Stevenson and Samararatne, 1988a, 1988b.) Social Circumstances of the Families Concerned Nearly all the cases occurred among families of Asia who lived in villages or small towns. This means that for the period when most of the cases developed (between 1960 and 1985) the child and its family had no access to television by which the child might have become acquainted with a model for the unusual behavior it showed. It cannot be excluded that in some cases a model for the behavior, for example, that of a shopkeeper, might have existed in the village or town where the child lived even though not in the child's own family. In each instance the child's behavior was unique among the play of the children of its family. Results Estimated Prevalence of Play in a Series of Cases Among the 278 cases selected, the feature of unusual play was reported in 66 (23.7%) cases. This is almost certainly a minimal figure of the prevalence of play in the cases. The Registration Form that we use in investigating these cases includes a checklist of features about which we would like to learn. It may happen, however, that the informants may not mention play that they observed in a child, and the investigator may also sometimes have failed to inquire about it, even though the checklist includes an inquiry about play. Examples of Unusual Play More than half the examples of unusual play that follow are derived from the 278 cases mentioned above. To these I have added some additional exam- Play and Previous Lives 561 ples from cases in our files of which neither I nor my colleagues have yet published reports. After each example I will add the reference to a published report, if there is one. I have myself investigated all the cases from which I draw my examples. After each example, I have added the symbols (S) and (US) to indicate whether the case is solved or unsolved. For many cases I have no information about how long the play persisted during the child's early years. For the cases where this is reported it generally occurred during the period of the child's most active speaking about the previous life and ceased when he or she stopped talking about it, which usually happened between the ages of 5 and 7 (Cook et al., 1983). In a few instances the play lasted longer. In five cases the child's play gave its parents their first indication that the child might be remembering a previous life. I have included two examples of this group in the present report; in another case of this report the child's play was one of the first indications-but not the very first-of its having memories of a previous life. The types of play observed corresponded to several features in the life and death of the previous personality. The most numerous feature was the previous personality's profession or vocation, and I describe 17 examples of this. Less frequently the child engaged in play that was typical of the play of the opposite sex (shown by children who claimed to remember a life as a person of the opposite sex) and in play that corresponded to the avocations or hobbies of the previous personality. Another small group of children named their dolls or other play objects after the previous personality's children. In a fourth small group the child reenacted the mode of death of the previous personality. I describe two examples of each of these four smaller groups. Play Corresponding to Vocations ofthe Previous Life The most frequent category of play was that of reenacting the vocation of the previous personality. Among the examples of this are the following: Shopkeeper. P.S. was the son of a professor in a college in Bisauli, a small town of northern India. P.S. remembered the life of a prosperous businessman who had owned shops, the most important of which was one (in the city of Moradabad) where biscuits were manufactured and soda water produced (Stevenson, 196611974) (S). At the age of about 2 112, P.S. began making models of what appeared to be shops, with wires running around them. He made "biscuits" from mud and served these with "tea" (actually water). He began to talk of soda water. While engaged in this kind of play, he gradually described a previous life in which he had owned a shop where biscuits and soda water were provided for the customers. (At that time in India bottled soda water was not readily available; it was mainly found in specially equipped shops that made it and sold it directly to customers.) Tea would almost certainly have also been offered at such a shop. P.S. played little with other children; he became so totally absorbed in his play at managing a biscuit and soda water shop that he fell 562 Stevenson markedly behind at school. His mother blamed his failures at school-which restricted his later vocational opportunities-to the time he had lost in childhood when he was playing at shopkeeping. In Bisauli, where P.S. lived, biscuits would have been sold in some shops, but it would have had no shop where biscuits were made and soda water produced. S.K., a young girl of Burma, was the daughter of a cotton farmer. (Myanmar is now the name of the country formerly called Burma; but for the period when most of these cases were investigated the country was called Burma.) She remembered the life of a woman who had sold pickled tea, a stimulant much appreciated in Burma (unpubl. data) (S). When she was young, S.K. played at setting up a stall from which she sold pickled tea and also dried tea leaves. She would not play at other games and would not change the commodity she sold. Schoolteacher. L.A. was a young girl of Sri Lanka who began speaking about the previous life of a housewife and schoolteacher at the age of about 2 112 (Stevenson, 1977) (S). She began to play at being a schoolteacher when she was about 3 years old, before she had seen any adult teaching. (Her father was an instructor in a carpentry school.) She would wrap herself in a garment to imitate a teacher's sari. Then, using a cane as a pointer and a door for a blackboard, she would teach the imagined class. She would ask the members of the class to give her their copybooks. L.A. did not draw other children into her teaching, but acted the scene alone. Her play at teaching school was continuing at the age of 5 112, by which time she had begun school herself. Nightclub owner. E.K. was the son of an unskilled laborer in Adana, a city of south central Turkey. He remembered the previous life of a man who owned a nightclub in Istanbul (Stevenson, 1980) (S). As a young child, E.K. regularly played at managing a nightclub. He arranged boxes to represent a bar and placed bottles on them. He assigned roles in the club to neighborhood girls and gave one a stick that represented the microphone held by a singer. He put out two chairs for the wives of the club's owner. (Polygamy was then illegal in Turkey but still practiced by some men, including the man whose life E.K. remembered. He had had two wives, although we may doubt whether he ever brought them together in his nightclub.) Operator of a flour mill. V. was the son of a small farmer in northern India. As a young child, V. remembered the life of a prosperous owner of a flour mill (unpubl. data) (S). When he was about 2 years old, he was playing with sand. He erected what looked like a flour mill with sand and said to his grandmother: "Bring some grain for grinding." This was his family's first indication that V. was remembering a previous life about which, on being encouraged to say more, he gave numerous details. Physician. V.R. was the son of a businessman of northern India. He remembered the life of a physician, Dr. S.S.D. (unpubl. data) (S). Like many physicians in India, Dr. S.S.D. had a shop where he both examined patients and sold the medicines that he prescribed. As a child V.R. played at being a doctor. He set up a clinic shop with bottles and a thermometer. He would take tempera- Play and Previous Lives 563 tures with a stick and then shake down the stick as one shakes a medical thermometer to lower the level of the mercury. He examined his playmates as patients. I did not learn V.R.'s age when he engaged in this sort of play. One informant told us that he continued with it for about a year. Many years later, in an interview, V.R. recalled his play in childhood at being a doctor. He said that a woman he knew had a high fever, and he mixed salt and pepper into water and "prescribed" this for the patient; she took it and recovered. Well digger. M.S. was a young boy of Lebanon whose father was a small farmer and collector of pine cones for their seeds. M.S. remembered the previous life of a man who had been a professional well digger (unpubl. data) (S). He had died when a heavy stone that was being raised out of a partially dug well slipped out of the lifting basket and fell on his head. M.S.'s mother observed him playing at digging a well in sand. I did not learn other details of this play or how long it lasted. Garage mechanic. D.J. was the son of a technician at a radio station in Lebanon. As a young child, D.J. remembered the previous life of an automobile mechanic (unpubl. data) (S). When he was about 2 1/2, he began to speak some names that his parents did not recognize. During the next year he said that he was from a place called Kfermatta, and he spoke about a vehicular accident near the sea. His parents had still not identified D.J.'s statements with the life and death of anyone they knew when they observed him lying under furniture, such as a sofa, and seeming to be unscrewing something. His family did not understand this behavior and became concerned that he would damage the furniture. When they told him to stop, he replied imperturbably: "I am working." They only understood his behavior when he was able to give enough information to permit them to learn that the previous life he was remembering was that of an automobile mechanic who had worked in a garage in Beirut. Vehicle driver. V.M. was the son of a farmer of northern India. When he could speak, he began to make statements about the life of a tonga (pony-cart) driver, Kallu, whom his father had known (unpubl. data) (S). V.M., at this age, would put a towel on his shoulder, as tonga drivers in India do, take a piece of cord to serve as reins, and act as if he were guiding a horse. While playing in this way he would say "tic, tic" repeatedly, in imitation of the sound the tonga drivers make when they wish to signal to pedestrians that they are coming. They do this by letting their whip strike against the spokes of the tonga's wheels, thus producing the repetitive sound that V.M. was imitating. At such times V.M. would also say "I am a tonga driver." Once he remarked: "I used to charge 112 a rupee; now I will charge a rupee." (This presumably referred to the charge for taking passengers from the railway station to their homes, which Kallu had done with his tonga.) Washerman. G.N. was the son of an Ayurvedic physician of northern India. The family were Brahmins. G.N. remembered a previous life as a washerman, who are low-caste persons in India (Stevenson, 1997) (US). As a young child, when G.N. saw his mother washing clothes, he would offer to help her, saying 564 Stevenson "I will do the ironing." At other times he said: "Give me your clothes. I will wash them for you." He became such a nuisance with his insistence on participating in the washing that his mother in exasperation beat him to make him go away. Sometimes he would talk to himself and play at being a washerman. His mother would hear him say: "My wife is sitting here and cooking food, and I am washing the clothes." Nun. T.T.A., a young girl of Burma, remembered the previous life of a Buddhist nun (unpubl. data) (S). During her early childhood, up to the age of 4 or 5 she would play at being a nun. She would put a tray on her head, walk around, say that she was a nun, and ask for alms. (Buddhist nuns do not carry the black bowls that the monks carry to receive the rice or other food that householders give them; they may carry what they are given on a tray.) Sweepress. Sweepers, the persons who sweep the streets, clean latrines, and dispose of garbage, belong to the lowest classes (castes) in India. I have investigated two cases in which children of middle-class families remembered the lives of sweepresses, and both played at being a sweepress. I include here one of these. S.L. was a young girl of northern India who happily cleaned up the stools of her younger brothers when they defecated in the house (unpubl. data) (S). She was a sister of G.N. (mentioned above, who played at being a washerman). Their family were Brahmins. S.L. liked to clean the house, saying "We used to do this work." At times she took a broom and swept the floor. She also made a broom for herself, using twigs and neem leaves, and would sweep with it. Sometimes she dressed herself in a petticoat, put a scarf on her head, and carried a basket. Asked what she was doing, she would say: "I am a sweepress." (Sweepers typically wear a scarf that they can pull over their noses when cleaning latrines; and the basket is to convey refuse that cannot be swept away.) Bandit. C.F. was a young boy of Turkey who remembered the life of a bandit, Cemil Hayik, who had committed suicide rather than be captured (and probably executed) by the police (Stevenson, 1997) (S). As a young child C.F. threw stones at soldiers and policemen. He played with a stick as if it were a rifle. Soldier. I have already said that play at being a soldier is so common among boys of all countries that we should not consider it unusual. Our series includes nine children who claimed to remember previous lives as soldiers and who, when young, played at being a soldier. Most of these might be explained as imitation of soldiers they had seen or learned about normally. Nevertheless, other features in four of these cases make them deserving of attention, and I give one example of this subgroup. B .B. was born in Bareilly, Uttar Pradesh, India, in 1918. He had diminished pigmentation of the skin and hair to the extent that he should be considered an albino. As a young child he said that he had been a soldier named Arthur and had been killed in the "German War" (World War I) (Stevenson, 1997) (US). Play and Previous Lives 565 He had numerous Western behavioral traits. From the age of about 3, he played at being a soldier. He would give military commands, such as "Left, right" and "March." He would pretend that a stick was a rifle, and he asked to have a gun. I cite his case here because his parents were Indians who could not speak English. His father was a scrivener. No one can imagine that they or anyone in their circle would have inculcated or encouraged B.B.'s play at being a soldier. The British Army had a cantonment in Bareilly for many years, and soldiers stationed there had fought and been killed in Europe during World War I. B.B. seems to have been referring to the life of a professional officer in the British Army. Bomber pilot. C.E. was born in Middlesbrough, England. When he became able to speak he said "I crashed a plane through a window." He gradually stated additional details. He said that he had been the pilot of a Messerschmitt and had been on a bombing mission. When he was between 2 and 3 years old he began to draw insignias and badges of military uniforms. These were at first crude, but improved as he became older. He drew an airplane with a swastika on it. He demonstrated the Nazi salute with his arm straight and raised and the German goose-step march. His classmates at school mocked him and he gradually stopped speaking about a previous life (unpubl. data) (US). For several of the vocations illustrated above we have observed additional examples. Also, limitations of space prevent the inclusion here of examples of children who showed in their play the following occupations corresponding to the vocations of the presumed previous life: mason, policeman, building constructor, elephant driver, monk, and snake charmer. Play at Parenting S.G. was a young girl of India who remembered the life of a woman who had died, leaving an infant daughter called Minu (Stevenson, 196611974) (S). The woman's last words before dying, spoken to her aunt, were: "Who will look after Minu?" (The aunt said that she would look after Minu.) When S.G. was about 1 112 years old and barely able to talk, she was observed cradling a block of wood or a pillow and addressing it as "Minu." Someone thought to ask her who "Minu" was, and S.G. replied: "My daughter." Thereafter, she stated additional details of the life of a young mother who had died while her daughter, Minu, was still an infant. S.G.'s play was her family's first indication that she was remembering a previous life. I.A. was a young girl of Lebanon who remembered the previous life of a woman, Selma, whose husband shot and killed her 5 days after she had given birth to a boy called Gandi (unpubl. data) (S). When she was a young child, I.A. used to hold a doll to her breast as if the doll were suckling milk from her. She called the doll Leyla, which was the name of one of Selma's daughters. One day her family found her missing and traced her to a neighbor's home, where there was a boy who happened to be called Gandi. I.A. said that she wanted to feed Gandi from her breasts. 566 Stevenson Play Appropriate for the Sex of the Previous Life Nearly all the children who claim to remember a previous life as a person of the opposite sex engage in cross-dressing when they are young children. I do not subsume this behavior under play. Reports that a girl played "rough like a boy" also seem insufficient to count as examples of play appropriate for the opposite sex. I do count as such examples (a) a distinct or exclusive preference for the play of the opposite sex and (b) a preference to play with persons of the opposite sex. R.K. was a young girl of Sri Lanka who remembered the life of a boy who drowned in a well when he was a little more than 7 years old (Stevenson, 1977) (S). When R.K. was young, she showed a preference for boys' activities such as kite-flying and cadju, which is a game somewhat like marbles as played in the United States. She was skillful at these games. R.K. also joined boys when they played cricket. She rode her brother's bicycle, and, perhaps most masculine of all, at least for Sri Lanka, she climbed trees. A.P. was a young girl of Thailand who, like R.K., remembered the life of a young boy who had drowned (Stevenson, 1983b) (S). When she was young, A.P. liked boys' games and sports such as boxing. Boxing is associated with males everywhere, but nowhere more than in Thailand, where the rules permit the use of elbows, knees, and feet. At a follow-up meeting that I had with A.P. when she was 15 years old, she told me that she still engaged in boxing. Naming Dolls or Other Objects After the Previous Personality's Children or Other Relatives In the section above on play at parenting, I mentioned that S.G. and I.A. had given a block of wood and a doll respectively the names of a daughter of the woman whose life each remembered. We have studied five other examples of this behavior, and I will mention two of these. S.B. was a young boy of Syria who remembered the life of a relative, Said (Stevenson 196611974) (S). The names of Said's seven children were almost the first words that S.B. spoke. When he was still extremely young-I did not learn his exact age-he found five eggplants and two potatoes. He then gave the eggplants the names of Said's five sons and the potatoes the names of his two daughters. If anyone touched these vegetables, he became angry, and he wanted to keep them indefinitely. H.I. was a young girl of Lebanon who remembered the life of a woman called Wadad, who had had five children (unpubl. data) (S). When she was still a young child, her mother bought for her a small toy coffee mill. This had figures of three persons on its top. H.I. gave these figures the names of three of Wadad's children: May, Raja, and Samia. Play at the Games or Avocations of the Previous Personality M.M.T. was a male child of Burma who remembered the life of the abbot of Play and Previous Lives 567 a Buddhist monastery, Ven. U Warthawa (Stevenson, 1997) (S). The abbot was fond of theatricals, and he both wrote and produced plays. He organized a dancing troupe, and he taught his students to sing, dance, and play musical instruments. He himself was an accomplished player of the flute and xylophone. As a young child, M.M.T. showed much interest in music and was fond of singing and dancing. He often played with dolls and set up a miniature toy stage. He presented plays with the dolls and other toys. G.P., a female child of England, remembered the previous life of her older sister, Joanna, who had been killed (at the age of 11) in a vehicular accident. Joanna's younger sister, Jacqueline, had been killed at the same time. G.P.'s monozygotic twin sister, J.P., remembered the life of Jacqueline (Stevenson, 1997) (S). Joanna had liked to wear costumes, and she had acted in little plays that she wrote. As a young child G.P. showed an interest in playing at acting with costumes. J.P. took no initiative in this play, but entered into it with her sister. Play Reenacting the Mode of Death in the Previous Life M.S. was a young boy of Burma who remembered the life of a man who had drowned when a ferry on which he was a passenger capsized and sank (Stevenson, 1997) (S). When M.S. was between 2 and 3 years old, he would sometimes in his play enact a scene of a man trying to escape from a sinking ship. He would shout: "The boat is sinking. Help! Help!" He enacted this scene with playmates, but I did not learn what role he assigned to them. His mother tried to suppress this play, because she was afraid that M.S. might cause a panic, or even an accident, if he behaved in this way when they were actually on a boat. R.S. was a young boy of Lebanon who remembered the life of a man, Sami Abutin, who had killed himself by holding the muzzle of a gun under his chin and somehow activating the trigger (unpubl. data) (S). He was alone when he killed himself and left no note behind. He had quarreled with his brother and was also frustrated because his wish to marry a woman he loved had been thwarted when she became interested in another man. When R.S. was about 3, he put a stick under his chin as if it were a gun and told his brothers: "Do not do this." He continued with this behavior for more than a year. When he was about 5 and playing at pointing a stick to his chin, his father asked him what he was doing. He said that he had done this to himself. He explained: "I did this for my cousin. They promised to give her to me, but they didn't." 1 Discussion Much of the play of the children who remember previous lives appears to repeat, automatically, what was done many times before. It suggests the unconscious expression of a habit. As I was writing a draft of this paper, I had a minor operation on my left hand, and for several weeks I had to wear my wristwatch on my right wrist instead of the left one, which was in a splint; I noticed 568 Stevenson i that when I wished to know the correct time I would, from habit, hold up my left arm, as if the watch were still on that arm. The interpretation that the play of children who remember previous lives is the expression of a habit seems applicable to the types of play that I have listed under vocations, under avocations, and under play appropriate for the sex of the previous life. We need a different explanation for the cases of the children who remembered the lives of parents who died leaving behind an infant or other young children. Their play attempts to recreate and continue the uncompleted role of parenting, as if no death had intervened. The reenactment by a child of the death in a previous life may express memories of a traumatic event that are forceful enough to manifest, not only in inwardly experienced images, but in the physical activity that we call play. Such children seem to have involuntary memories similar to those experienced by persons traumatized in this life, such as the victims of the Holocaust (Kuch and Cox, 1992). Children may express in play memories of a traumatic event in this life (Saylor et al., 1992; Terr, 1981, 1988, 1991). The cases I have described differ only in that they seem to derive not from a trauma in this life but from one in a previous life. Although all but two of the examples of unusual play that I have described occurred in Asia, I believe more examples will be found in Europe and North America when they are looked for more systematically than has hitherto been done. An important motive for presenting these cases is the hope of stimulating persons concerned with child development to observe and report unusual play in children. Some of the children who engage in unusual play may conceivably talk spontaneously about previous lives. If they do, parents should listen to them carefully. If they do not, parents may legitimately ask them why they engage in such unusual play. Acknowledgments Thanks are gratefully given to Emily Williams Kelly, Patricia Estes, Dawn Hunt, and Arthur Hastings for comments on drafts of this paper that improved it. I thank Jiirgen Keil, Antonia Mills, and Satwant Pasricha for information about cases that they have investigated. The research of the Division of Personality Studies is supported by the Bernstein Brothers Foundation, the Nagamasa Azuma Fund of the University of Virginia, the Japan-U.S. Fund for Health Sciences, the Lifebridge Foundation, Richard Adams, and several anonymous donors. References Bornstein, M. H., Haynes, M. O., O'Reilly, A. W., & Painter, K. M. (1996). Solitary and collaborative pretense play in early childhood: Sources of individual variation in the development of representational competence. Child Development, 67, 2910-2929. Play and Previous Lives 569 Cook, E. W., Pasricha, S., Samararatne, G., Win Maung, & Stevenson, I. (1983). A review and analysis of "unsolved" cases of the reincarnation type: 11. Comparison of features of solved and unsolved cases. Journal of the American Society for Psychical Research, 77, 115-135. Cotton, N. S. (1984). Childhood play as an analog to adult capacity to work. Child Psychiatry and Human Development, 14, 135-144. Doering, R. W., Zucker, K. J., Bradley, S. J., & MacIntyre, R. B. (1989). Effects of neutral toys on sex-typed play in children with gender identity disorder. Journal of Abnormal Child Psychology, 17, 563-574. Fagot, B. I. (1974). Sex differences in toddlers' behavior and parental reaction. Developmental Psychology, 10, 554-558. Freud, S. (1961). Beyond the pleasure principle. New York: W. W. Norton. (First published in 1920.) Haraldsson, E. (1991). Children claiming past-life memories: Four cases in Sri Lanka. Journal of Scientific Exploration, 5, 233-261. Haraldsson, E. (1995). Personality and abilities of children claiming previous-life memories. Journal of Nervous and Mental Disease, 183, 445-45 1. Haraldsson, E. (1997). A psychological comparison between ordinary children and those who claim previous-life memories. Journal of Scientific Exploration, 11, 323-335. Jacklin, C. N., Maccoby, E. E., & Dick, A. E. (1973). Barrier behavior and toy preference: Sex differences (and their absence) in the year-old child. Child Development, 44, 196-200. James, W. (1 890). The principles ofpsychology (2 vols.). New York: Henry Holt. Kuch, K., & Cox, B. J. (1992). Symptoms of PTSD in 124 survivors of the Holocaust. American Journal of Psychiatry, 149, 337-340. Lazarus, M. (1883). Die Reize des Spiels (2nd ed.) [The stimuli of play]. Berlin: Diimmlers Verlagsbuchhandlung. Lillard, A. S. (1993). Pretend play skills and the child's theory of mind. Child Development, 64, 348-371. Millar, S. (1968). The psychology ofplay. Harmondsworth, Middlesex, U K : Penguin Books. Mills, A., Haraldsson, E., & Keil, H. H. J. (1994). Replication studies of cases suggestive of reincarnation by three independent investigators. Journal of the American Society for Psychical Research, 88, 207-2 19. Rekers, G. A., & Morey, S. M. (1990). The relationship of measures of sex-typed play with clinician ratings on degree of gender disturbance. Journal of Clinical Psychology, 46, 28-34. Saylor, C. F., Swenson, C. C., & Powell, P. (1992). Hurricane Hugo blows down the broccoli: Preschoolers' post-disaster play and adjustment. Child Psychiatry and Human Development, 22, 139-149. Stevenson, I. (1974). Twenty cases suggestive of reincarnation (2d rev. ed.). Charlottesville, VA: University Press of Virginia. (First published in 1966 in Proceedings of the American Society for Psychical Research, vol. 26.) Stevenson, I. (1 975). Cases of the reincarnation type. Ten cases in India (vol. 1). Charlottesville, VA: University Press of Virginia. Stevenson, I. (1977). Cases of the reincarnation type. Ten cases in Sri Lanka (vol. 2). Charlottesville, VA: University Press of Virginia. Stevenson, I. (1980). Cases of the reincarnation type. Twelve cases in Lebanon and Turkey (vol. 3). Charlottesville, VA: University Press of Virginia. Stevenson, I. (1983a). American children who claim to remember previous lives. Journal of Nervous and Mental Disease, 171,742-748. Stevenson, I. (1983b). Cases of the reincarnation type. Twelve Cases in Thailand and Burma (vol. 4). Charlottesville, VA: University Press of Virginia. Stevenson, 1. (1986). Characteristics of cases of the reincarnation type among the Ibgo of Nigeria. Journal ofAsian and African Studies, 21, 204-216. Stevenson, I . (1987). Children who remember previous lives: A question of reincarnation. Charlottesville, VA: University Press of Virginia. Stevenson, I. (1990). Phobias in children who claim to remember previous lives. Journal of Scientific Exploration, 4, 243-254. Stevenson, I. (1997). Reincarnation and biology: A contribution to the etiology of birthmarks and birth deficts (2 vols.). Westport, CT: Praeger Publishers. Stevenson, I. (2000). The phenomenon of claimed memories of previous lives: Possible interpretations and importance. Medical Hypotheses, 54, 652-659. 570 Stevenson Stevenson, I., & Samararatne, G. (1988a). Three new cases of the reincarnation type in Sri Lanka with written records made before verification. Journal of Nervous and Mental Disease, 176, 741. Stevenson, I., & Samararatne, G . (1988b). Three new cases of the reincarnation type in Sri Lanka with written records made before verifications. Journal of Scientific Exploration, 2, 217-238. Terr, L. (1981). Forbidden games: Post-traumatic child's play. Journal of the American Academy of Child Psychiatry, 20, 741-760. Ten; L. (1988). What happens to early memories of trauma? A study of twenty children under age five at the time of documented traumatic events. Journal of the American Academy of Child and Adolescent Psychiatry, 27, 96-104. Terr, L. (1991). Childhood traumas: An outline and overview. American Journal of Psychiatry, 148, 10-20. Vygotsky, L. S. (1978). Mind in society: The development of higherpsychological processes. Cambridge, MA: Harvard University Press. Journal of Scientific Exploration, Vol. 14, No. 4, pp. 571-581,2000 0892-3310100 0 2000 Society for Scientific Exploration A Scale to Measure the Strength of Children's Claims of Previous Lives: Methodology and Initial Findings Division of Personality Studies, Department of Psychiatric Medicine University of Virginia Health System l?0. 800152, Charlottesville, VA 22908-0152 Box e-mail: jbt8n @ virginia.edu Abstract-To assess the relative strength of children's claims to remember previous lives, a strength-of-case scale has been developed. It assigns weights to features of each case that are more suggestive of a paranormal explanation. The scale was found to have a high degree of internal consistency. Analysis of 799 cases with the scale indicated that the strength of a case correlated with the economic status of the child but not his or her social status or caste. It did not correlate with the initial attitudes of the child's parents toward the case but did correlate with an early onset of statements about the previous life, the amount of emotion shown by the child when recalling the past life, and the amount of facial resemblance between the child and the deceased individual. These results are more consistent with a paranormal explanation for the cases than with a normal one. Keywords: reincarnation Introduction In various parts of the world, children have been found who claim to remember previous lives. Cases have been studied for over 30 years by Stevenson (1974, 1987) and more recently by other investigators as well (Mills et al., 1994; Pasricha, 1990). Over 2,500 cases have now been collected, and they have a number of typical features. The children begin talking about the life of a deceased individual, or "previous personality," at a very young age, usually between the ages two and five and occasionally earlier, and they generally stop between the ages of five and eight. At times, they speak of the life of a deceased family member or family friend, but other children speak of total strangers in other locations. In such situations, the children can be quite insistent that their families take them to the previous personality's home, and when this is done, the child is often said to recognize friends or family from the previous life. In addition to the statements, many children bear birthmarks that correspond to wounds suffered by the previous person, and the correspondences are confirmed by postmortem reports whenever possible (Stevenson, 1997). The children also demonstrate behaviors that appear to be associated with the previous life, such as phobias related to the mode of death of the previous individual or play that mimics the previous person's occupation. 571 572 Tucker When cases become known to an investigator, he or she then interviews the child along with as many firsthand witnesses as possible. Members of the child's family as well as members of the previous personality's family are interviewed. The two families have usually met before the investigator has arrived on the scene, but there are cases in which a written record of the child's statements has been made before the concerned families have met. Also in a very few cases, the investigator has participated in controlled tests of recognition by the child of family and friends of the previous individual. These cases generate two general types of hypotheses regarding their etiology, the "normal hypotheses" and the "paranormal hypotheses." In the sociopsychological hypothesis (Brody, 1979), a normal hypothesis, it is pointed out that most of the children live in cultures with a belief in reincarnation. They are said to hear of deaths, and they then say that they were the deceased individuals out of either fantasy or deception. After the two families have met, the family members remember the child's statements and recognitions as being more impressive than they actually were since they have the wish that the deceased individual has reincarnated as the child. The birthmarks are said to be mere coincidences, and the behaviors indicative of the previous life are thought to be part of the self-delusion. A previous study comparing cases having written records made before the two families met with ones without records did not find support for such a theory (Stevenson and Schouten, 1998). The paranormal hypothesis most often considered is that of reincarnation (there are others, such as possession and the "super-psi" hypothesis). Investigators have reported cases in which the children appeared to have knowledge about the previous personality that they could not have obtained through normal means, and in some cases (as mentioned), written records were made before the previous individual's family was located (Haraldsson, 199 1; Stevenson, 1974, 1975, 1977; Stevenson and Samararatne, 1988). This knowledge, considered along with the children's identifications as the deceased individuals, the matching birthmarks, and the unexpected behaviors, appears to support the reincarnation explanation for the cases. Some cases produce stronger evidence for a paranormal explanationthrough more verified statements about the deceased individual, through birthmarks that correspond to lesions on the deceased, and so forth-than others do. From the two types of hypotheses, predictions can be generated about what factors might be associated with the stronger cases. For instance, the normal hypotheses would predict that in cases with more apparent evidence for reincarnation, the children have had more opportunity to obtain information about the deceased individual through normal means than in the weaker cases. The present paper describes a scale to assess the strength of the evidence for a paranormal explanation for a case. It then demonstrates the use of this scale to determine what factors are associated with the stronger cases and to test predictions generated by the two types of hypotheses. Strength-of-Case Scale for Claims of Past Lives 573 Methods Cases have been investigated by several researchers who have been associated over the past 30 years with the University of Virginia Division of Personality Studies. A case will typically include a registration form and the researcher's interview notes, and this information is transferred to a coding form to be entered into a computerized database. The coding form, completed as much as the information for each case allows, includes demographic information about the subject and the previous personality along with other features of the case. 799 cases have now been entered into the database, and a strength-ofcase scale (SOCS) has been developed that assigns relative weights to those features of a case that may bear on the strength of the evidence for a paranormal explanation (Table 1). As Table 1 shows, there are four broad categories of factors that provide evidence for the paranormal hypothesis. The first is birthmarks and birth defects that are said to correspond to lesions on the deceased individual. The second category involves statements that the subject has made about the previous life, with points given for statements that are verified to be accurate and points subtracted for statements that turn out to be incorrect. The third category involves behaviors that the subject exhibits that seem to relate to the previous life. Finally, the fourth involves the connection between the subject and the previous personality, since a closer association would increase the chances that a child could have learned about the previous personality through normal means. Thus, a case in which the two families are more than 25 km apart scores more points than ones in which they are closer, and ones with no contact between families score more than ones in which the families had a close association before the case developed. In addition to spatial proximity, consideration is given to closeness in socioeconomic status and, for Indian cases, in caste placement, since those at different ends of the spectrum might have just as much functional distance between them as cases that are many kilometers apart. The points that a case gets for each item are then summed to give a total score on the SOCS. The points that each item scores were determined subjectively, with an attempt made to weigh items in a balanced way. Thus, for example, a birthmark verified by medical records to correspond with a fatal wound on the previous personality receives eight points just as the presence of more than 20 verified statements does. With each case receiving a score on the scale, other features in the cases were then examined for correlations with the strength of a case. For this initial use of the scale, 799 cases were available in the database. The sex ratio and country of origin of the cases are outlined in Table 2. Of note is that there are some countries, such as Thailand and Myanmar (Burma), from which we have numerous cases, but these cases have not yet been entered into the database. The SPSS 9.0 for Windows statistical package was used to obtain the statis- Tucker TABLE 1 The Strength-of-Case Scale (SOCS) Item Birthmarkslbirth defects 1. Corresponding to fatal wound on deceased individual Verified by medical records Verified by friends or relatives of deceased Claimed by subject but not verified 2. Corresponding to nonfatal wound on deceased individual Verified by medical records Verified by friends or relatives of deceased Claimed by subject but not verified 3. Disease or infirmity related to deceased individual Verified by medical records Verified by friends or relatives of deceased Claimed by subject but not verified Statements about the previous life 4. Subject claims to remember previous life Yes No, claim made only on basis of other evidence 5. Statements about places or people as they appeared during previous life, these appearances having since changed 6. Statements about the previous life verified as correct minus incorrect statements made >20 11-20 6-10 1-5 0 -1-(-5) -6-(- 10) <-lo Behavior 7. Unusual dietary cravings or avoidances related to previous life 8. Unusual methods of eating or table manners related to previous life 9. Unusual use of intoxicants related to previous life 10. Unusual philias related to previous life 1 1. Unusual skills or aptitudes related to previous life 12. Unusual animosities related to previous life 13. Unusual phobias related to previous life 14. Behavior related to that of the opposite sex According to friends or relatives According to subject or investigator only 15. Desire or reluctance to return to previous family or place Strong desire Moderate desire Neutral Moderate reluctance Strong reluctance 16. Memories of previous life expressed in play Connection to deceased individual 17. Identification of deceased individual By academic investigators By other investigators By family or friends Points Strength-of-Case Scale for Claims of Past Lives TABLE 1 (continued) The Strength-of-Case Scale (SOCS) Item 575 Points ! I 18. Association between the two families before case developed Close association or same family Slight association Knew about each other but no association Total strangers unknown to each other 19. Distance (in km.) between child's birthplace and deceased individual's main residence 0.0-1.5 1.6-24.9 25.0 or more 20. Difference in social status between child and deceased individual Slight Moderate Great 21. Difference in economic status between child and deceased individual Slight Moderate Great 22. Difference in caste of subject and deceased individual with score equal to difference in ranking by Brahmin-Kshtriya-Kayastha-Vaishya-SudraUntouchable tics involved in this report. Since the SOCS produced results in which the cases do not have a normal distribution, a nonparametric correlation test, the Spearman correlation coefficient, was used. Results A SOCS score was obtained for each of the 799 cases in the database. The scores ranged from a low of -3 to a high of 49, with a mean of 10.40, a median TABLE 2 Sex Ratio and Country of Origin of Cases Analyzed Frequency Sex of subject Male Female Total Subject country of origin India Sri Lanka Turkey Lebanon United States Canada Total Percent 576 Tucker strength-of-case scale Fig. 1Distribution of strength-of-case scores. of 8.00, and a standard deviation of 9.48. The distribution was significantly skewed (Figure 1). To assess the internal consistency of the scale, each of the individual items was compared to the rest of the scale. This produced results in which all but two of the items significantly correlated with the remainder of the scale (Table 3). The two items that did not show a significant correlation, items 4 and 14, did show trends in that direction. In addition, each of the four parts of the scale-birthmarkslbirth defects, statements about the previous life, behaviors related to the previous life, and connection between the subject and the deceased individual-showed a significant correlation to the remainder of the scale (Table 4). Thus, the scale showed considerable internal consistency. The scale was then used to assess how other factors in the cases correlate with the strength of the cases. To look at the possibility that the cases originate from fantasies of poor children wishing they were in better circumstances, the social status, economic status, and caste of the subject (for Indian cases) were compared with scores on the SOCS. In this instance, subjects from the extremes of the socioeconomic range have an advantage on the scale since they are more likely, by chance, to have a greater distance from the circumstances of the previous personality. The economic status of the subject was found to have a marginally significant correlation with the SOCS score ( p = .045), but when the item for the difference in economic status between the child and the Strength-of-Case Scale for Claims of Past Lives TABLE 3 Correlations of Each Item With Remainder of SOCS Item Birthmarks matching fatal wounds Birthmarks matching nonfatal wounds Disease related to deceased individual Subject claims to remember previous life Statements about changed appearances Verified statements about previous life Unusual dietary cravings related to previous life Unusual methods of eating related to previous life Unusual use of intoxicants related to previous life Unusual philias related to previous life Unusual skills or aptitudes related to previous life Unusual animosities related to previous life Unusual phobias related to previous life Behavior related to that of the opposite sex Desire or reluctance to return to previous place Memories of previous life expressed in play Identification of deceased individual Association between the two families before case Distance between child and deceased individual Social status difference between the two Economic status difference between the two Caste status difference between the two Spearman coefficient 577 Significance previous personality was removed from the scale, then that correlation was no longer significant (Spearman correlation coefficient = 0.082, p = .075). The social status and caste of the subject showed no correlation at all (Table 5). To assess the possibility that enthusiasm of parents is creating these cases, the initial attitude of each of the subject's parents was compared to the strength of the case, and there was not a significant correlation (Table 6). (Even in countries with a belief in reincarnation, many parents try to suppress the child's statements about a previous life, believing, for example, that such statements will lead to a shorter life.) This is consistent with a previous study that found that in 41 % of cases in India, parents tried to suppress the children from talking about previous lives, but this had no effect on the duration of the child's speaking of them (Stevenson & Chadha, 1990). The amount of acceptance of the case that the family of the previous personality shows does correTABLE 4 Correlations of Sections With the Remainder of SOCS Section Birthmarks and birth defects Statements about the previous life Behaviors related to the previous life Connection to the deceased individual Spearman coefficient 0.178 0.360 0.429 0.360 Significance <0.001 <0.001 <0.001 <0.001 578 Tucker TABLE 5 Correlations Between the SOCS Score and the Social Status, Economic Status, and Caste of the Subject Number of cases coded Spearman coefficient -0.008 0.092 0.009 Significance 0.863 0.045 0.892 Social status of subject Economic status of subject Caste of subject 493 471 247 TABLE 6 Correlations Between the SOCS Score and the Attitude of Families Toward the Case Number of cases coded Initial positive attitude of subject's mother Initial positive attitude of subject's father Amount of acceptance from deceased's family 497 457 275 Spearman coefficient -0.061 0.032 0.234 Significance - - 0.172 0.493 <0.001 late significantly with the SOCS score, implying that those families use criteria similar to the ones in the scale in assessing the legitimacy of the case. Several other characteristics of the cases were studied (Table 7). The age at which the child began to talk about a past life shows a significant negative correlation to the SOCS score, so that in the stronger cases, the child is more likely to start talking earlier. In addition, there is a strong correlation between the SOCS score and the amount of emotion that the subject shows when recalling the past life. Since item 15 of the scale, desire or reluctance by the subject to return to the previous family or place, might be related to the amount of emotion the subject shows, this correlation was tested with item 15 removed from the scale, and it remained significant (Spearman correlation coefficient = 0.205, p < .001). Finally, the amount of facial resemblance between the subject and the deceased individual, as judged by those who knew the deceased, proved to be significantly correlated with the strength of the case. TABLE 7 Correlations Between the SOCS Score and Other Characteristics of the Cases Number of cases coded Age when subject first talked about previous life Amount of emotion subject shows during recall Facial resemblance between subject and deceased 636 484 300 Spearman coefficient -0.106 0.235 0.386 Significance 0.007 <0.001 <0.001 Strength-of-Case Scale for Claims of Past Lives 579 Discussion The above results can be considered in relation to the normal and paranormal hypotheses. In looking at the internal consistency of the scale, one area that produced surprising results w?s how the connection between the subject and the previous personality correlated with the remainder of the scale. A case is given points for distance between the two, be it lack of association between the families, geographical distance, or socioeconomic or caste difference. All of these correlated with the remainder of the scale, both as individual items and as a group. Thus, greater distance between the child and the previous personality is associated with more evidence for a paranormal explanation. This is the exact opposite of what the socio-psychological hypothesis would predict, since it supposes that the children have acquired their knowledge about the previous personality through normal means. These results suggest that this is likely not the case. As for the reincarnation hypothesis, it would not necessarily predict a positive or negative correlation between distance and the strength of a case. What the positive correlation may suggest is that, while the long distance cases may well be reincarnation cases, some of the close proximity reincarnation cases may be diluted by others best explained through normal means. In looking at how other factors may correlate with the SOCS, the correlation between the scale score and the economic status of the subject provides some support for the idea that the cases are the result of poor children fantasizing about a better life, but as noted above, this correlation is not significant if the difference in economic status between subject and previous personality is removed from the scale. In addition, there is no correlation between the strength of the case and the social status and caste of the subject. This lack of correlation would be perfectly consistent with the reincarnation hypothesis. The lack of correlation between the SOCS score and the reaction of the subject's parents to the child's claims is also inconsistent with the prediction of the normal hypotheses that the parents help create these cases, either accidentally or intentionally for an unknown motivation. The negative correlation between the SOCS score and the age when the child begins talking about the previous life is consistent with the reincarnation hypothesis, in which children with more residue from the previous life would be more likely to talk about that life as soon as the child is able. As for the normal hypotheses, one might argue that younger children with less developed language skills would make statements that would be more likely to be misinterpreted as possessing paranormal knowledge, but it could also be argued that older children could be more reliably made part of a fraud. It seems most reasonable that the normal hypotheses would not make a prediction regarding a correlation between the strength of the case and when the child starts talking about the past life. Likewise, the reincarnation hypothesis would predict that children with strong impressions from the previous life would be more likely to express their 580 Tucker memories with strong emotion. The normal hypotheses would not predict a correlation between the strength of a case and the amount of emotion the child demonstrates about the memories, but in fact, a correlation was found. Lastly, the correlation between the SOCS score and the facial resemblance between the subject and the previous personality is consistent with the reincarnation hypothesis, in which cases with more past life residue would carry with them some facial resemblance just as they do birthmarks and birth defects. The socio-psychological hypothesis might predict that people who believe a child is the reincarnation of a deceased person would be more likely to see a facial resemblance than they would otherwise. It would not, however, predict this finding that facial resemblances are more commonly reported in the stronger cases than in the weaker ones. I Conclusions A scale has been developed that measures the evidence that each case of past life claims provides for a paranormal explanation, and it shows good internal consistency. Several factors have been found to correlate with the scale score, including distance between the subject and the deceased individual, earlier onset of statements about the previous life, the amount of emotion the subject shows when discussing the previous life, and the amount of facial resemblance between the subject and the deceased individual. These results are more consistent with the reincarnation hypothesis than with one of a normal origin. While they are not definitive, they do indicate that more research is justified in the cases of children who appear to remember previous lives. I Acknowledgements Sources of support include the Nagamasa Azuma Fund, the Bernstein Brothers Foundation, the Perrott-Warrick Fund, the Japan-U.S. Fund for Health Sciences, Richard Adams, and the Lifebridge Foundation. In addition, the author wishes to thank Ian Stevenson, MD and Bruce Greyson, MD for their suggestions about earlier drafts of this paper and Ms. Dawn Hunt, RN for her assistance with the SPSS statistical package. References Brody, E. B. (1979). Review of Cases of the reincarnation type.Volume 11. Ten cases in Sri Lanka, by I. Stevenson. Journal of Nervous and Mental Disease, 167, 769-774. Haraldsson, E. (1991). Children claiming past-life memories: Four cases in Sri Lanka. Journal of Scientific Exploration, 5, 233-261. Mills, A., Haraldsson, E., & Keil, J. (1994). Replication studies of cases suggestive of reincarnation by three different investigators. Journal of the American Society of Psychical Research, 88,207-219. Pasricha, S. (1990). Claims of reincarnation. An empirical study of cases in India. New Delhi: Harman Publishing House. Stevenson, I. (1974). Twenty cases suggestive of reincarnation (2d rev. ed.). Charlottesville, VA: University Press of Virginia. (First published in 1966.) Strength-of-Case Scale for Claims of Past Lives 581 Stevenson, I. (1975). Cases of the reincarnation type. I. Ten cases in India. Charlottesville, VA: University Press of Virginia. Stevenson, I. (1977). Cases of the reincarnation type. 11. Ten cases in Sri Lanka. Charlottesville, VA: University Press of Virginia. Stevenson, I. (1987). Children who remember previous lives. A question of reincarnation. Charlottesville, VA: University Press of Virginia. Stevenson, I. (1997). Reincarnation and biology. A contribution to the etiology of birthmarks and birth defects. Westport, CT: Praeger. Stevenson, I., & Chadha, N. K. (1990). Can children be stopped from speaking about previous lives? Some further analyses of features in cases of the reincarnation type. Journal of the Society for Psychical Research, 56, 82-90. Stevenson, I., & Samararatne, G. (1988). Three new cases of the reincarnation type in Sri Lanka with written records made before verifications. Journal of Scientific Exploration, 2, 217-238. Stevenson, I., & Schouten, S. A. (1998). Does the socio-psychological hypothesis explain cases of the reincarnation type? Journal of Nervous and Mental Disease, 186, 504-506. Journal of Scientific Exploration, Vol. 14, No. 4, pp. 583-622,2000 0892-33 10/00 0 2000 Society for Scientific Exploration 1 Reanalysis of the 1965 Heflin UFO Photos ANN DRUFFEL 257 Sycamore Glen, Pasadena, CA 91105 ROBERT M. WOOD 1727 Candlestick Lane, Newport Beach, CA 92660 ERIC KELSON 10201 Lindley Avenue, Apt. G-99, Northridge, CA 91325 Abstract- Photography is one of the few sources available that can demonstrate the scientific validity of UFO phenomena. The Heflin photos, taken in Santa Ana, California, in 1965 were regarded as most probably genuine until 1968. Then questions arose from scientists and other UFO researchers that remained unanswered until the 1990s' mainly because the four original photos had been taken by unidentified persons posing as government personnel. In 1993, Heflin's Polaroid originals surfaced unexpectedly under mysterious circumstances, and reanalysis was resumed by a three-member team, including two scientists and a veteran UFO investigator. State-of-the-art computer enhancement has revealed new data that answer all prior doubts and questions and discloses additional information that could not have been available at the time they were taken. In this article, we address the validity of the objections as originally put forth and the results of the computer-enhancement analysis. A second article, to follow, details more technical aspects of the computer-enhancement analysis performed. Keywords: UFO photographs-vortex rings-Rex E. Heflin-photo computer enhancement-Fort Belvoir, VA-James E. McDonald, Ph.D. Introduction Historical Elements of UFO Photo Analysis In the earliest era of UFO research, investigators realized that only those reports in which documentedproof could be obtained would convince the scientific community that UFOs were actual objects. The main problem has been, from the beginning, to convince the scientific community at large that the UFO phenomenon constitutes a question worthy of serious study. Only by doing this can adequate attention and funding be brought to the problem. One class of sightings that offered hope of proof that a scientific mystery existed was photo cases that held up under careful scrutiny. Very few earlier UFO photos survived scientific analysis. Some had been presented as genuine UFOs that proved Frisbees, hubcaps, disc-shaped lamps 584 A. Druffel, R. M. Wood, & E. Kelson reflected in windows, lens flares, balloons, lighted air vehicles, and more recently, computer-generated hoaxes. The four Heflin photos taken in 1965 allegedly depicted a close passage of a UFO, and an all-out effort was launched to study them. The major UFO research organization of the time, the National Investigations Committee on Aerial Phenomena (NICAP), spared no effort to have them examined by the best photogrammetrists available, as did other smaller research groups. By the time the case was considered authenticated by the Los Angeles NICAP Subcommittee (LANS), the pictures had been studied by six photographic experts or teams and the integrity of the photographer, Rex E. Heflin, had been established beyond doubt. The event received media coverage all over the world, and the photos were reprinted in UFO journals in many countries. One of the photos even appeared in the 1968 Encyclopedia Britannica under a section titled "Spacecraft" (Encyclopedia Britannica, 1968). Many aspects of the investigation, including an intense reanalysis that has been on-going since 1993, are revealed here for the first time. Essential Facts of the Heflin Sighting On August 3,1965, at about 12:30 p.m. PDT, Rex E. Heflin, age 38, stopped his work van near the intersection of Myford Road and Walnut Avenue, on the boundary of Orange County, near the city of Santa Ana in southern California. Heflin wore no watch and had to estimate the time. He was about one half a mile outside the perimeter of El Toro Marine Base, and his van was facing in a northerly direction. Some branches of a tree were obscuring a railroad-crossing sign, and Heflin had stopped to photograph it, for he was a highway maintenance engineer for the Orange County road department; part of his job involved keeping all traffic signs clearly visible. He attempted to report the obscured sign to his supervisor on his van radio but was unable to get through due to an unexplained failure of the radio. Still inside his van, he was about to photograph the sign with his Model 10 1 Polaroid work camera, which was loaded with ASA 3000 film. He suddenly saw a flash of motion on the periphery of his left eye and then viewed a silvery craft flying slowly from left to right across Myford Road. The object's general appearance caused him to assume that it was an experimental aircraft from El Toro. It appeared to be about one eighth of a mile away and at about 150 feet altitude. Heflin judged its approximate size by comparing it to the traffic lanes over which it flew. These lanes, he knew, were 20 feet wide, and the craft was about the same apparent size (Heflin, 1965). As the flying object crossed Myford Road, it seemed to hover and Heflin photographed it through his windshield as it crossed his line of vision (Figure 1). As the craft moved toward the east, it tipped and revealed a dark underside. "This is when I saw a rotating beam of light emitting from the center of the UFO on the bottom side," Heflin stated in his September 18, 1965, signed statement to the first UFO investigator, engineer Edward Evers, who inter- ~ Reanalysis of the 1965 Heflin UFO Photos 585 i ~g I t ~ r \ ~,hotographof'Iici11n \ ~ g l ~ t (Heflin Photo 1 ) t ~ng viewed him on behalf of NICAP and LANS (Heflin, 1965). Heflin described the beam as greenish white and rotating clockwise around the bottom from the center out to the ri111; each rotation took about 2 seconds. He snapped a second photo through his passenger window (Figure 2 ) . Throughout its passage, the craft seerned to "wobble" slightly. In Hcflin's words, its n~otion was ""sn~ilar a gyroscope when losing its stability." As to the object moved f~irther east, i t began a turning motion, and Heflin snapped a third photo (Figure 3). "The LIFO continued moving away, slowly gaining altlt~iiiz bee~mcdto gain stabilily, then it increilbcd I L v e l o ~ r i yand altliude ... ~ more rapidly leaving a dcposit o f sinokclike vapor" (Heflin, 1965). It then disappeared into the distance traveling north northeast over the Santa Ana Freeway that cut across the landscape about seven eighths of a mile away. A ring of bluish-black smoke hung in the sky at the same altitude and azimuth as the craft's position before it began its rapid departure. Hef'lin drove north on Myford Road 100 or 200 yards toward the smoke ring, which was slowly rising in altitude and traveling northeast with the wind at an angle of about 50-degrees elevation. When he got close to the smoke ring, he got out of his van and photographed it (Figure 4). The only reference points in this fourth photo, besides clouds from an overcast sky, were a tele- 586 A. Dr~iffel, M. Wood, & E. Kelson R. Fig. 2 Second photograph of I l e f l ~ n sightrllg ( I lctlin I'hoto 2 ) phone wire and a small portioi~of an orange tree in a nearby commercial grove. Heflin j~tdged ring to be three to four times larger than the diameter the of the craft with which he assunled it was associated. The smoke ring was being bent out of shape by light winds. He was nevertheless surprised at the ring's "solid" appearance; it did not dissipate like ordinary smoke. Still thinking he had photographed an experi~nental plane, he turned his attention back to his work and found that his radio now worked norlnally (Heflin, 1965). He was not particularly excited about the pictures he had taken and didn't mention the incident to anyone until he showed the photos that afternoon to a few coworkers in his Santa Ana office. They caused considerable interest, but some of his workmates seetncd agitated by the fourth picture of the smoke ring; it wasn't anything one would expect to see associated with experimental aircraft. Ijeflin decided that "three pictures were enough for one day" and stopped showing the fourth (McDonald, 1968). In fact, he tho~ight little more about any of them until some of his relatives and friends became intrigued with them. Thinking that Heflin might have photographed a UFO, some of his coworkers asked to borrow the pictures, and he loaned them the originals to be copied. T e also loaned the originals to his sister, Mrs. Williani Fincher. FinalT Reanalysis of the 1965 Heflin UFO Photos 587 Fig 3 1 htrd photograph ofllellin \ightlng ( H c l l ~ n Photo 3) ly, Mr. and Mrs. Pat H~irIey, friends of thc Finchers, borrowed them and took them to a drugstore in Santa Ana, of which Mr. Hurly was part owner. Mrs. R. C. Hoiles, publisher and co-owner of the Suntln Ancr Register newspaper saw copies of thc photos in I lurley's drugstorc a n d asked to borrow them for possible publication. Mr. Clay T. Miller, chief photograpl~el- the Rcgi,ste~; about tracking of set Heflin down at his work. He persuaded him to bring the originals down to the newspaper office to be copied, but only after some hesitation on Heflin's part because he sought no publicity or remuneration. He still regarded them, at the time, as pictures of an experin~ental craft f r o ~ n Toro. Six sets of negatives of El the first three photos were made by Clay 'I-. Miller, but they were croppcd for easier publication, cutting off the foreground, which showed the shadow of a nearby telephone pole, in Photo 1 . However, Mrs. Fincher had also loaned the originals to a ~ ~ o t h li-iend, who took thciu to Wayne Thornhill, an amateur er photographer. Thornhill also made copies, which were of lesser quality than the Register's but were not cropped. The foreground features latcr proved invaluable in ascertaining the exact t i n ~ c sighting (later determined through of Griffith Observatory to be about 12:37 PD'T). 1 588 A. Druffel, R. M. Wood, & E. Kelson Pig. 4. Fourth photograp11 of Hcflin qtght~ng (tfcllin Photo 4). Aftermath of the Photos' Publicity The Register checked at El Toro Marine Rase, inquiring if any personnel there had seen the craft. El Toro officials stated that no other reports had been received and denied that the object was an experirncntal aircraft from their base. Heflin's first three photos of the unidentified craft were published for the first time in the Register on September 20, 1965, 6 weeks after the event had occurred. Heflin was never asked for permission to print the photos, and even though they were subsequently published widely in journals and magazines throughout the world, he never copyrighted them or asked for any remuneration for their use for 30 years. As a result of the Register h story, years of investigation by UFO researchers began, as well as continual harassment of Heflin by curiosity seekers. Through it all, Heflin never sought publicity, but when he was sought out by objective researchers, he was always forthright and cooperative. Primary among these was the investigative team from LANS, which was headed at the time by Dr. Leslie K. Kaeburn, a noted biophysicist. Several qualified NICAP investigators participated in the LANS study, including John Gray and Edward Evers, two aviation engineers employcd at North American Aviation. Idabel Epperson, an objective researcher who was cochairman of L A N S (and Reanalysis of the 1965 Heflin UFO Photos 589 chairman after Dr. Kaeburn's death in 1968), was also an active investigator on the case. Other subcommittee members also participated, and various UFO researchers around the country launched independent investigations. By the time LANS became involved in the case in September 1965, Marine Corps Intelligence officers had come by Heflin's house, interviewed him, and borrowed the first three pictures to make copies. Heflin did not ask the Marines for a receipt, yet they requested him to sign a receipt when they returned the photos unmarked. (The only marks on the four originals were the numbers 1,2,3, and 4, which Heflin had written in blue ink to demonstrate the sequence in which they had been taken.) The U.S. Air Force also contacted him and conducted an official inquiry. The investigating officer, Captain Charles F. Reichmuth, interviewed Heflin for more than 3 hours and also copied and returned the photos unmarked. He checked with Heflin's supervisors and learned that he was a valuable employee-mature, alert, and trustworthy. Reichmuth noted in his official report that he "could find no evidence to disagree with this estimate." He sent his report to Project Blue Book at Wright-Patterson AFB in Dayton, Ohio (Reichmuth, 1965; Young, 1967). In spite of Reichmuth's generally positive report, a Project Blue Book "Photo Analysis Report" describes a comparison photo, made by other Air Force officials, of a 9-inch tray tossed into the air at 15-20 feet from the camera. The results of the Air Force experiment evidently satisfied Project Blue Book personnel that the photos were hoaxed. Blue Book was, at the time, the only publicly known official Air Force group responsible for studying UFOs, and it officially listed the Heflin photos as a hoax (Project Blue Book Files, 1965). This Blue Book assessment was made despite that Captain Reichmuth's report had stated, "From all appearances, he [Heflin] is not attempting to perpetrate a hoax" (Reichmuth, 1965). A man representing himself as a NORAD colonel phoned Heflin on September 20 and arranged to meet him 2 days later, warning him not to discuss the event further with the press. On the evening of the 22nd, two men in civilian clothes, claiming to be from NORAD, came to his door. One of them flashed a salmon-and-green card, which, to Heflin, looked similar to those carried by El Toro Marines. Heflin did not remember the name on the ID but noted that it did not have a photo. The second man did not participate in the conversation. Heflin obligingly lent "the NORAD man" the three pictures of the craft, fully expecting that they would be returned, as they had been returned by the U.S. Marine Corps (USMC) and the U.S. Air Force. The self-proclaimed NORAD men never returned the photos. Heflin tried to track them down with NICAP-LANS assisting in the search. NORAD disclaimed any knowledge of the photos. Heflin even contacted his Congressman, Representative Utt, who inquired on behalf of Heflin and was assured that NORAD offices had been searched from top to bottom with no results. The identity of the "NORAD men" remains a mystery to this day. 590 A. Druffel, R. M. Wood, & E. Kelson The First Photogrammetric Analyses As a consequence, Heflin and UFO researchers were left with only copies of some of the best UFO photos that had come to light. The first groups of photogrammetrists who studied them could be assured of first-generation copies made directly from the originals; now the originals of Photos 1,2, and 3 were no longer available. Heflin was criticized by some in the UFO field for lending out his original photos in what they considered a careless fashion. Heflin, however, was used to working with military officials in the course of his work and was by nature a trusting individual. He had been basically a skeptic regarding UFOs and for weeks after the event continued to think that the object was probably an experimental aircraft. Not until scientists and engineers connected with NICAP-LANS and other organizations took interest in the photos, and particularly after they disappeared, did he begin to think he had photographed something highly unusual-i.e., a possible UFO. As the number of researchers who wanted to analyze them grew, it soon became impossible to ensure whether or not the copies they were working from were first-generation copies or not, since the originals were no longer available. The danger of "tampering" in second- and thirdgeneration copies grew. In the spring of 1966, a highly respected scientist, Dr. James E. McDonald, actively entered the field of UFO research. McDonald was senior physicist at the Institute of Atmospheric Physics at the University of Arizona at Tucson. To him, the government's apparent neglect and disinterest toward UFOs was untenable, and his own research had convinced him that UFO reports presented a serious scientific problem. He had studied the UFO phenomenon since 1958 and had worked closely with NICAP personnel and other smaller scientific UFO groups and individuals throughout the United States, Australia, New Zealand, Canada, France, and Russia (McDonald, 1 968a). Since 1966, he had been lecturing on the UFO problem to scientific organizations including the American Institute of Astronautics and Aeronautics, the American Academy for the Advancement of Science, the American Meteorological Society, and other scientific groups whose interests bordered on atmospheric physics. Besides his scientific talents, he was a prolific writer who had hundreds of highly technical and semitechnical articles published in referred journals. His interest in informing the public also prompted him to contribute to popular magazines such as Scientific American (McDonald, 1953). The numerous objective UFO researchers with whom he became personally acquainted were intrigued by his scientific excellence, flawless reputation, and persistence, as well as his cordiality and humor. By March 1967, he had taken an active interest in the Heflin case, and long letters regarding every aspect of the investigation traveled between NICAP-LANS and Tucson, Arizona. McDonald also investigated the case by telephone, holding long conversations with any source that could give him information about Heflin's character and reliability. Reanalysis of the 1965 Heflin UFO Photos 591 Eventually he concluded that Heflin's photos were among the very few UFO pictures that could be considered "probably genuine." He included them in his list of "100 best cases," which he sent to the Condon Committee's staff with the suggestion that they be investigated by that government UFO study group, which at the time had been newly funded. Extremely dubious of the Air Force "hoax explanation," McDonald wrote in his "Heflin" file, "Wonder if anyone had taken Rex's camera, set it for 15 feet, shot a 9-inch pan, then checked for blurring of the freeway power lines visible in Heflin's photos?" (McDonald, 1968). Photo analysis had already shown that the telephone lines (and the UFO) were in sharp focus, indicating that Heflin had held his camera steady as he snapped his four photos. He also investigated the interference on Heflin's van radio, because this was a reported physical effect that could be checked out. He confirmed that the radio malfunction had occurred and that other Highway Department radio systems in the area had also been affected. When Heflin had called in after the incident, he was told that others on the system had failed to get any signal at the time his own radio went dead. His superiors considered the incident very odd and encouraged Heflin to have the radio checked, even though the radio had worked perfectly after the UFO flew out of sight. The radio was checked, but the technician had found nothing wrong. The radio trouble, therefore, was not normal static or interference-the system was "just dead," as McDonald wrote in his "Heflin" file. Consulting with a University of Arizona colleague, Walt Evans, McDonald satisfied himself that it was quite conceivable that a strong [electromagnetic] field at the same frequency as Heflin was calling out on could "blank" the system because the type of amplifier usually used in first stage would simply block or saturate at very high receiver signal strength and transmit nothing at all" (the word "strong" was underlined by McDonald). McDonald was puzzled, however, by the fact that the first three pictures of the craft in flight, taken from inside Heflin's van, showed "flat," apparently unclouded skies while the fourth, which Heflin stated was taken outside his van only about a minute or two after the first photo, showed what McDonald's meteorologist's eye judged to be "substantial clouds" in the vicinity of the ring. This did not trouble other researchers who worked on the photos, for they understood that Heflin's automatic light meter on his Polaroid work camera tended to "flatten" a bright sky in photos taken from inside his work van. McDonald, however, continued to pursue the question. He checked with every available weather service within 50 miles of the Myford Road site. From their logs, and from scientific data concerning humidity and temperature in the locality for that date and time, he became convinced that the apparently thick clouds in his copy of Photo 4 could not possibly have formed in the 592 A. Druffel, R. M. Wood, & E. Kelson sky at that location on the date in question. In spite of the fact that copies of the same photo in the hands of NICAP investigators showed the sky as being overcast, he began to suspect that Heflin had taken the fourth picture at another time and place. Idabel Epperson had also checked weather data. G. W. Kalstrom of the U.S. Weather Bureau at LAX International Airport had assured her that thick clouds could, and did, form in portions of the Los Angeles Basin when the rest of the sky was merely overcast. Epperson also noted that her own copy of Photo 1 showed an overcast sky while McDonald's copy showed a flat, apparently clear sky. The differences between Epperson's and McDonald's copies lie in the fact that they had been made separately by different film developers, who had used different degrees of exposure. McDonald still was not convinced. He was confident that the apparent thick clouds in his copy of Photo 4 could not possibly have formed under the temperature-dew point difference that prevailed in the area at the time of Heflin's sighting. Pursuing the issue, Epperson learned from the National Weather Records Center and other weather stations close to El Toro that an overcast sky could look essentially "flat" in pictures taken at a low angle from inside an enclosed vehicle. LANS' photogrammetric consultants confirmed this. In addition to this possibility, there could have been fairly thick overcast in the easterly sky where the ring was photographed, according to weather conditions Epperson obtained from the U.S. Weather Bureau. At the 50-degree angle at which Photo 4 had been taken outdoors-where the camera's automatic light meter performed in normal fashion-this fairly thick overcast would have been plainly visible. McDonald remained doubtful, possibly because his expertise was cloud physics, not photography. As the investigation continued, Heflin was bothered incessantly by curiosity seekers as well as members of the media; the Road Department also was swamped with calls. By now, he realized that he had photographed an "unknown" and began to wish heartily that he'd never told anyone that he had taken the pictures. He vowed to LANS that he "wouldn't tell a living soul" if he ever had an opportunity to photograph a UFO again. McDonald's growing doubts about Photo 4 also puzzled him. By now, LANS members had become well acquainted with McDonald's fierce persistence when bird-dogging a case. They assured Heflin that it was typical of McDonald to affirm and reaffirm every possible aspect of sightings that captured his interest. Besides this, McDonald had written personally to Heflin on December 27, 1967: "As you know, your 1965 photos remain the outstanding photographic evidence yet submitted concerning UFOs" (McDonald, 1967). By the time McDonald was able to investigate the case on-site, LANS had done a 2-year check on Heflin's character and work record and had been assured by his superiors and coworkers that Rex Heflin was a straightforward individual with 15 years of responsible duty in the County Road Department and was not the type to pull a hoax that could jeopardize his job. The fact that Reanalysis of the 1965 Heflin UFO Photos 593 I Heflin had an offbeat sense of humor and joked at times in a deadpan fashion, particularly when irritated, in no way detracted from his truthful and responsible nature. Early in November 1967, McDonald arranged with Dr. William Hartmann, a University of Arizona faculty member who was in charge of photo cases for the Condon Committee, that they would investigate the Heflin case on-site together. Hartmann had copies of most of the documents concerning the Heflin sighting and had also confirmed the radio interference that had occurred at the time of the sighting, having interviewed Heflin's supervisor, Mr. Herm Kimmel, as McDonald and NICAP had done. Kimmel had told him he heard the radio cut off in the middle of the transmission and that the sudden cutoff was akin to "button release," except that there was no blip of noise, as occurred when the button was released. NICAP investigators, McDonald, and Hartmann were all satisfied that Heflin's description of the radio failure was accurate. In January 1968, McDonald and Hartmann traveled to California and interviewed officials and radar technicians at El Toro Marine Base. Reviewing the Marines' official reports, they learned that the investigating officer had checked Heflin's character, work record, and reputation and had interviewed Heflin's relatives and friends. They had also shown the photos to the Commander of the 1lth Navy District in southern California, although the Navy never expressed any open interest in the case. The most cogent question on the scientists' mind was, had the unidentified craft been viewed on radar from El Toro or adjacent military facilities? If so, the Heflin photos would constitute a unique piece of empirical evidence, a UFO caught on three observing channels at the same time-i.e., visual, film, and radar. McDonald and Hartmann, however, were told that the radar log at the Marine Corps Air Station (MCAS) had noted nothing unusual at the time of the sighting. They were also taken to the Electronics Communications Maintenance on the base where they met with First Lieutenant Leahy and Mr. Paul Schaen, who wore civilian clothes. Schaen had to "get clearance," apparently from unidentified superiors, to talk with McDonald and Hartmann. The three men went to the Radar Air Traffic Control Center, a joint FAA-USMC facility (McDonald, 1968). This facility had no record of Heflin's UFO on its logs, either. They told McDonald that according to their records, the winds were blowing at 4 knots from the north northwest at the time of Heflin's sighting, which directly contradicted wind information LANS and McDonald had searched out independently. That evening, a LANS meeting was held at the Epperson home for discussion of the Heflin photo case. McDonald's list of the 17 attendees read as follows: "Hartrnann, Idabel Epperson and her daughter, Marilyn; Dr. Robert M. Wood; Maralyn Benton, Wood's secretary; Dr. Darryl Harmon; George Kocher and Mrs. Kocher; Dr. Robert Nathan; the Druffels, Charles and Ann; 594 A. Druffel, R. M. Wood, & E. Kelson the Eugene Coltons; John Gray." Also present were Dr. Stephen Black and Philip Daly, two BBC personnel who were in the U.S. shooting segments for a British UFO documentary. This particular meeting was packed with newcomers, mostly scientists who had been invited to meet McDonald and Hartmann. Some of them were not schooled in interviewing UFO witnesses with proper objectivity and courtesy. When they began interrogating Heflin, the questions flew hot and heavy. Heflin typically kept his cool. Hartmann expressed doubt that anyone could take three Polaroid photos in, roughly, 20 seconds, the conditions under which Heflin said he had taken his first three pictures of the flying object. LANS investigators had asked Heflin the same question 2 years prior, and he had demonstrated his ability to do it, using his own work camera. Another of the visiting scientists asked whether Heflin knew if the object was really coming from directly behind him? Heflin, who wondered privately about the logic of the question, replied, "No, I don't. I didn't see it until it was on the periphery of my left eye." And on and on. John Gray, the principal LANS investigator on the Heflin case, was also questioned closely. Gray was an engineer who checked ground equipment for the Apollo Project at North American Aircraft. His technical qualifications and methodical investigative skills were exceptional. Nevertheless, because the smoke ring in Photo 4 had been photographed over an orange orchard with only a telephone wire and a small branch of one tree as reference points, McDonald asked why Gray wasn't sure which branch of which tree Heflin had photographed, discussing this "problem" for 40 minutes. Gray had identified the tree and the branch 2 years ago when the investigation was in its first stages, but because twigs on fruit trees change over the seasons, Gray was no longer sure exactly which branch was represented in the photo. This answer did not satisfy McDonald, for he wished to make his own check on-site to confirm the elevation and azimuth at which Photo 4 had been taken. This would depend largely on the direction and speed of the wind at the time, and NICAP's and McDonald's own investigations had revealed that the wind had been from the southwest at 4 knots. Yet Hartmann and McDonald had been given conflicting information from the joint FAA-USMC radar facility the prior day. We might wonder who the "Mr. Paul Schaen" mentioned in McDonald's Heflin notes really was, and if he could possibly have been deliberately attempting to throw McDonald and Hartmann off the track. Soon after his return to Tucson, Arizona, McDonald rechecked wind data using his own sources and reconfirmed that the winds were blowing from the southwest at the time and date of the sighting. Robert Nathan, an independent scientist working at a southern California aerospace facility, had computer enhanced the photos for LANS. His interest was purely unofficial and had nothing to do with his work association. At the time, he had considerable private interest in the UFO phenomenon. He often Reanalysis of the 1965 Heflin UFO Photos 595 attended LANS meetings and analyzed several photos for the subcommittee. He had told LANS investigators that his computer enhancement had revealed that the black band around Heflin's UFO was particulate matter-possibly atmospheric pollutants picked up by the craft as it flew through smoggy Orange County. Other scientists present at this January 15, 1968, meeting speculated that the black band was held around the craft by some type of electrostatic effect, possibly associated with the craft's propulsion system. Dr. Nathan had determined that the smoke ring in Photo 4, which he referred to as a "vortex ring," was also composed of particulate matter. He stated that he saw no reason why the "smoke ring" could not logically be the original black band that surrounded the object in the first three photos that had been left behind by the craft upon its rapid departure. He also saw no reason to doubt that Heflin's Photo 4 was part of a set depicting a UFO event, all photographed within 2 minutes. McDonald brought up his growing doubts about what was to him conflicting cloud data. He voiced objection to Photo 4 on the basis that he had received positive information from the most reliable weather sources available that there could not possibly have been any substantial clouds at the time of sighting. He emphasized that he'd cross-checked all possible cloud observation sources concerning Los Angeles area mesometeorology and the role of mean inversion depth and dry suprainversion air. The scientific terms didn't impress Heflin. "The meteorologists are going to have to find some clouds to go in these photos!" he stated firmly. He was not claiming there were clouds in the atmosphere above the dark smoke ring because he simply hadn't noticed. But he had photographed the smoke ring only about a minute after taking his three photos of the UFO, and if the photo contained clouds, then there had to be clouds. LANS investigators backed Heflin. They tried to interpret to McDonald that in their own copies of the fourth photo the clouds were not as dark as in McDonald's copy, again explaining that the various copies had been made from different generations of the originals and had been processed at different degrees of exposure. McDonald was still unconvinced. The question of whether the manufacturer had put numbers on the back of the Polaroid film on which Heflin's photos were taken was also discussed at length. Most of the scientists present at the meeting insisted that all film packs bore numbers, but Heflin stated there were no sequence numbers on the originals. He explained that the Polaroid Special 3000 film packs used by the County Roads Department at the time had no numbers and that this was a serious issue in the department because the work cameras were often exchanged among the employees. Later research by Heflin, LANS members, and McDonald verified that Polaroid ASA 3000 film was manufactured with no sequence numbers around the time of Heflin's sighting. The strange account of a recent visit of a "U.S. Air Force captain" to Helfin's home was also discussed. Heflin had confided the event to LANS. On I 596 A. Druffel, R. M. Wood, & E. Kelson the evening of October 11, 1967, a man in a U.S. Air Force uniform, identifying himself as "Captain C. H. Edmonds," came to his door. His ID had no photo and was salmon and green, like the ID the "NORAD" men had used 2 years earlier. As Heflin talked with him on the porch, he noted that about 30 feet away, directly in his line of sight, was a 1965 or 1966 Chevy parked at the curb. The auto was dark blue with dark lettering on the door, which blended into the surface. Heflin was unable to read it. He saw movement in the back seat of the vehicle, which he assumed was a second man, and a strange purplish glow emanated from that area. Edmonds chatted with Heflin, stating that he'd been involved in numerous important UFO cases, including one in the Bermuda Triangle! He also asked various personal questions, such as what UFO groups Heflin was affiliated with; Heflin told him he'd joined NICAP only. While this was going on, Heflin heard crackles and pops coming from his hi-fi, which he'd been listening to when Edmonds knocked on his door. He'd never before heard interference like this and inwardly wondered whether it was somehow linked to the strange purplish glow in the back seat of Edmonds' vehicle. Later, he speculated that he might have been secretly photographed or recorded. When Edmonds left a few minutes later, Heflin called John Gray and told him what had happened. By this time, he was atypically angry, not only at Edmonds' unwarranted intrusion but also because he'd been having a lot of trouble with his phone. His friends had been complaining they were never able to reach him at home. Heflin had been home at the time they'd called, but the phone hadn't rung. He told Gray he'd had his number changed more than once, obtaining unlisted numbers each time. Somehow, strangers managed to get his new telephone numbers, and the crank calls continued. A neighbor woman had told him that she'd seen Marine Corps and Air Force officers "nosing" around his home while he was at work (McDonald, 1968). After the LANS meeting, McDonald tried to mend his strained relationship with Heflin, pointing out the numerous aspects of the case that were positive. He explained that as a scientist, he needed to maintain strict objectivity in the investigation so that the facts would emerge. He also pointed out that another Orange County multiple-witness case tended to corroborate Heflin's sighting. This incident had occurred during the first week of August 1965, the same week as Heflin's sighting. About 9:00 p.m., Mr. and Mrs. Ralph Joseph were driving on the Santa Ana Freeway in Anaheim, a city about 10 miles from the location of Heflin's sighting. They saw a large glowing object sitting atop high-voltage lines that ran high above the freeway. The object had a rounded dome and was disc shaped with a rounded bottom. It glowed bright white with self-contained light, which did not illuminate its surroundings. Many cars were stopped on the shoulder of the freeway, and motorists had gotten out and were staring at it. The Josephs slowed down but did not stop. At Reanalysis of the 1965 Heflin UFO Photos 597 their closest point, they were about 40 feet from the object yet could hear no sound from it. They judged its size as "much bigger than a car." It was about 3 times the width of the power-pole yardarms and its width was four times its own height. No features, such as seams or "windows," were visible. The Josephs expected that the occurrence would be reported by the media, and they listened to the car radio as they drove, but nothing was broadcast about it. Neither was there any mention in the newspaper or TV news programs during the next few days. Yet the object had been totally visible from a heavily traveled freeway. This phenomenon of underreporting from busy locations is a reflection of lack of clarity of uniform reporting procedure. It is still true in 2000. LANS and McDonald had conducted interviews with the Josephs, which left them satisfied that the witnesses were solid and reliable. Other witnesses to the sighting were sought but without success (Epperson, 1966). By 11:30 the morning after the LANS meeting, a group of varied investigators gathered at the Myford Road site, including the two BBC newsmen. Dr. William Hartmann and one of the BBC documentarians tried to replicate the photos with a similar camera, using both a small model hung on a string and a Leica camera lens cap suspended by a fine thread. Both photos came out surprisingly sharp even though the lens cap was only 2 feet from the camera. The strings used in the first test shot showed plainly, although the thread was less visible. Heflin watched silently. Dr. Robert Nathan, however, had demonstrated 2 years before that there were absolutely no strings or other supporting mechanisms visible in Heflin's photos. Hartmann was fully aware of this. He also was aware that Heflin's camera focused the frame of the truck window remarkably well, considering the fact that other features of the landscape miles away (and everything in between) were also in focus. These facts didn't matter to Hartmann, however. Later, writing up the case in the Condon Report, he judged Heflin's photos "of little probative value," claiming that his replication shot with the lens cap was a good match with Heflin's Photo 2 (Condon, 1969). This is an early example of the naive assumption that just because a photograph could be ostensibly faked, that it was faked. The two BBC documentarians also neglected to show Heflin common courtesy. McDonald wrote a complete account of their interaction in his "Heflin File," which is included in coauthor Druffel's biography of McDonald's UFO research, Scientist vs. the System: Dr. James E. McDonaldk Fight for UFO Science, publication of which is forthcoming. Other aspects of the complex interactions in this complex investigation are included in this book, manuscript and appendix, which reaffirm Heflin's total reliability and honesty. When the two BBC representatives pressured Heflin for a full, filmed interview at the site, Heflin refused, saying that an American producer, John MacDonald, had already done a creditable job for ITV. He suggested they borrow his film. He did allow them to film a very brief segment, in which he stated that he understood why various investigators were interested in his photos and that 598 A. Druffel, R. M. Wood, & E. Kelson people had the right to reach their own conclusions. He also explained how the automatic light meter on his camera had allowed the sky to appear flat and featureless in Photo 1, Photo 2, and Photo 3, which had been taken inside his work van, but showed the overcast sky in Photo 4, which was taken outside the van. Ending his interview with the BBC, Heflin pointed out to McDonald the area of the sky where the object had emitted the bluish-black smoke ring and drove with him to the location where he had taken Photo 4. McDonald estimated mathematically that the smoke ring had been photographed at about 400 feet altitude. The position of the smoke ring convinced McDonald that the wind data gathered by both himself and LANS were accurate, that the wind had been blowing from the southwest, and that the Marine Corps Air Station had given him inaccurate data. When he arrived home in Tucson, Arizona, McDonald tried every way available to him to check out the ID presented by "Captain Edmonds," who had visited Heflin. Because no official investigating agency had identification cards without photos and none were salmon colored, this satisfied him that Edmonds was an imposter. The eerie purplish glow that had apparently affected Heflin's hi-fi remained a mystery. McDonald had been surprised when Hartmann demonstrated that Heflin's 101 Polaroid camera was capable of sharply focused photos at close range and had begun to think that Heflin could have photographed a small model. However, he learned from LANS that two and a half years before the very start of case investigation, LANS photo expert Zan Overall had discovered that Heflin's work camera sharply focused nearby objects; when set at infinity, the photos turned out sharp at 3 feet (McDonald, 1968). There were other interesting aspects of the Heflin photos. Dr. Nathan had noticed an unexplained "blurring" around the UFO, which was not caused by motion of the craft or the camera. As pure speculation, he wondered if the blurring might be due to ionized air around the craft. UFO researchers had long hypothesized that the propulsion systems of UFOs might ionize the air surrounding them while they were in flight. Also, while enhancing the underside of the craft in Photo 2, Nathan had brought out a wedge-shaped portion of light. It was consistent with Heflin's description of a revolving light ray sweeping around the bottom of the UFO.' NICAP photogrammetrists and most UFO researchers remained convinced that Heflin's Photos 1,2, and 3 were most probably genuine, and that Photo 4 was part of a set taken within about 2 minutes of each other. It is impossible, of course, to declare a UFO photo absolutely authentic unless one had the actual UFO nearby with which to compare it. Significantly, Dr. William Hartmann's alleged "replication" of Photo 2 showed no such wedgeshaped feature. ' Reanalysis of the 1965 Heflin UFO Photos 599 The Fort Belvoir Photos McDonald's doubt about Photo 4 served to diminish his acceptance of the case, however. He redoubled his efforts to solve the "smoke ring" problems. UFO literature held references to a few other cases of dark smoke rings associated with UFOs, such as a blue-black smoke ring in Ohio, which had a ball of fire swinging around the outside, described in the Apro Bulletin (Lorenzen and Lorenzen, 1966). But another smoke-ring photo case, occurring at Fort Belvoir Army Base in Virginia, was far more interesting. This case proved to be an important part of the Heflin debate. At 9:00 a.m. on a cold, cloudy day in September 1957, U.S. Army Private George L. Stofko, Junior, witnessed a solid-looking black ring hovering low over a building of the base. Intrigued, he grabbed a camera from his car and took six photos. The first showed the ellipsoidal black ring hovering against a clouded sky. This photo turned out dark on one side, but the second showed the ring clearly. In the third photo, taken just a short time later, the ring was being surrounded by whitish vapor. In the fourth, this apparent "clouding" had almost obliterated the ring, but a curious "ribbed" effect was noticeable as if "projections" on the black ring are poking out through the vapor. In the sixth photo, the black ring was completely hidden in the small oval cloud, which was apparently at much lower altitude than other clouds in the sky (Dell, undated). The Fort Belvoir photos had come to NICAP's attention through NICAP members Z. J. Halaut and Don Berliner, and NICAP Photographic Adviser Ralph Rankow had followed up on the case. He learned from Stofko that the black ring had been viewed by other soldiers, that it seemed "solid," and that it was approximately 60 feet in diameter. This group of soldiers were on a brief break from their work in the Post Engineer Section and viewed the phenomenon for 1-2 minutes. They had to return to work and could not stay to see what eventually happened to the small, oval cloud. Stofko had no idea what the "object" was. He told Rankow that some of the other soldiers had also taken pictures, but he wasn't well acquainted with the coworkers and hadn't kept track of them. The witnesses agreed among themselves that they had inadvertently witnessed a secret military experiment and did not tell any officials on the base that they'd taken pictures (McDonald, 1968b). Several years passed before NZCAP learned about the Fort Belvoir photos. When contacted, Stofko couldn't find the negatives. They had been misplaced, because he had been transferred several times since the occurrence. He'd shown the photos to various friends over the years but had met with ridicule and disbelief until he showed them to NICAP. Stofko's first two photos were strikingly similar to the black ring in Heflin's Photo 4. Stofko's ring had appeared "solid," and it maintained its ovoid shape during the 1-2 minutes it was in sight. Heflin, too, had noticed the "solid" appearance of the smoke ring associated with the UFO he'd photographed, even 600 A. Druffel, R. M. Wood, & E. Kelson though it had been blown out of shape by 4-knot winds. However, Stofko had taken his pictures in meteorological conditions that were cold, damp, and windless. McDonald reasoned that these conditions could have helped produce the clouding effect. He was puzzled, however, by what he termed the "rosette shape" in the sixth photo. His meteorological expertise told him that clouds that formed under such conditions should have flattened bottoms, yet this small cloud was oval. McDonald pursued the puzzle relentlessly, thinking that if he could solve the Fort Belvoir incident, it might shed light on Heflin's fourth photo. Charles A. Maney, a professor at Defiance College in Ohio and a top NICAP official, was a careful researcher who also took great interest in the Fort Belvoir photos. He suggested to Ralph Rankow that ionization of the air near the aerial object had produced the vapor that formed around and inside the ring. He also referred to a 1952 newspaper clipping, which described experiments by Noel W. Scott, an army physicist at Fort Belvoir. Scott had produced a partial vacuum in a bell jar then allowed a tiny stream of ionized air to enter from the bottom of the jar, producing a cloud of vapor that later filled the jar. According to this newspaper article, Scott produced orange-red fireballs, discs, and other shapes of various colors, according to what gas he introduced into his bell jar. This article stated that Scott "was trying to simulate a flying saucer" and claimed that these "miniature masses of illuminated air" had sufficient body or substance to be picked up on radar. Scott's experiments were cited by the Air Force as "possibly" explaining some of 1952 sightings, when groups of UFOs overflew the Capitol and White House, with the result that military jets had been sent up to investigate them (Toledo Blade, 1952). McDonald was not convinced that the Fort Belvoir photos depicted a true UFO, but he considered them an interesting problem of cloud physics. Seeking more data on ring-shaped UFOs, he learned from Dr. Jacques Vallee that donut-shaped UFOs had been reported in France. However, in these cases, the hole was about one third the diameter; they did not have the "hula-hoop" appearance of the Fort Belvoir object. In the meantime, the Fort Belvoir photos were published in a newsstand publication, Dell UFO Magazine #4, which seemed to knock information loose from other quarters. A man calling himself Jack Strong wrote to Ralph Rankow, stating that he was an ex-Army man and had seen the smoke ring while stationed at Fort Belvoir around the same time Stofko took his photos. In technical terms, Strong stated that the black ring had been caused by an "atomic bomb simulator," which Strong claimed was a device consisting of a charge of high explosives that produced a small black mushroom cloud. Strong contended that on the morning Stofko photographed the black ring, the atmosphere was "dead calm" and the air "cold and moist." He stated that at a height of about 40 feet, the "cap" of the mushroom developed into a perfect Reanalysis of the 1965 Heflin UFO Photos 601 smoke ring vortex, detached itself from the main column, and continued to rise into the cold, saturated air. Strong's letter also alleged that the vortex ring swept moisture out of the air and that the smoke particles on the periphery of the vortex provided condensation nuclei for the formation of a visible cloud, forming "the lens-shaped body" with the vortex still visible within it. He speculated that the "ribbing effect" in the Fort Belvoir photos may have been due to slight variations in velocity at different points around the periphery of the ring. Even though Strong's letter to Rankow was expressed in technical terms, he claimed not to know enough about atmospheric physics to explain why the center of the ring clouded in so rapidly. He stated, however, that "the presence of a stable ascending vortex continuing particles suitable for condensation nuclei must, in connection with such conditions, be very rare." Rankow wondered if the government might be behind Jack Strong's explanation. He asked McDonald to study Strong's letter and determine if it stood up from a technical point of view (Rankow, 1967). McDonald wrote back, saying I find it ludicrous that the Army would be using anything like what he [Jack Strong] describes as "the atomic bomb simulator." Disney might, the Army scarcely, in my opinion. Secondly, the suggestion that anyone could find a means of generating a "perfect smoke ring vortex," with the sharp outlines and a perfectly circular form that appears in [Stofko's] Photos 1 and 2, is not reasonable. Vortex rings are intrinsically quite unstable and maintain circularity for only a very brief time.. . . [Alny vortex ring of such small dimensions and low altitude would be most unlikely to produce condensation of atmospheric moisture, and even if it did, would generate a much more fragmentary mass of cloud than is exhibited by the Stofko photos.. . . [All1 ordinary clouds of any substantial size (restricting attention to cumuliform types) develop essentially flat bases, whereas the base of the cloud rosette in the Belvoir photos is distinctly rounded. [Strong's letter]. .. strikes me as having been formulated by somebody with rather more than casual knowledge of vortex ring dynamics. [He] ought to know enough to realize that parts of his argument are just not very sensible. That inference may be quite incorrect, and for that reason, I repeat my request that you keep all of this confidential for the time being. I think that the suspicions that you hint in your letter are by no means unreasonable. The letter has a peculiar ring (McDonald, 1 9 6 8 ~ ) . This represents, to our knowledge, the first time McDonald expressed in writing the possibility that the government might be seriously involved in a UFO cover-up. Always before, McDonald preferred to argue that the government's apparent neglect of the UFO question was due to a "grand foul-up" rather than a widespread official cover-up, reflecting the widespread attitude of the academic community. It is common even today that academics have little knowledge or respect for the astonishing achievements of the intelligence community. McDonald set about trying to track down Jack Strong. He also expressed doubt about the Fort Belvoir photos to Dr. William Hartmann. Contacting Fort Belvoir officials, Hartmann was told that the black smoke ring photographed by Stofko had resulted from a "demonstration device" for visiting firemen and 602 A. Druffel, R. M. Wood, & E. Kelson that the main function of this device was not to simulate a mushroom cloud but rather, under favorable atmospheric conditions, to blow smoke rings. Hartmann visited Fort Belvoir officials on-site. They told him that the device had been in use for many years, including September 1957, but was not presently in operation. He was told by an unnamed sergeant at Fort Belvoir that "there were some days when it [the device] didn't work at all." In a long memo to Hartmann, McDonald wondered how it ever worked, considering the physics involved. He also expressed his opinion that neither explanation, i.e., "atomic bomb simulator" or "demonstration device," was convincing (McDonald, 1968d). An Andrew H. Taylor of Pittsburgh, Pennsylvania, also wrote to Dell UFO Magazine, claiming he was one of a group of soldiers who set off an explosive device at Andrews Air Force Base in May 1959, during an Armed Forces Day show. He described how he and his crew had filled ten 55-gallon drums with No. 2 diesel fuel, arranged them in a 60-70-foot circle, and set them off by a small charge, to simulate an atom bomb explosion. Taylor claimed that a black ring formed, exactly like Stofko's photos and that a cloud had developed around the ring. Bill Hartmann was ready to accept the Army's explanation for the puzzling Fort Belvoir photos, but McDonald was not convinced. Army officials offered no written reports describing the history of the device, no photos or other documentation proving it had ever e x i ~ t e dYet Fort Belvoir officials claimed .~ that their device could, under the right atmospheric conditions, produce a stable smoke ring that remained perfectly circular and tight while rising hundreds of feet into the air. These claims disturbed McDonald. His own calculations on the Stofko photos indicated that the Fort Belvoir smoke ring retained its exact shape while rising to an altitude of a few thousand feet. If there had been enough moisture in the surface layers of the air to produce condensation of the vortex ring, as stated by Jack Strong, there should have been clouds at roughly the same altitude as the ring itself. Instead, the ring cloud was distinctly lower than the background cloud deck. McDonald wrote to a colleague, Dr. Clayton H. Reitan of the Department of Meteorology at the University of Wisconsin, asking him to try to locate Strong at his given address in Wisconsin and stating his concern that Strong's explanation did not seem meteorologically reasonable. He added a confidential postscript to Reitan, pointing out that some of the other information that has been recently obtained from military sources on the background of these photos has a slightly phony ring. There is, in the back of my mind, a certain suspicion that "Jack Strong" might somehow be involved in these maneuvering~ (McDonald, 1968e). The authors request any reader who might have personal knowledge that these devices existed at Fort Belvoir, or at Andrews Air Force Base in California, to please contact them at 257 Sycamore Glen, Pasadena. CA 9 1105. Reanalysis of the 1965 Heflin UFO Photos 603 With Reitan's help, both "Jack Strong" and "Andrew Taylor" were contacted, and McDonald held long phone conversations with both of them, which he recorded with their permission. In a turnaround that greatly surprised NICAP investigators and other UFO colleagues, he apparently accepted Strong and Taylor as independent, reliable witnesses and seemed assured of their intelligence, observational powers, and honesty. Yet Strong, Taylor, and Fort Belvoir officials never offered any firm documentation, and many NICAP investigators and other UFO researchers remained unconvinced. Strong's description of windless, cold, and humid meteorological conditions on the day he reportedly saw the smoke ring at Fort Belvoir also apparently convinced McDonald that a stable smoke ring could result from an explosion of a device such as described by Taylor. In addition, a written confirmation to Ralph Rankow from Colonel J. H. Jackson at Fort Belvoir led McDonald to accept that a stable vortex ring could form and be enveloped in cloudlike vapor under the right circumstances (Jackson, 1968). McDonald rapidly developed an intense interest in researching this new aspect of vortex rings, for to him it was new scientific information. By early 1971, he was planning to present a proposal to the National Science Foundation (NSF), which included the subject of vortex rings. A brief note found in his files during archiving revealed his intentions, but no evidence has been found that the NSF ever accepted the proposal (Druffel, 1997). McDonald suspected that Heflin had photographed the smoke ring while attending an air show at some different location and was attempting to pass it off as part of a set depicting the overflight of a UFO. Many UFO investigators, including those from NICAP, objected to the "atomic bomb simulator" explanation because there was no documentation that it had actually existed, only verbal testimony and Colonel Jackson's letter. Strong's and Taylor's credentials were never completely checked out, and Colonel Jackson's verity on the matter could not be confirmed. NICAP investigator John Gray wrote to McDonald regarding Photo 4, emphasizing two points: First, the most logical theory pertaining to the origin of the black smoke ring is that it was the same band of particulates that enveloped the object in Photos 1 through 3; and second, McDonald's objection to the clouds in Photo 4 could be answered by the fact that the cloud formations depicted in McDonaldS enlarged copy of that photo were darker than those in NICAP-LANS' copies because of differences in processing. Gray said If Photo 4 is not acceptable evidence to support the sighting, then.. . it should be set aside as questionable evidence considered not relevant to the case. With the human element involved in an inexplicable event, science should not expect its explanation to be more than the expression of an opinion. This is the practice you have followed, and rightly so, in your public statements concerning the reality of UFOs. With Rex Heflin's reputation at stake, any explanations issued by science must be based on certainty. If his sincerity cannot be disproved or his insincerity cannot be proved, then it behooves 604 A. Druffel, R. M. Wood, & E. Kelson science to shelve his case until disclosure of further evidence rather than run the risk of besmirching his character (Gray, 1968). Ground Saucer Watch Declares Heflin Photos Hoaxed Idabel Epperson, John Gray, Zan Overall, and other LANS members remained convinced of Heflin's integrity, as did most researchers in the UFO field. His first three photos of the metallic craft survived as an example of an apparently genuine UFO and continued to be printed in UFO research journals and magazines world wide. McDonald died in June 1971, but his doubts about Photo 4 caused some lay researchers to attempt reanalysis of the entire set. In the mid-1 970s, William Spaulding, a lay UFO researcher who headed a group called Ground Saucer Watch (GSW) in Arizona, obtained copies of Heflin's photos from an unspecified source and had them computer enhanced. Like most copies circulating at that time, they were probably at least fourth generation. First generation copies were rare and carefully guarded by the researchers who possessed them, because the originals of Photo 1, Photo 2, and Photo 3, which showed the flying craft, were no longer available. Heflin had retained his original of Photo 4, which showed the smoke ring. The GSW computer enhancement was apparently conducted by Fred Adrian, the organization's Photographic Consultant. In a mid-1970s issue of the GSW "News Bulletin," the Heflin photos were included among a list of UFO photos, which in GSW's words, "represent both crude and grandiose hoaxs [sic] or photographic anomalies and should not be considered evidence of UFO existance [sic]" (Spaulding, 1975). This widely disseminated assessment of the Heflin photos was based on what Spaulding termed "a string," which allegedly extended from the top of the photo down to the object. Spaulding claimed that "the string" had been found on four separate sets of the Heflin photographs. Idabel Epperson and David Branch of LANS and researcher David Schroth, among others, challenged Spaulding's findings, pointing out that several photo experts, including Robert Nathan, had looked for but found no evidence of a string or any supporting mechanism associated with the object. All of NICAP-LANS' photogrammetrists had worked with confirmed first-generation copies, which had been made directly from the originals before they disappeared. The strong objections to GSW's methods of analysis resulted in an admission by Spaulding that possibly the linear structure seen on his enhanced copies was a scratch rather than a string. He also stated that there was a "slim chance" that all of the four sets of photographs were from the same unspecific source. Spaulding never publicly withdrew his assessment that the Heflin photos were hoaxed, however, and the situation caused continuous controversy in the UFO research field. Many otherwise objective researchers tended toward the hoax explanation while the original investigators and many other Reanalysis of the 1965 Heflin UFO Photos 605 UFO researchers continued to regard them as among the best UFO photos ever taken. The Original Photos Reemerge The disappearance of the original Polaroid photos thwarted attempts to restudy them in greater depth as photogrammetric technology advanced. Heflin continued to work for the Orange County Traffic Department for 15 additional years. Eventually, after 3 decades of working outdoors on the streets and freeways of southern California, he developed a serious health condition, i.e., an accumulation of tetraethylene lead in his bone marrow. It is a condition that is being diagnosed in increasing numbers of people who have worked for decades outdoors in heavy automobile traffic. It has no cure and no standard medical treatment. Suffering from fatigue, breathing difficulty, and other symptoms associated with this medical condition, Heflin moved to a small town in northern California around 1985. Here, the air was relatively pure, facilitating breathing, and an experimental treatment for the condition was available at a local hospital on an outpatient basis. His doctors informed him that lead poisoning in bone marrow, although becoming more and more prevalent, was not recognized as a verified medical condition by the AMA or any health insurance programs, including Medicare. The serious health conditions it caused, such as high blood pressure, strokes, and other debilitating illnesses, were not covered either. Heflin became solely responsible for his own medical bills associated with the bone marrow condition, which he could sorely afford on his fixed retirement income. One day in 1993, while Heflin was sitting in his northern California home, the phone rang. When Heflin answered, a woman's voice asked, "Have you checked your mailbox lately?" Before he could answer, the call was abruptly terminated. He went to the mailbox and found it empty. About half an hour later, the same unidentified woman called again with the same question. This time, Heflin found a plain manila envelope in his mailbox. It had no postage or other marks that would indicate how it had been delivered. To his surprise, it contained what appeared to be his three lost Polaroid photos. He examined them carefully and concluded they were indeed the three original photos that had been taken by the self-styled NORAD agents in 1965. Their size, texture, and general appearance precisely matched Photo 4, which he still retained. He had numbered the photos when they were first in his possession, marking them 1 through 4 with blue ink in the lower left-hand corner. The three photos that had been mysteriously returned had his markings 1 through 3 in the identical place as the 4 on his smoke ring photo. The 4 is the only marking on the back of this photo. There were, however, other markings on the backs of the originals of Photos 1, 2, and 3. Each had "original" printed in capital letters across the top, by someone who had used a white grease pencil; the imprint of the letters showed 606 A. Druffel, R. M. Wood, & E. Kelson slightly on the photo side. Each of the three originals also had the number 13 written lightly on the back with soft black pencil; these markings did not disturb the photo side. Heflin had not put any such marks on the backs of the photos; they had apparently been marked by those unknown persons who had possession of them for the previous 28 years. There were also several distinct fingerprints on the surfaces of the photos. The photos were in very good condition considering their age and apparently had been well taken care of. Heflin's Photo 4 had developed light brown stains around the smoke ring. He considered this the result of its age and due to the fact that several photo analysts had made close-up copies of it, subjecting the smoke ring image to bright light. The three returned originals also showed similar light brown stains, particularly along the lower portions. The object (UFO) in these returned photos was not so deeply stained as the smoke ring in Photo 4. Heflin was now in a quandary. To add to his financial problems, his pension and other funds resulting from his long-term employment in Orange County had been temporarily cut off because that California county was in bankruptcy. Heflin suspected he would not live many more years and wondered what to do with the newly returned originals of his UFO photos. He, like many UFO researchers of his acquaintance, considered them to be perhaps the finest photographic UFO evidence in existence. Reanalysis of the Heflin Photos in the 1990s Soon after the original Polaroid photos surfaced, one of the authors of this paper (R.M.W.), an aerospace scientist and physicist who had conducted an independent analysis of the photos in the 1960s, learned about their return from author Martin Caiden. Another of the authors (A.D.), a veteran UFO researcher who participated in the original LANS-NICAP investigation, contacted Rex Heflin with the suggestion that the originals be analyzed with state-of-the-art computer-enhancement equipment. The third author (E.K.) offered his expertise in computer enhancement techniques. Heflin agreed to a reanalysis of his photos. He also expressed his concern that all four originals be preserved for the future, so that they would be available to other scientists who wished to study them, particularly as new photogrammetric technology develops. Heflin knew and trusted Wood and Druffel as veteran UFO researchers who were thoroughly familiar with the case. He entrusted the four originals into Druffel's care, who in turn offered him $850 for his upcoming series of medical treatments-an act of friendship rather than a formal sale. They are now among the acquisitions of Druffel's SKYNET Library. After the authors' analyses of these photos are completed, first-generation copies will be available to other scientists and accredited researchers who wish to study them. It was now possible to reanalyze the photos, working from the originals rather than copies of uncertain generation. Also highly advanced computer- Reanalysis of the 1965 Heflin UFO Photos 607 enhancement equipment is now accessible, technology not available in the 1960s and 1970s. Our three-person team restudied the facts of the case from original NICAP documentation and applied our efforts toward unanswered questions, such as the problem of the Photo 4 smoke ring, which McDonald had raised; the problem of the apparent differing sky conditions; the "string," which Spaulding and GSW claimed to have found; the "wedge of light" on the bottom of the UFO in Photo 2; and any other aspects that could be found andlor answered. General Techniques Photographic preparation. In preparation for digitization of the images, several large high-resolution prints were prepared from the original Polaroid photos. A supervised professional photographed the original photos and prepared negatives by standard techniques. Large prints (8 by 10) were then made with standard exposure times that gave optimal overall contrast, and then two additional sets of prints were prepared with twice and half the standard exposure times. Though the standard exposure prints best reproduced the main features of the photographs, the longer and shorter exposures better represented contrasts in the darker and lighter regions, respectively. Image digitization. All of these prints were then scanned into digital format with a commercial UMAX 6-SE scanner interfaced to a Macintosh PowerBook 1400c running System 7.5.3 and Adobe Systems PhotoShop 4.0.1 with the Vistascan 2.3.3 scanner control plug-in. The sensitivity of the scanner was calibrated against a pattern of standard grays (0%, 20%, 40%, 60%, 80%, 100% reflectance), and a gamma setting of 1.60 was found to deliver a linear digital response over this pattern. This ensured that the digital data best represented the photometric intensities in the image. Each 8-by-10-inch print was then scanned at 300 dpi (dots per inch) spatial resolution and 16-bit grayscale resolution. Since the grain of the Polaroid resolution was visible at 150 dpi, higher resolutions were not practical. The resulting data were stored in standard, uncompressed TIFF graphics files. Though this resulted in exceptionally large files, this format ensured data integrity. Nonlinear scale -stretch enhancement. These data were initially examined within PhotoShop and specific issues were explored through nonlinear scalestretch enhancement. The latter technique is particularly important in emphasizing low-contrast details within intermediate shades of gray while retaining the overall contrasts of the original image. Nonlinear scale-stretch enhancement involves the reassignment of the narrow range of gray that contains the detail in question to a much wider range. Intensities outside this range are linearly compressed to tighter ranges. For example, to emphasize detail consisting of shades of gray between 45% and 55% reflectance, these data can be reassigned as shown in Table 1 and Figure 5. Essentially, detail represented by variations of only 10% reflectance can be enhanced to a plainly visible 50% variation. This technique is demonstrated in Figures 6 and 7. Figure 6 contains 608 A. Druffel, R. M. Wood, & E. Kelson TABLE 1 Reassignment of Intensity Values by Nonlinear Scale Stretching Original reflectance range Reassigned reflectance range a bright "A ' and a dark "C" against a gray background. Between these letters is a diffuse "X" defined by a contrast of only 10% reflectance. Reassignment of the intensity values as detailed in Table 1 and Figure 5 results in Figure 7, in which all three letters are clearly visible. This enhancement reveals the "X" without sacrificing the original clarity of the "A ' and "C." This is not yet a packaged technique and does require some experience. When applied correctly, the enhanced image still fairly represents the original data. 7 7 General Observations of Original Images HeJlin Photo I . The standard exposure image (Figure 1) clearly shows the well-known object. The object (enlarged in Figure 8) has a general shape like that of a hat seen from the side with a domelike feature on its top surface. The top edge and lower brim appear to be reflecting the diffuse sunlight, suggesting they have a metallic luster. The large reflection off the top of the object Input Reflectance (Percent) Fig. 5. Plot of reflectance in nonlinear scale stretching that emphasizes detail between 45% and 55% reflectance. Reanalysis of the 1965 Heflin UFO Photos 609 Fig. 6. " O r ~ g ~ n a image with diffuse "X" between a bright "A" and dark "C." l" and the somewhat darker region just left of it is consistent with a dome-like structure. The surrounding context o f t h e object reveals a good deal about the conditions of the sighting. This picture was shot through the front window of a truck parked on the side of a road lincd with telephone poles. I n the distance, there appears to be a freeway with three cars and one truck clearly discernable. This is consistent with the picture being taken north on Myford Road with the Santa Ana Freeway in the distance. The obvious shadow of a pole in the foreground and the regular spacing of others alongside the road suggest that the truck was parked near the base of a pole. The angle of the shadow is consistent with the claimed date, time, and location of the sighting. The contrast of the shadow is weaker than that expected for direct sunlight; the reflectance within the image of this shadow is 50% 60% of that o~itside suggesting the shaded area was it, lit with a significant a~nourlt indirect light and the direct sunlight was someof what attenuated. This seems consistent with a thin haze or a lightly overcast sky. The intensity of light on the road stretching into the distance appears to be Fig. 7. Nonlinear scale-stretch enhancement (Table 1) of "X" in Figure 6. 610 A. Druffel, R. M. Wood, & E. Kelson Fig. 8. Enlargement of ob-ject in Heflin Photo I (Figure I). relatively even indicating that the hazelovercast is also mostly uniform. Also of note, distant features (including what appears to be a line of trees) seem partly obscured by a ground level haze. The brilliant and directional reflections from the rim of the object suggest that the incident light is somewhat directional, and it too should be casting a shadow. There is indeed a diffuse dark patch to the left of the white pipe below the object, but it is not clear if this is indeed a shadow. If this patch were the shadow of the object, this would place the object almost at the distance of the pipe. Hqflin Photo 2. This picture (standard exposure shown in Figure 2) was clearly taken out the passenger side of the tr~lck still parked on the side of a road lined with poles (as can be seen in the mirror reflection). Given the apparent shape of the object from the first photograph, this image seems to show the object (enlarged in Figure 9) with its shaded underside tipped toward the camera but still showing its upper rim. Like the first image, this scene appears to be lit with diffuse, mostly indirect light consistent with a hazy or lightly overcast sky. The further pole visible in the rear-view mirror seemed to cast a diffuse shadow though the closer one did not cast a clear shadow. The light in- ~ Reanalysis of the 1965 Heflin UFO Photos 61 1 Fig 0 l~nlargemcnt object in Iieflin Photo 2 (Figurc 2) of I tensity in the shadow of the f i ~ r t l ~pole appeared to be about 70%-80% of er that outside it, suggesting less direct light than in the first image. This variability in the shadows suggests the overcast may have been broken enough to allow through variable amounts of direct light while the indirect light remained roughly uniform. As in the first image, distant tree lines are partly obscured with ground-level haze. Neflin Photo 3. This photograph (standard exposure shown in Figure 3) is much ilke Heiiin Photo 2 In that i t was taken out the passenger side window at the same scene except from a slightly different angle. The foreground shrubbery and background features in Heflin Photos 2 and 3 easily match indicating the truck had not moved. Using the background tree lines as a reference, the object appears further in the direction faced by the truck. When measured against background features, the relative diameter of the disk rim appears to be nearly identical in Heflin Photos 2 and 3, suggesting the object is about the same distance in both photographs. Given the apparent shape of the object from the first photograph, the object (enlarged in Figure 10) appears to have its upper surface slightly tipped toward the camera. The diffuse general light- 612 A. Druffel, R. M. Wood, & E. Kelson Fig. 10. Enlargement of object in Hcflin Photo 3 ( F ~ g u r 3). e ing and ground-level haze in this image appear almost identical to that in Heflin Photo 2 suggesting similar conditions. Overall, these photographs appear to have been taken in the same location and under the same hazylovercast weather conditions. The shape of the object is also consistent through the images. These features support the claimed continuity of the Heflin Photos 1-3 sequence. Heflin Photo 4. This photograph (standard exposure shown in Figure 4) exhibits a dark, distorted ring against a patchy overcast sky. The ring consists of a thin puckered torus surrounded by an uneven sheath of haze. Its general appearance seems consistent with the slow disintegration o f a tightly circulating smoke ring. As mentioned, the obvious mottled appearance of the background sky has led some researchers to challenge the claim that this photograph was part of the same sighting. This issue is addressed in the next section of this paper. Enhancements qf'the Skj1 Around the Object in Hejillin Photos 1, 2, arld 3 It is important to note that the Polaroid camera used to take these photographs used a unidirectional automatic light sensor to set film exposure. The Reanalysis of the 1965 Heflin UFO Photos TABLE 2 Details of Nonlinear Scale-Stretch Enhancement of Heflin Photo 1 Original reflectance range Scaled reflectance range 613 Note: Results shown in Figures 11 and 14. shaded interior of the truck likely led to a low-light reading and inadvertent overexposure of the first three images. Such an overexposure would "wash out" important contrasts in the relatively bright sky but not obliterate them. To reveal the contrasts of the background sky, nonlinear scale-stretch enhancement techniques were applied to the longer exposure reproductions of the first three Heflin images as outlined in Tables 2 through 4. (The longer exposures accentuate contrasts within the lighter regions.) The results from these enhancements of the Heflin Photos 1, 2, and 3 are shown in Figures 11, 12, and 13, respectively. The stretch parameters were selected from average intensities representing the light and dark regions of interest. Figures 14 through 16 show corresponding enlargements of the object and surrounding sky. The first important observation that can be made from these enhancements is that no supporting strings or wires are evident in the background sky around the object in any of the photographs (Figures 14 through 16). Since the enlargement and enhancement clearly show the grain in these photographs, further scale-stretch enhancement would be unproductive. Prior claims of a supporting wire in the Heflin Photo 1 were likely due to problems introduced during multiple-generation reproduction and/or handling of intermediate generations. The second important observation is the mottled texture of the background sky in all three photographs. Though some of this mottling may be due to degradation of the photographs, comparison of the second and third images (Figures 12 and 13) shows general similarities in the background sky. This is TABLE 3 Details of Nonlinear Scale-Stretch Enhancement of Heflin Photo 2 Original reflectance range Scaled reflectance range Note: Results shown in Figures 12 and 15. 614 A. Druffel, R. M. Wood, & E. Kelson TABLE 4 Deta~ls Nonlinear Scale-Stretch E~ihancemcntof Heflin Photo 3 of Original reflectance range Scaled reflectance range Note: Results sliown III F~gures 16, and 17 13, all, this seelns to support k-leflin's account of overcast conditions during the sighting. Surprisingly, the enhancement of Heflin's Photo 3 revealed what appeared to be a tattered smoke trail (visible in Figure 16 and better shown in Figure 17). The position of this trail is consistent with the deduced travel of the object roughly in the direction faced by the truclc (discussed above). No trail is evident in either of the Heflin Photo 1 or 2. This is consistent with Heflin's claim that the object blew off, or otherwise left behind a ring of smokelike vapor as it moved quickly away. The presence of smoke in Heflin Photo 3 also offers some continuity with the smoke ring pictured in the Heflin Photo 4. Fig. 11. Nonlinear scale-stretch enhancement ('Table 2) of background sky detail in Heflin Photo 1 (Figure I ) . Reanalysis of the 1965 Heflin UFO Photos 6 15 Fig. 12. Nonlinear scale-5trctch cnhancclncnt ('Table 3) of background sky detail In IIeflin Photo 2 ( F ~ g u r 2). e Fig. 13. Nonlinear scale-stretch enhancement (Tablc 4 ) of background sky detail in Heflin Photo 3 (Figure 3). 616 A. Druffel, R . M. Wood, & E. Kelson Fig. 14. Nonlinear scale-slretcli e~~liancement (Table 2) of background sky d c t a ~ l Hcflin Photo In 1 (Figure 8). Fig. 15. Nonlinear scale-stretch enhanccnient ('Fable 3 ) or background sky detail behind object Iiefl~n Photo 2 ( F ~ g u r e 9). kt1 Reanalysis of the 1965 Heflin U F O Photos 617 Fig. 16. Nonl~ilear mile-5tretch cnhanceincnt ( Iable 4 ) of background sky detall behind object In IIcflin Photo '3 ( I Igure 10) As noted earlier, another interesting feature of the Heflin account is the description of a rotating bean1 of'light f r o n ~ underside of the object about the the time of the second photograpl~.To best reveal such a feature, the enlarged view of the object from the Heflin Photo 2 was subjected to nonlinear scalestretch enhancement (Table 5 ) . To test for this possible feature, nonlinear scale-stretch-enhance~nent techniques were applied to the short-exposure rcproductions of the second Heflin image, as outlined in Table 5 . The result is shown in Figure 18. This image does exhibit a line or wedgelike feature in the lcat to rnpper left porticw ofthc objcct's dark undcl-<id2 Tk115 supports X:lefli~~'s account. TABLI: 5 Details of Nonlinear Scale-Stretch Enhancement of 1 leflin Photo 2 Original reflectance range Scaled reflectance rangc Note: Results shown in Figure 18. 618 A. Druffel, R. M. Wood, & E. Kelson Fig. 17. Nonlinear scale-stretch enhancement of background sky detail b e h ~ n d object in Heflin Photo 3 showing length of "smoke trail." Further Enhnnc.ernent While direct examination and nonlinear scale-stretch enhancement of the Heflin photographs allow a number of important issues to be explored in this case, enhancements ofthe object itself are still b l ~ ~ r r e d possible focus error, by lens aberrations, motion, and particle content in the air. Detailed examination of this blur by techniques such as edge analysis could reveal important information abo~lt object such as its distance from the camera, physical size, the and relative motion. This analysis would also facilitate image restoration through public domain techniques such as the Maximum Entropy Method (MEM). The MEM would both confirm characterization of the blur distorting the object and could clarify the image to reveal important details of the object. This work is underway and will be reported in a later communication. Reanalysis of Heflin Photos Answers Historical Objections As noted above, the computer-enhancement process established what appears to be a definite link between Heflin Photos 3 and 4 (Figures 3 and 4). The absence of a definite link had eluded James E. McDonald during the 4 years he studied the photos. The trail of black particulate matter in Heflin's Photo 3, which is seen streaming behind the unidentified craf't (Figure 16), and which is also denser than the immediate area just behind the craft (Figure 17), confirms Heflin's statement that the ob-ject left behind the smoke ring as it was departing. Although McDonald's refusal to accept Heflin Photo 4 as part of a set was directly responsible for loss of interest in the photos by some persons in the UFO field, this discovery is very important in reestablishing the series of Reanalysis of the 1965 Heflin UFO Photos 619 Fig. 18. Nonlinear xale-\trctcli enhancement ol'ctblect underside in Hcfllli Photo 2. events, as described by Rex Heflin. It also solidly reconfirms Heflin's integrity and reputation for those who had tended to believe that lie had "hoaxed" Photo 4. In the raw data frorn the third Heflin photograph, the intensity values for the trail and background sky are close enough to each other so that the corresponding shades of gray are difficult to distinguish. The trail readily appeared, however, when a computer program reassigned the output intensity range such that the darkest sky pixel becanie black and the brightest sky pixel became white. In the processed image, the new shades of gray were easily distinguishable and the trail was clear. Ileflin snapped his Plioto 3 just bet'orc the lJFO cinitted the smoke ring. In coinptxter-enhanced enlargcd images, t l ~ c tinaterial in the smoke trail appears very similar to the particulate matter in tlie black band surrounding the object, as well as the particulate matter in the smoke ring (Figures 4, 16, and 17). This suggests that the black ring around the object was beginning to separate h ~ f i w ~ the object emitted the circular sinoke ring. If McDonald had had access to this information and to the infor~iiationabove about the sky background, his doubts about Heflin's Photo 4 would most probably have been satisfied, and he would have been able to accept it as part of a set taken within a time frame of 2 minutes, as estiniated by Heflin. Continued computer enhance~nent should help us determine the size and distance of the object. Author Kelson has independently detected an unusual "blurring" around the object, particularly in Photo 1, which is not due to ordinary phenomena such as motion or camera focus blur. This problem has impeded efforts to positively confirm size and distance. This blur issue correlates with Robert Nathan's finding, 30 years prior, of a "fuzziness" in the craft 620 A. Druffel, R. M. Wood, & E. Kelson image, which he could not relate to motion either of the camera or of the craft. Just what causes this "blurring" remains unexplained. Contrast studies of the sky backgrounds in all four Heflin photos revealed similar overcast conditions in all of them. Thus, the major point of contention, which prevented James E. McDonald and others from accepting the fourth photo as part of a set, has been answered. Our analysis has also revealed the presence of the "wedge of light" on the dark bottom of the craft in Photo 2, in exactly the same position where Robert Nathan first detected it (Figure 18). This light feature was also detected by author Robert M. Wood in 1968, using the Marine Corps' copies. And our state-of-the-art computer enhancement has failed to reveal any sign of a string or other supporting mechanism, thus verifying NICAP's original analysis and the analyses of other early researchers who studied the photos. Our own enhanced version of the same area puts to rest William Spaulding's contention that the Heflin photos were hoaxed by photographing a small model hung on a string. It also renders unacceptable William Hartmann's "replication" of Photo 3 in the 1969 official report of the Condon Committee (Condon, 1969). Some Puzzles Still Unanswered Regarding Fort Belvoir Photos Although our reanalysis of the Heflin photos has answered most of the questions posed in the 1960s and 1970s, some puzzles connected with the whole milieu remain. For instance, the "projections" on the Fort Belvoir smoke ring were never explained to anyone's satisfaction. As described above, they were regularly spaced, appeared solid, and seemed to be projecting out of the vapor cloud. It must be remembered that Jack Strong's easy "solution" of this feature had seemed so implausible to McDonald that he warned his colleague, Clayton H. Reitan, to be wary of possible "phony maneuverings" by Strong. The fact that he later conceded that vortex rings could remain stable under the right meteorological conditions in no way explains the regularly spaced "projections" in the Fort Belvoir photos. Information about UFO studies at Fort Belvoir, which has surfaced in the field during the past 10 years, was not available to McDonald. In 1985, researcher William L. Moore obtained a formerly secret document, which reveals that the 1127th Field Activities Group (FAG) at Fort Belvoir, Virginia, was an Air Force special projects group connected with UFO investigation. The partially declassified document is entitled "History: HQ 1127th USAF Field Activities Group, Ft. Belvoir Virginia: 1 July-31 December 1960." Although parts of the text are blacked out, it divulges that the 1127th conducted UFO investigations out of Fort Belvoir as far back as 1959 and possibly even earlier (U.S. Air Force, 1960). Was the 1127th UFO study group active at Fort Belvoir in September 1957, when six "smoke ring" photos were taken by Private Stofko? Were the Fort Reanalysis of the 1965 Heflin UFO Photos 621 ulator?" Why did a vortex ring from an ordinary explosive device develop regularly spaced projections?3 Might the Fort Belvoir photos have been associated in some way with UFO activity, just as Heflin's "smoke ring" was associated with a solid, metallic-appearing UFO? The fact that the 1127th was based at Fort Belvoir, of course, does not prove that it was definitely linked with the Fort Belvoir smoke ring photos any more than Noel Scott's ionized air experiments at Fort Belvoir can be definitely linked with UFO investigations by the 1127th. Many other activities were conducted out of Fort Belvoir besides the Air Force's 1127th. However, the possibilities and questions remain. The declassified 1127th document, generously shared by researcher William L. Moore in 1994, constitutes additional proof that Project Blue Book was not the only official group investigating UFOs. This fact is not surprising to us in the 1990s' but it would have been extremely enlightening to researchers in the 1960s. The 4602nd Army Intelligence Special Services (AISS) was originally charged with the responsibility of recovering "space vehicles" as well as the investigation of UFO reports, according to a 1961 document obtained by researcher Clifford E. Stone, a retired U.S. Army sergeant (Stone, 1995). It is unclear just when the responsibility for UFO investigation switched from the 4602nd to the 1127th or if both units possibly were involved in UFO investigations at the time of the ~ o r Belvoir photos. Documents obtained by Mct Donald in 1970 from Blue Book archives at Maxwell AFB in Alabama, designated the 4602nd as one of the investigative units for the August 13-14, 1956, Lakenheath-Bentwaters sightings in England, an overseas sighting in which McDonald was highly interested (McDonald, 1970). If McDonald had known about Fort Belvoir's 1127th, he would not have trusted the assurances of Fort Belvoir officials so readily and would have persisted in his efforts to document Colonel Jackson's, Jack Strong's, and Andrew Taylor's statements. Conclusions Our team's reanalysis of the set of four Heflin UFO photos lead us to draw the following conclusions: (a) The clouds in all four photos are consistent; (b) we have detected a wake, never before reported to our knowledge, impressively suggesting a real object moving through the atmosphere; (c) the mid-1 970s GSW analysis that showed a "string" was a stunning error, certainly not based on legitimate copies, and possibly hoaxed itself, by persons unknown, on the copies Spaulding and GSW used; (d) the notations on the back of the first three photos suggest they were part of a covert analysis; (e) James E. McDonald, erroneously led to believe that the fourth photo was taken at a different time and place from the first three Heflin photos, missed a golden opportunity 622 A. Druffel, R. M. Wood, & E. Kelson to identify the finest photo evidence available at the time; (f ) the trail of particulates detected behind the craft in Photo 4 clearly demonstrates that the smoke ring in that photo is apparently the same as the black band of particulates detected in Photos 1 and 3; and (g) all of the photographic evidence remains totally consistent with the statements of the photographer, Rex Heflin. Our reanalysis of the August 3, 1965, Heflin photos confirms that Heflin's account of the sighting is entirely consistent with his pictures and reconfirms that the witness/photographer was not involved in a hoax. This analysis represents a general study that specifically addressed the historical issues behind these photographs. An in-depth analysis is underway that will characterize the blur of the object and incorporate this information into determinations of size and distance. This analysis will be offered for a forthcoming issue of this journal. References Condon, E. U. (1969). Scientific study of unidentifiedflying objects. New York: Bantam Books. Druffel, A. (1997). James E. McDonald's UFO files. Mufon UFO Journal, No. 345,3-9. Edwards, F. (1 966). Flying saucers-Serious business. New York: Lyle Stuart. Encyclopedia Britannica, 1968 Book of the Year (Special Rep.). (1968). UFO's & the Colorado project, reevaluating the unidentified (p. 123). Epperson, I. (1966). Personal communication. Gray, J. (1968). Personal communication. Heflin, R. E. (1965). Personal communication. Jackson, Col. J. H. (1968). Personal communication. Lorenzen, J., & Lorenzen, C. (1966). APRO Znvestigator. Tucson, AZ: Aerial Phenomena Research organization. McDonald, J. E. (1 953). The earth's electricity. Scientific American, 188 (4). McDonald, J. E. (1967). Personal communication. McDonald, J. E. (1968). Handwritten notes on meeting of Los Angeles NICAP Subcommittee, in "Heflin" file. (McDonald's UFO research files have been recently archived and are available to the public at University of Arizona Library, Tucson.) McDonald, J. E. (1 968a). File "USSR: A. A. Tikhanov; Y. A. Fomin; Fedor Y. Zigel; Gen. Porfiri Stolyarov." McDonald, J. E. (1968b). "Ft. Belvoir" file. . McDonald, J. E. ( 1 9 6 8 ~ )Personal communication. McDonald, J. E. (1968d). Memo to Dr. William Hartmann. McDonald, J. E. (1968e). Personal communication. McDonald, J. E. (1970). UFOs over Lakenheath in 1956. FSR: Flying Saucer Review (London), 16(2), 9-1 7,29. Project Blue Book Files. (1965) Heflin, Rex E. Photo Case. Washington, DC: National Archives. Rankow, R. (1 967). Personal communication. Reichmuth, Capt. C. F., U.S. Air Force. (1965). In Young, M. (Ed.). UFO Top Secret. (138-141) New York: An Essandess Special Edition, 1967. Spaulding, W. (1 975). August-Summer news bulletin: Result of computerphoto analysis. Tucson, AZ: Ground Saucer Watch, 2. Spaulding, W. (1 977). Personal communication. Stone, C. E. (1995). Point Piece: Air Force lied about projects, researcher says. Sunland, California UFO Magazine, 10(4), 4,43. Strong, J. (1 967). Personal communication. Toledo Blade. (1952). They're homemade "flying saucers": Expert makes saucers in glass jar. U.S. Air Force. (1960). History HQ 1127th USAF Field Activities Group. (Declassified government document.) Journal of Scientific Exploration, Vol. 14, No. 4, pp. 623-641,2000 0892-3310100 O 2000 Society for Scientific Exploration Should You Take Aspirin to Prevent Heart Attack? Research Professor Chemistry University of the Sciences in Philadelphia 600 South 43rd Street Philadelphia, PA 19104 "Nothing is Simple."-Harry Rowe Mimno, Professor of Applied Physics, Harvard University Abstract- The majority of physicians in the USA recommend aspirin for prevention of first heart attacks to almost everyone over the age of 50, even though women have not been included in the clinical trials of aspirin. While aspirin does prevent about 113 of first heart attacks, its side-effects are so severe as to cause a higher death rate overall than placebo. Non-fatal side-effects, such as internal bleeding and cataracts, are significant after years of aspirin use. The major study on which most recommendations are based did not utilize aspirin alone; therefore, the calcium and magnesium present in the buffered aspirin actually taken may have been responsible for some of the beneficial effects. Supplemental magnesium and vitamin E have been shown to be more effective than aspirin in lowering heart attack rates as well as overall death rates. Aspirin does reduce the incident of second heart attacks by about 115 when taken for a few weeks. Supplemental magnesium and coenzyme Q10 have been shown to be more effective than aspirin in treatment of cardiovascular disease. Keywords: Aspirin-heart attack-cardiovascular disease-stroke-vitamin E-coenzyme Q 10-magnesium-potassium. Introduction Recent advice given in books and articles for general audiences is contradictory. Many practicing cardiologists and other physicians still do not understand the findings in clinical studies and believe that this was all settled 10 years ago. Flaws have been claimed to exist in some of the largest and formerly bestregarded studies, including the supposed lack of an exact specification of what was taken that was supposed to be aspirin. As a result, studies of the supposed benefits of aspirin in preventing heart attacks continued during the 1990s. A skeptical examination of the benefits of aspirin attributed to the peer-reviewed medical literature shows frequent misinterpretation or worse by writers for lay audiences. This article will show that careful examination of original peer-reviewed papers will allow you to draw conclusions about who could benefit by taking aspirin which are at odds with some strongly held opinions. (The mean- ~ i 624 Kauffman Fig. 1. Structural formula of aspirin (acety,lsalicylic acid or ASA). ing of medical terms may be found in ordinary or medical dictionaries, or www.netscope. drkoop. com/search/queryl asp.) What Is Aspirin? The structural formula for aspirin is shown in Figure 1. The most common chemical name for this organic compound is acetylsalicylic acid (ASA). There are at least 32 other names for it, mostly trade names (Windholz, 1976). It was first synthesized by Carl R. Gerhardt in 1853 (Mustard, 1982). The major therapeutic use of ASA in providing relief from the pain of rheumatoid arthritis was recognized by Felix Hoffman, an employee of Bayer AG, in 1897, who administered ASA to his father, who tolerated ASA much better than other salicylates already in use. ASA was not "invented" in 1897 as in the book The Aspirin Wars (hereafter Wars, Mann & Plummer, 1991) (Wars, p. 7 and cover). First trademarked in 1899 by Bayer AG (Vane & Botting, 1992), Leverkusen, Germany, the name AspirinTM became a generic term for ASA in the manner of Kleenex and Frigidaire. For most of a century aspirin has been the preferred treatment for arthritis pain and has been used for headache, fever, and, in the last decade, prevention of heart attacks. It has been called the most successful drug in history. A decade ago 1 in 5 Americans took aspirin every day (Wars, cover). Not until the 1970s was the mode of action of aspirin worked out! Sir John Vane was awarded the Nobel Prize for uncovering the mode of action of ASA (Feinman, 1993). ASA inhibits the enzyme cyclooxygenase, preventing the cells of the body from making certain prostaglandins that cause inflammation and other ones that cause the clumping of blood platelets to form clots. The clots, or thromboses, are responsible for "ischemic events," which are the local anemias, or blood shortages, caused by blockage of arteries. When these are coronary arteries, the blockages are called "heart attacks" of the myocardial infarction (MI) type. The common slogan "aspirin thins the blood" is not strictly true; aspirin prevents clot formation by platelets. The ASA content of a standard aspirin tablet is 325 mg. Extra-strength or Does Aspirin Prevent Heart Attack? 625 arthritis-strength tablets contain 500 mg. For other uses tablets containing 160 and 81 mg are available. Enteric-coated aspirin tablets resist the acidic environment of the stomach; the aspirin is absorbed in the alkaline small intestine. You would not expect "fast, fast, FAST relief of headache" with these, but some studies showed that stomach erosions and ulcers were less frequent (McDonald, 1982). "Buffered" aspirin is no faster than plain aspirin (Wars, p. 164) and only slightly less irritating, if at all (www.mayohealth.org). Besides containing 325 mg of ASA, a BufferinTM tablet has an actual alkali content of 158 mg of calcium carbonate, 63 mg of magnesium oxide, and 34 mg of magnesium carbonate; the latter pair provide a total of 48 mg of magnesium, which may be important for preventing heart attacks. Bayer Aspirin with Stomach Guard is the same. Primary vs. Secondary Prevention of Heart Attacks One must be skeptical about any recommendation for or against aspirin that does not distinguish between primary and secondary prevention. Primary means that people not at any particular risk of MI may prevent a fraction of potential MIS from occurring by taking small doses of aspirin for a long period. Any side effects of aspirin can be serious if people begin taking it at age 45-50 and continue for 30-40 years. Secondary prevention means that actual victims of MI or unstable angina, a high-risk group, may prevent a fraction of further cardiovascular problems by taking moderate doses of aspirin for a limited period. Any recommendation for or against aspirin that does not make the distinction can be disregarded as superficial. The Aspirin Wars distinguished between primary prevention of first heart attacks and secondary prevention on p. 11 quite well and described the U.S. Food and Drug Administration (FDA) decision to allow advertising for second heart attacks, but not for first heart attacks, due to an unusual number of strokes in the aspirin-using group in a large study on primary prevention, a prescient decision. But by p. 334 in Wars: Aspirin is the drug of doctors' dreams. It is hugely effective. One aspirin a day, or every other day, will save hundreds of thousands of lives a year. It can be taken safely by more people than almost any other drug ... It is likely to remain the only heart attack preventive sold in grocery stores for years to come. Surrogate End Points in Clinical Trials: Are We Being Misled? This is the title of an unusual paper by Fleming and DeMets (1996). Clinical trials are the standard scientific method for evaluating a new drug or a new use for an old drug. The true end point in most trials would be cure of a disease or condition, or, at least, reduction of symptoms, as indicated by longer lifespan of good quality. A surrogate end point is a laboratory measurement or a physi- 626 Kauffman cal sign used as a convenient substitute for a clinically meaningful end point that measures survival directly. Changes induced by a therapy on a surrogate end point are supposed to reflect changes in a clinically real end point, but all too often, they do not. Examples of surrogate end points are reduction of cholesterol level or blood pressure, two parameters easy to measure in the short term. A meta-analysis of 50 cholesterol-lowering interventions, including diet, resins, and lovastatin, showed that cholesterol levels were lowered an average of lo%, but there was a 1% increase in overall mortality. (This should not have been a surprise, since high blood cholesterol level and low LDL + HDLIHDL ratio are merely correlated with cardiovascular trouble, not an actual cause of it, as homocysteine is; McCully & McCully, 2000, and references to peer-reviewed papers therein). A meta-analysis of trials of calcium channel blockers, even though they really do lower blood pressure, showed possibly harmful effects overall. In addition, two antiarrythmia drugs approved by the FDA, encainide and flecainide, clearly suppressed arrythmias, probably as seen by electrocardiograms, as the surrogate end point. However, it was found that 3 times as many patients in the drug group died as in the placebo group. In evaluating aspirin, it is, therefore, not enough to show reduction in the rate of MI or other undesirable vascular events; one must determine total death rates for a reasonable period of several years in order to find whether some toxic effect of aspirin is countering a positive effect on MI. On the other hand, one does not want to carry on for too many years since the ultimate death rates of treatment and placebo groups converge-to 100%. Whisper Down the Alley This is one name for a grade-school game in which someone in a classroom whispers a phrase of a few words to the nearest student, who whispers the same phrase (supposedly) to the next student. The output of the 30th or so student is compared with the input and all have a good laugh, since the two are never equal. Adult scientists are not supposed to scramble the input-but some do. A massive meta-analysis of 25 completed clinical trials of secondary prevention of MI was reported in the British Medical Journal ( B M J ) in 1988 (Antiplatelet Trialists' Collaboration, 1988). The title: "Secondary Prevention of Vascular Disease by Prolonged Antiplatelet Treatment," makes clear that most of the patients involved had already suffered from MI, transient ischemic attack, unstable angina, or minor stroke. "Antiplatelet Treatment" indicates that aspirin was not the only drug tested; these facts are, of course, confirmed in the text and tables, of which one of the key tables is reproduced here as Figure 2. Note that only 12 of the trials employed aspirin alone. Overall reduction in mortality was about 25%, mostly in the first 2 years of treatment. A special note was made that men aged 55-74 with no history of vascular disease for whom aspirin treatment was actually primary showed no difference in mortality. Does Aspirin Prevent Heart Attack? Aspirfn + + dlpyrldarnofe I I UK - TlA Aspirln Asplrln dlpyrldamde, aspirin Aspirid, sulphlnpyrazone, both Asplrin Suloctldll I 3 s * I 1 AlCLA CCSG Sweden McMaster Toulouse AlTlA Toronto . - Asplrln Aspirin + dlpyrldamole, asplrin 2 ; ocs Stoke 5 .= b Sulphinpyrazone I Asplrln Olpyridamole Sulphlnpyrazone Aspirin Typical odds r e d u d o n 22% (SD 5%) Tennessee German TIA All cerebrovascular trials Aspirin Aspirin Aspirin Asplrin I AMlS PARIS - It ..- 2 2 PARIS - I Cardiff -!I ART + + dlpyridamole dipyrldarnole, aspirin Sulphlnpy razone Asplrin Aspirin Asplrin Sulphinpyrazone Aspirin 2 g COP - A GDR Cardiff- I ARlS GAMIS I A I All myocardial infarction trials Ia $2 7 I I Typlcal odds reductlon 25% (SO 4%) VA B a McMaster Aspirin Aspirin, sulphlnpyrazone. bom I I - 1 r, I I I I I All unstable angina trials All availaMe trials Typlcal odds reduction 36% (SO 13%) Typlcal odds reduction 25% P O 3%) 41 Odds ratio (active treatment : control) 1 Fig. 2. Meta-analysis of antiplatelet treatment from McCully & McCully (1999). Odds ratios (active treatment : control) for first stroke, myocardial infarction, or vascular death during scheduled treatment period in completed antiplatelet trials. Filled square: Trial results and 99% confidence intervals (area of square proportional to amount of information contributed). Open diamond: Overview results and 95% confidence intervals. Dashed vertical line represents odds ratio of 0.75 suggested by overview of all trial results. Solid vertical line represents odds ratio of unity (no treatment effect). This BMJ article was cited in Science, a publication of the American Association for the Advancement of Science, with reproduction of that same figure, as an excellent example of how to do a meta-analysis, along with an explanation of how to do one (Mann, 1990). The secondary nature of the trials was in- 628 Kauffman dicated only by the word "recurrence," and the end point was implied to be only "heart attack," while the legend (in Figure 2) includes as end points MI, stroke, and other vascular death. The Science article was cited by Dean Radin in the book The Conscious Universe as an example of the power of meta-analysis (Radin, 1997). Now Radin wrote implying that only aspirin was involved, and only for heart attacks, and the secondary nature of the treatment was not mentioned at all, which led me to believe, when I read this, that I should have continued using aspirin myself after all. Recommendations for You to Take Aspirin for Primary Prevention of Heart Attacks In publications for the general public there are a number of sources of advice to take aspirin for primary prevention of heart attacks. Here are a few: Consumer Reports (CR), with =5 million subscribers and =20 million readers, recommended that postmenopausal women, men over 35 with risk factors such as smoking cigarettes, and possibly men over 45 without risk factors all take aspirin. No dose level was given, although the study quoted was based on "one 'aspirin' tablet every other day," and the use of enteric coated aspirin was advised only if uncoated aspirin caused damage to the stomach. "The ... study found that one aspirin tablet every other day cut the rate of initial heart attacks almost in half ... The implications were stunning." But then CR was very cautious, noting that the clinical study they were citing showed significantly more hemorrhagic strokes (rupture of blood vessel in the brain), ulcers, and allergic reactions, and that no benefit was observed in another trial on healthy male physicians in the UK (Consumer Reports, 1988). While the studies used did not get proper citations, the first was certainly the Physicians Health Study in which 22,071 male physicians were studied for 5 years (Steering Committee of the Physicians Health Study Research Group, 1989; hereafter P H s 89.) Julian Whitaker, MD, in his popular newsletter Health & Healing, properly referenced PHs 89 and recommended that everyone take aspirin for primary protection from MI, but at the rate of 162 mg every other day, or 81 mg every day, half the dose used in PHs 89. While the study involved only male physicians, Whitaker did not restrict his recommendations to males. Whitaker wrote that the usual side effects of aspirin could be avoided by taking the low dose he recommended with a meal (Whitaker, 1996).' In the February, 2000 issue of Life Extension magazine, the recommendation for taking 81 mg of aspirin per day with food is unequivocal: Dr. Whitaker is one of the most courageous advocates of new or alternate treatments for many conditions. He has risked life, liberty, and financial ruin in trying to protect other practitioners from the FDA and in campaigning for easy availability of supplements. ' Does Aspirin Prevent Heart Attack? 629 A lot of people in alternative medicine criticize The Life Extension Foundation for recommending the daily use of low-dose aspirin, but The Foundation stands firm on the recommendations it made in 1983: most healthy people should take low-dose aspirin to specifically reduce their risk of heart attack. Aspirin may protect in ways that supplements do not. (Knorr, 2000) Of the 34 references cited at the end of the article in such a way that one cannot tell which one backs each aspect of the article, just nine are to peer-reviewed journals. PHs 89 is not cited, nor is any peer-reviewed paper that shows lower total mortality in low-risk subjects. The article is cleverly laid out with a large space taken up by artwork so that it ends on the top half of its last page. The bottom half of the page is an advertisement for Life Extension Foundation's brand of aspirin. Does this fact make you skeptical? Recommendations for You Not to Take Aspirin for Primary Prevention of Heart Attacks From an anonymous author on a website (www.internetwks.com/pauling/ lie/mag.html): We have been told that all the aspirin studies that 'prove' an aspirin a day keeps a heart attack away-were with buffered aspirin, i.e., with added magnesium. Our sources point out that it is unlikely that further studies using 'plain' aspirin will be undertaken because preliminary studies always show 'plain' aspirin does not show the same protective effect against heart attacks. So if you still believe what you read in the mass media, make sure that your daily aspirin is buffered! (Or much better yet, take a magnesium tablet instead!) From the editors of the newsletter What doctors don't tell you (McTaggart, 1995): Possibly the largest collaborative study ever performed in medicine, this meta-analysis (Antiplatelet Trialists' Collaboration, 1994) pooled the results of some 174 clinical trials from around the world, testing an aggregate of 110,000 patients ... The overview was designed to determine whether medium-dose aspirin (75 mg to 325 mg per day) . .. could prevent ... nonfatal heart attacks, strokes, or deaths in [mostly] high-risk patients .... The researchers reckoned that this sort of therapy reduced the risk of [premature] death [a solid endpoint] from one of these causes by one-sixth .... This isn't the case with low-risk patients; the study showed that among those taking aspirin as 'primary prevention', although heart attacks were reduced by a third, there was a 'non-significant' increase in nonfatal strokes. However, that increase was cited as 21% (hardly a 'non-significant' increase in our view) .... However, the study makes quite clear that for low-risk people or for those with so-called risk factors like high cholesterol, hypertension, or smoking, but without vascular disease, there is no evidence that this socalled preventive therapy does any good. In fact, the risks (particularly of hemorrhage or stroke) may outweigh the benefits. Therefore, there is no scientific justification for your doctor's view that you should start taking aspirin just in case. 630 Kauffman And from William Campbell Douglass, Jr., MD: I'm sure you've heard about the study [ P H s 891 showing that a n aspirin a day prevents heart attacks. I n that study, m e n who took a daily aspirin had 47% [sic] fewer heart attacks than men who didn't. W h a t you haven't heard, and what I ' m sure t h e aspirin companies don't want you to know, is that the subjects in that study took buffered aspirinaspirin mixed with magnesium. Numerous studies have proven that magnesium has a powerful protective effect o n your heart. It dilates blood vessels ... aids potassium absorption into your cells (preventing heartbeat irregularities) ... acts a s a natural blood thinner ... and keeps your blood cells from clumping together [the anti-platelet effect]; indeed autopsies of heart attack victims almost always find a magnesium deficiency! .... Not only that, but recent studies link aspirin t o macular degeneration- the # 1 cause of blindness in people over the age of 55! B u t t h e biggest strike against aspirin may come from the very study touting its heart benefits. If you read the study's fine print, you'll find that even though the group taking aspirin had 47% fewer heart attacks, there was n o difference in the death rates of t h e t w o groups. That means that death from other causes was 47% higher in the aspirin group! S o stop taking that daily aspirin! Stick to magnesium instead. (Douglass, 1988) Are these people crazy? Not entirely. Now we know enough to divide the original question that is the title of this article into two separate questions: on primary as distinct from secondary prevention of heart attacks. Let us go to the peer-reviewed literature to answer the first of the properly posed questions: TABLE 1 Actual Deaths in PHs 89 Trial From PHS 89. Aspirin group 81 Total cardiovascular deathsb Acute myocardial infarction (4 10) 10 Other ischemic heart disease (411-4 14) 24 Sudden death (798) 22 10 Stroke (430,43 1,434,436)' Other cardiovascular (402,42 1, 424,425,428,429,437,440, 441) 15 1 24d Total noncardiovascular deaths Total deaths with confirmed cause 205 Total deathse 217 Person-years of observation 54,894.6 Placebo group 83 28 25 12 7 11 133 216 227 54,864.2 Relative risk 0.96 0.3 1 0.97 1.96 1.44 1.38 0.93 0.95 0.96 - 95% Confidence interval P value "Numbers are code numbers of the International Classification of Diseases, ninth revision. All fatal cardiovascular events are included, regardless of previous nonfatal events. This category includes ischemic (three in the aspirin group and three in the placebo group), hemorrhagic (seven aspirin and two placebo), and unknown cause (zero aspirin and two placebo). This category includes one death due to gastrointestinal hemorrhage. Additional events that could not be confirmed because records were not available included 23 deaths (12 aspirin and 11 placebo), of which 11 were suspected to be cardiovascular (seven aspirin and four placebo) and 12 noncardiovascular (five aspirin and seven placebo). Does Aspirin Prevent Heart Attack? 631 Should You Take Aspirin to Prevent a First Heart Attack? The Antiplatelet Trialists' Collaboration (1988) reported there were some low-risk men aged 55-74 for whom aspirin treatment was actually primary. The paper concludes with the opinion that the absolute benefits in primary prevention of MI were uncertain because they might be outweighed by a small increase in cerebral or other serious hemorrhagic disease. "Thus only for patients with an appropriate history of vascular disease is there at present clear evidence that antiplatelet treatment reduces the overall incidence of fatal or disabling vascular disease." This opinion recognizes that the real end point is life extension, not merely minimizing MIS. Table 1 is reproduced from PHs 89. This massive study on 22,071 male physicians, half taking 325 mg of "aspirin" every other day, showed that total deaths in the aspirin group over the 5-year period of the study were 4% fewer total deaths than in the placebo group (P = .64); thus the difference was not considered significant. A big reduction in fatal MIS of 69% (P = .004) was countered by nearly equal increases in the totals for sudden death (P = .09), stroke, and other cardiovascular deaths. The reduction in MI was seen only in those aged 2 50. Using the end point of life extension, not MI, there was hardly any benefit from taking aspirin. With respect to nonfatal bleeding of several types, the aspirin group had a relative risk of 1.32 (P = .00001). Furthermore, 48 in the aspirin group and 28 in the placebo group required blood transfusions (P = .02, all 95% conf.). There really was a significant ( P < .00001) reduction in nonfatal MIS of 44%. But what did this mean in real benefit? It meant that in a 1-year period the chance of having a nonfatal MI was cut from 0.44% to 0.25%. There is no doubt that PHs 89 used Bufferin, not aspirin. Monthly calendar packs containing either Bufferin or placebo were provided by Bristol-Myers Products. Domenick Mellace of Bristol-Myers was acknowledged for his logistic support. Bristol-Myers contrived to have a 1.5 page advertisement placed just ahead of this paper in the journal, in which advertisement they were careful to advertise Bufferin only for secondary prevention as directed by the FDA. Is it possible that the reduction in MIS was due to the magnesium present in the Bufferin and not the ASA content? By 1994 the Antiplatelet Trialists' Collaboration published a meta-analysis that was now up to 100,000 patients, of whom 30,000 were in the low-risk category (Antiplatelet Trialists' Collaboration, 1994). The doses were 75-325 mg of ASA per day, but the exact source of the ASA was not given. "There was no clear evidence on the balance of risks and benefits of antiplatelet therapy in primary prevention in low-risk subjects." In fact a graph was shown with "% free from a vascular event," including fatal, as the ordinate, and "years to first vascular event" as the abscissa. For low risk subjects after 4 years the treated group had 0.4% fewer events, that is, 4 per 1,000. But this included all of the antiplatelet treatments, including 2 trials with drugs that were more effective than aspirin, so it is likely that aspirin was of no benefit in low-risk subjects. 632 Kauffman Randomized clinical trials testing aspirin in 5,011 elderly people, 58% of whom were women, mean age 72 years, followed for a mean of 4.2 years, showed that use of aspirin caused a fourfold increase in hemorrhagic stroke (P = .003) and a 1.6- to 1.8-fold increase in ischemic stroke (Kronmal et al., 1998). Based on the Nurses' Health Study involving 79,319 women aged 34-59 years at the beginning, the role of aspirin in primary prevention of stroke was uncertain (Iso et al., 1999). This was based on a questionnaire, so the reduction, mostly in older women, of large-artery occlusive infarction by half (1 to 6 aspirin per week) or a doubling of the risk of hemorrhage (15 or more aspirin per week) might have included the use of a large fraction of buffered aspirin. This was not thought important. Total death rates were not included. "No conventionally used prophylactic aspirin regimen seems free of the risk of peptic ulcer complications ... Alka-Seltzer may be associated with higher risk (2x) and enteric-coated aspirin with lower risk ( 0 . 5 ~compared with plain ) aspirin" (Weil et al., 1995). Users of aspirin for long periods to relieve arthritis pain have suffered so badly from side effects that a multitude of alternates, such as ibuprofen and naproxen, were introduced. And, most recently, reported in 1998, a study of about 5,500 physicians in the UK on primary prevention of ischemic heart disease (which causes MIS) was carried out with 75 mg of aspirin daily in a controlled-release formulation for a median time of 7 years. The main effect of aspirin was a 32% reduction in nonfatal MI (less effective than PHs 89, which used double the dose), but there was an increase of 12% in fatal MI leading to an overall rise in death from all causes of 696, which was not considered significant (Meade, 1998). The absolute reduction in all MIS per year was 0.23%. Note that there is a 10% increase in overall death rate in the aspirin group in this study compared with the Bufferin group in PHs 89. Could this "nonsignificant" difference have been a lack of the beneficial effect of the magnesium in Bufferin? Another difference from P H s 89 is that the men in this study were recruited from the quintile considered to be at highest risk for MI based on heredity, smoking, blood pressure, and obesity; but this is still a lower risk group than the one composed of actual victims of MI. If delaying death is the real end point, not reduction in heart attacks per se, then it seems pointless to take aspirin for primary protection, with its certainty of obnoxious side effects, which may include gastritis, peptic ulcer, other internal bleeding, hemorrhagic stroke, fatal MI, and sudden death, to which has been added wet macular degeneration (in 1988) and 2-5 times the risk of cataracts in those people who were 2 55 and then took aspirin for 2 10 years (Cumming & Mitchell, 1998), in trade for a probable reduction of only 0.2% absolute per year in total (mostly nonfatal) MIS, especially when safer alternatives exist, such as magnesium. Now it is time to ask the more difficult question.. . Does Aspirin Prevent Heart Attack? 633 Should You Take Aspirin to Prevent a Second Heart Attack? Five earlier studies on secondary prevention of MI by ASA were reported from 1974 to 1980. There was said to be no beneficial effect overall (Gent & Carter, 1982). One multicenter study, nevertheless, the earliest of this type I have seen, had positive results. A single daily dose of 300 mg of aspirin in a gelatin capsule or a similar-looking placebo was to be taken before breakfast to ensure rapid absorption by 1,239 men who had had a recent MI. The aspirin group showed a reduction in total mortality of 12% at 6 months, 25% in 1 year, and 28% at 2 years. The authors modestly acknowledged that the results were statistically inconclusive, but they were in the range of what was observed in later trials. The much larger size of the later trials was needed to obtain results that would be statistically solid (Elwood et al., 1974). Reported in 1988, the second International Study of Infarct Survival (ISIS2) Collaborative Group in the UK determined the effect of aspirin vs. placebo in 17,187 people entering 417 hospitals after the onset of suspected acute MI. The aspirin used was clearly stated to be 162.5 mg in an enteric-coated tablet given daily for 1 month. All-cause mortality was said to be similar to vascular mortality. After 5 weeks aspirin produced 23% fewer vascular deaths overall (2P < .00001), cut MISfrom an absolute value of 2% to 1%, cut nonfatal stroke from 0.6% to 0.3%, and did not cause any increase in cerebral hemorrhages. Survival rates after 2 years were 81.7% in the aspirin group vs. 80.0 % on placebo (ISIS-2, 1988). In 1988 the Antiplatelet Trialists' Collaboration (1988) on 29,000 patients, a majority with a history of transient ischemic attack, stroke, unstable angina, or MI, were treated by a variety of methods, including with 2 300 mg ASA daily, which did not differ greatly in results from other drug regimens employed in the trials, as shown by this meta-analysis (see Figure 2). The authors thought that vascular mortality was reduced by 116, and nonfatal vascular undesirable events by 113 in high-risk patients. By 1994, now up to 70,000 highrisk patients, the Antiplatelet Trialists' Collaboration (1994) found similar results, but now the daily aspirin dose was 75-325 mg. By 1998 ISIS-2 was still following 6,213 high-risk patients in the UK of the 17,187 originally in the trial. During the first 35 days of follow-up, the use of aspirin during the first month reduced the death rate by 22%. Hence all of the survival benefit of an early, 1-month course of oral aspirin (1 62.5 mg enteric coated, daily) seemed to accrue during the first month, with little further benefit between day 36 and the end of year 10, by when the death rate was down 1% relative to placebo (Baigent et al., 1998). This then was the background of the North of England Aspirin Guideline Development Group's recommendations to physicians: Aspirin should be used in patients with acute MI at 150 mg daily for one month, then 75 mg daily for several years. In patients with MI, anginas, stroke, or transient ischemic attack, aspirin should be used at 75 mg daily for several years. There was no evidence that higher doses were more effective (Eccles, Fremantle, & Mason, 634 Kauffman 1998). The group did not mention either buffer or enteric coating. The latter seems desirable to this writer. So the answer for secondary protection from recurrence of several types of undesirable vascular conditions is Yes, take aspirin in low doses, and not forever, in order to obtain a moderate (16-22%) protection from fatal MI. Take the first aspirin tablet, not enteric-coated, in a hurry; have small sealed packs in your home and car. Switch then to enteric-coated tablets, and consider stopping in 4-6 weeks if you do not have further attacks. But is aspirin the best protection there is, either from the standpoint of effectiveness or freedom from side effects? Probably not. What Else Could You Take to Prevent Heart Attacks? Vitamin Efor Primary Protection The Nurses Health Study involved 87,245 female nurses aged 34-59 in 1980, who were free from diagnosed cardiovascular disease and cancer, and who completed dietary questionnaires every 2 years up to 8 years. Women who took vitamin E supplements containing, on average, 200 IU (International Units) for more than 2 years had 41% fewer instances of coronary disease of several types, and overall mortality 13% lower than those who did not ( P = .05) (Stampfer et al., 1993). The amount of vitamin E in multivitamin capsules at that time was typically I 3 0 IU. It could not be ascertained whether the vitamin E supplements were mixed isomers (d, l) or the l .36 times more active d form. The Health Professionals Follow-up Study involving =40,000 males aged 40-75 in 1986 who were free of diagnosed coronary heart disease, diabetes, or hypercholesteremia completed detailed dietary questionnaires every 2 years until 1990. Men who took 100-250 IU of vitamin E as supplements for 2-10 years had 37% fewer instances of coronary disease of several types, including fatal (P = .05). Higher doses of vitamin E were no more effective. By contrast, the intake of vitamin C and beta-carotene did not lower risk (Rimm et al., 1993). These results are far more impressive than the ones for aspirin, especially because side effects were so minimal as not to be mentioned. And vitamin E could be bought at almost any grocery store. Vitamin Efor Secondary Protection The Cambridge Heart Antioxidant Study [CHAOS (English humor?)] was a single-center, double-blind, placebo-controlled study with 2,002 patients who had angiographically proven coronary atherosclerosis (fatty deposits). Doses of 400 or 800 1U of natural vitamin E were used in half, and the group was followed for a median of 5 10 days. Vitamin E gave a significant reduction in nonfatal MI of 77% ( P = .005); however, there was a nonsignificant excess (1 8%, P = .61) of cardiovascular deaths in the combined vitamin E groups. However, Does Aspirin Prevent Heart Attack? 635 the lower dose of vitamin was better on both counts, including a 13% lower death rate on 400 IU than on placebo versus a 35% higher death rate on 800 IU. The lower dose of vitamin E gave 86% fewer nonfatal MIS and the higher dose 52% fewer. So here, too, vitamin E is far more effective than aspirin, and, again, side effects were negligible at the lower dose (Stephens et al., 1996). These latter data do not appear in the abstract, showing the bias of the authors, as did their measurements of total serum cholesterol, rather than of the more meaningful homocysteine. The recent GISSI-P trial used 300 mg of synthetic (d, I) vitamin E (GISSI-P Investigators, 1999). In the composite end point of death and nonfatal MI, vitamin E reduced risk by a barely significant 1 1 % (Brown, 1999). Since the subjects were already eating a "Mediterranean Diet" high in vitamin E, with olive oil, which aids in absorption of vitamin E (McCully & McCully, 1999, and references to peer-reviewed papers therein), perhaps this result is not surprising. Again, measurements of total serum cholesterol, HDL, LDL, and triglycerides showed no changes at all. Magnesium for Primary Protection The Caerphilly Heart Disease Study of men aged 45-59 years at the beginning of a 5-year period examined the relation of magnesium in the diet to the incidence of MIS, both fatal and nonfatal. Of the 627 men in the study, 38 suffered MIS. The mean daily intake of magnesium in these was 12% lower than in men who did not have MIS (Elwood et al., 1992). This is a difference of about 38 mg per day, less than the amount in a Bufferin tablet. The inverse relation of magnesium concentrations in drinking water to rate of heart attacks has been noted many times (Purvis & Mohaved, 1999). The usual recommendations for dietary supplements are to take 300-600 mg of magnesium in a compound (not the metal) daily with food (Miller, 1999). The most common form in which to take magnesium is as the compound magnesium oxide, one of the alkalis in Bufferin, but equally persuasive is advice to take it as potassium magnesium aspartate for fast absorption (Douglass, 1995). Women at risk of osteoporosis are advised to take also about twice the Inass of calcium (Lieberman & Bruning, 1990). But calcium, as well as vitamin D and phosphates, increase the amount of magnesium needed (Seelig, 1980). Mildred S. Seelig, MD, also wrote that the typical daily intake of magnesium in American college students was 250 mg, not 2 385 mg recommended for a 140-lb woman, or 2 500 mg for a 185-lb man. Unfortunately, I have not found a report on a large clinical study on primary protection using magnesium supplements in humans. A prospective study of 10-year duration in 400 "high-risk subjects .(selected about as in Weil et al., 1995), of whom 93.5% were male, living in Moradabad, India, was carried out by assigning half the group to a high-magnesium diet (1,142 mg per day vs. 4 18 mg in the control group from fruits, green vegetables, cereals, and nuts) and tracking medical events. The high-magnesium 636 Kauffman group had 35% fewer deaths from all causes ( P < .001) and a 61% reduction of nonfatal cardiovascular events (P c .001), including a 54% reduction in strokes (Singh, 1990). Unfortunately, this report was marred by a number of arithmetical errors in the table of results. There was also a confounding factor in that the high-magnesium diet was also a high calcium diet (880 vs. 512 mg daily) and a high-potassium diet (3,080 vs. 548 mg daily). Since serum levels of magnesium and potassium were raised and those of calcium were not, it is most likely that the magnesium and potassium were responsible for the differences in outcomes, which also included significant reductions in serum total cholesterol and glucose. Use of magnesium supplements in many people is probably justified by inference based on their effectiveness on secondary prevention, the clinical experience of a number of physicians, the drinking water studies, and the above diet study. The diet study would support using potassium magnesium aspartate as the supplement most resembling the high-magnesium diet. Magnesium for Secondary Protection In a double-blind, placebo-controlled study involving 273 patients with suspected acute MI, 74 received placebo, while 130 received 1.2 g of magnesium as the chloride intravenously during 24 hours, followed by 0.3 g in the next 24 hours. Treatments were begun within 3 hours of hospital admission. During the first 4 weeks after treatment mortality was 7% in the magnesium group and 19% in the placebo group, a reduction of 63% (P = .045). In the magnesium group 21% of the patients had arrhythmias that needed treatment vs. 47% in the placebo group, a reduction of 55% (P = .004). No adverse effects of intravenous magnesium were observed (Rasmussen et al., 1986). Reported in 1992, the second Leicester Intravenous Magnesium Intervention Trial (LIMIT-2) on 2,316 patients with suspected acute MI found a 24% reduction (P = .05) in 28-day mortality from treatment with intravenous magnesium sulfate (Woods et al., 1992). Reported in 1995, the Fourth International Study of Infarct Survival (ISIS-4) showed no benefit of similar treatment of 29,000 patients (ISIS-4 Collaborative Group, 1995). By 1996 the discrepancy was explained as follows: LIMIT-2 was doubleblind and placebo controlled, and only 30% of the patients had received treatment for thrombosis (streptokinase) by the time magnesium was begun, on average 3 hours after onset of symptoms. ISIS-4 was nonblinded; had no placebo, the alternate treatments being the drugs isosorbide mononitrate or captopril; 70% of the patients had received treatment (which raises blood magnesium concentrations) for thrombosis (clotting in major blood vessel); and 94% had received aspirin by the time magnesium was begun, on average 8 hours after onset of symptoms. It is interesting that captopril is a product of Bristol-Myers Squibb, the sponsor of ISIS-4, at a cost of about $10 million (B axter, Sumeray, & Walker, 1996). A study appeared simultaneously involving 194 patients considered unsuit- Does Aspirin Prevent Heart Attack? 637 able for treatment for thrombosis. In-hospital mortality was 4.2% in the magnesium group and 17.3% in controls, a reduction of 76% (Baxter, Sumeray, & Walker, 1996). Where confounding treatments are absent, rapid treatment of patients suffering from MI with intravenous magnesium is of great benefit in secondary prevention, not only of MI, but of arrhythmias. The three studies not confounded showed, on average, a greater 4-week benefit than from aspirin, and LIMIT-2 showed that concurrent aspirin did not change the outcome. Side effects of magnesium were minimal and could be avoided altogether by controlling the rate of administration. The medical establishment has accepted the role of oral magnesium supplements in countering hypertension, MI, congestive heart failure, and arrhythmias (Lauler, 1989). Coenzyme QlO for Primary Protection Coenzyme QlO is an oily organic compound, like vitamin E, and is found in every cell of the body. It has a number of functions, among which are preventing the oxidation of LDL and transporting oxygen from hemoglobin into the parts of cells where ATP, the main source of cellular energy, can be formed. Sharing its status with magnesium, the value of oral coenzyme Q10 supplements for better health in a low-risk population has not been investigated in large-scale controlled experiments (Overvad et al., 1999). Its use in older people with some definite symptoms, such as congestive heart failure, is probably justified by inference based on its effectiveness on secondary prevention, and the clinical experience of a number of physicians. For details, see the website of The International Coenzyme Q10 Association (wwwcsi.unian.it). Coenzyme QlO for Secondary Protection The New York Heart Association (NYHA) has grouped heart failure into four classes of severity. A cardiac patient in class IV, the most serious, is unable to perform any physical act without discomfort, and symptoms of heart failure, including anginal pain, may be present even at rest. Cardiologists know that such patients are on a relentless downhill course to death in spite of all conventional therapy. In a study in which all patients in hospitals were in NYHA classes I11 and IV and all received conventional therapy (bypass surgery, digitalis, diuretics, vasodilators), about 25% survived for 3 years. The patients treated with Coenzyme Q10 had a 75% survival rate. Putting this finding in the same form as used above, the reduction in 3-year death rate was 67%! (Folkers, 1985, 1986) Congestive heart failure is always characterized by an energy depletion status correlated with lowered coenzyme Q10 levels. In a 1-year double-blind trial, 641 patients of mean age 67 with chronic congestive heart failure (NYAS classes I11 and IV) were randomly assigned to receive either 2 mglkg (=I00 mg) daily of coenzyme Q10 or placebo. The number of patients who required 638 Kauffman hospitalization for worsening heart failure was 38% lower (P < .001) in the Q 10 group; the incidence of pulmonary edema was cut by 6 1%, and of cardiac asthma was cut by 51% (both P < .001) (Morisco, Trimarco, & Condorelli, 1993). A similar study on 2,500 patients showed only 0.5% with side effects thought due to Q10. Summing Up Not only the medical adviser to Consumers Union (CU), but also some health professionals who recommend aspirin, believe that there is a "highrisk" but not-yet-diagnosed population who should take aspirin for primary prevention of heart attack. You may check for yourself in the studies citedoften there is no such group. True, males 2 50 years old are at higher risk than males or females 1 50, but those males 2 50 are actually the low risk group in most of the large studies on health professionals. The study with the most favorable results in terms of reducing MIS, P H s 89, used Bufferin, which contains a significant amount of magnesium, not plain aspirin. This fact was lost on CU, as well as many others, including Abramowicz (2000), who cited P H s 89 and did not believe that the later European studies that utilized plain aspirin were valid. There is no consensus even among cardiologists that use of aspirin in the general population is advisable. For one, Prof. F. Verheught, Dept. Cardiology, University Hospital, Nijmegen, Netherlands, warned that use of aspirin for primary prevention was inadvisable because its use was investigated only in men, that the risk of nonfatal MI is <0.5% per year [and would be cut only by 0.2%], and that there was risk of gastric discomfort and bleeding (Kmietowicz, 1998). The studies on low-risk males were carried on for 5-7 years. Based on life-expectancies, advice to take aspirin beginning at age 50 would mean =30 years of exposure to its side effects. The story is the same for diabetic patients: in primary studies the hazards exceed the benefits. However, because of their higher platelet turnover, diabetics who take aspirin for secondary prevention may benefit from higher doses300 mg of enteric-coated aspirin daily (Yudkin, 1995). Vitamin E is both more effective and safer than aspirin, and its value in primary protection has been demonstrated in both men and women. Magnesium intake is inversely correlated with incidence of cardiovascular problems, the effect being more pronounced in men than in women. Up to at least 1,100 mg daily along with up to 3,000 mg of potassium is strongly protective in the primary sense and unarguably protective in the secondary sense (Patki, et al., 1990). In secondary protection, aspirin has a limited but definite value and does not have to be taken forever; most of the benefit is obtained in the first month. Based on available evidence, aspirin is preferred for the majority of stroke or myocardial infarction (heart attack) patients at risk of recurrences, according to The American Heart Association. But studies have shown that vitamin E, Does Aspirin Prevent Heart Attack? 639 magnesium, omega-3 fatty acids (Guallar, et al., 1999), and coenzyme Q10 each provide much greater benefits than aspirin with lesser side effects. Not even the skeptical website www.quackwatch.com disagrees with this. Lowering homocysteine levels may be of the greatest value in preventing cardiovascular disease (McCully & McCully, 1999, and references to peer-reviewed papers therein), and both vitamin E and magnesium have roles in lowering those levels or preventing oxidation of LDL cholesterol. A skeptical outlook is of great value in evaluating medical claims of most types. Medical advice with no citations to peer-reviewed papers on well-controlled studies can be ignored. You should spot-check the original papers, but beware of the Internet trap-you can get abstracts free, but it is more difficult or costly to obtain the full papers from websites. While all of the peer-reviewed papers in this field seem very honestly presented in detail, some important facts often do not appear in the abstracts, and some studies were contrived or abstracted to favor a preconceived result. Acknowledgements The following faculty at The University of the Sciences in Philadelphia provided help: Eric G. Boyce, Donna Gagnier, Daniel A Hussar, Sarah Spinler, Jeannette McVeigh, and William A. Reinsmith were of great help, but do not necessarily agree with the conclusions. Additional aid was provided by Charles J. Kelley, Tom Miller, and Mildred S. Seelig, MD. Disclaimer Any recommendations herein are based on studies published in peer-reviewed scientific journals. I am not an MD and cannot engage in the practice of medicine. (My degrees are: BS in Chemistry from the Philadelphia College of Pharmacy and Science, and a PhD in Organic Chemistry from the Massachusetts Institute of Technology. My experience includes about 10 years of exploratory drug development at the former and 4 years at the Massachusetts College of Pharmacy.) References Abramowicz, M., Ed. (2000) The Medical Letter, 42, 18. Antiplatelet Trialists' Collaboration. (1988). Secondary prevention of vascular disease by prolonged antiplatelet treatment. British Medical Journal, 296, 320-331. Antiplatelet Trialists' Collaboration. (1994). Collaborative overview of randomized trials of antiplatelet therapy-I: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. British Medical Journal, 308, 81-106. Baigent, C., Collins, R., Appleby, P., Parish, S., Sleight, P., & Peto, R. (1998). ISIS-2: 10-Year survival among patients with suspected acute myocardial infarction in randomised comparison of intravenous streptokinase, oral aspirin, both, or neither. British Medical Journal, 316, 1337-1343. Baxter, G. F., Sumeray, M. S., & Walker, J. M. (1996). Infarct size and magnesium: Insights into LIMIT-2 and ISIS-4 from experimental studies. Lancet, 348, 1424-1426. 640 Kauffman Brown, M. (1999). Do vitamin E and fish oil protect against ischaemic heart disease? (Commentary). Lancet, 354,441-442. Consumer Reports, October 1988, pp. 616-618. Cumming, R. G., & Mitchell, P. (1998). Medications and cataract: The Blue Mountains eye study. Opthalmology, 105, 1751-1 758. Douglass, W. C. (1995). A strong case for magnesium. Second Opinion, 5, 1-3. Douglass, W. C. (1998). Health breakthroughs. Second Opinion, Fall, 4. Eccles, M., Fremantle, N., Mason, J., & the North of England Aspirin Guideline Development Group (1 998). North of England evidence based guideline development project: Guideline on the use of aspirin as secondary prophylaxis for vascular disease in primary care. British Medical Journal, 316, 1303-1307. Elwood, P. C., Cochrane, A. L., Burr, M. L., Sweetman, P. M., Williams, G. H., & Welsby, E. (1974). A randomized controlled trial of acetylsalicylic acid in the secondary prevention of mortality from myocardial infarction. British Medical Journal, 1 436-440. Elwood, P. C., Fehily, A. M., Sweetman, P. M., & Yarnell, J. W. (1992). Dietary magnesium and prediction of heart disease. Lancet, 340, 483. Feinman, S. E. (Ed.) (1993). Beneficial and toxic effects of aspirin (p. 11). Boca Raton, FL: CRC Press. Fleming, T. R., & DeMets, D. L. (1996). Surrogate end points in clinical trials: Are we being misled? Annals of Internal Medicine, 125, 605-613. Folkers, K. (1986). Contemporary therapy with vitamin B6, vitamin B2, and coenzyme QlO. Chemical Engineering News, 21 April, 55-56. Folkers, K., Vadhanavikit, S., & Mortensen, S. A. (1985). Biochemical rationale and myocardial tissue data on the effective therapy of cardiomyopathy with coenzyme Q10. Proceedings of the National Academy of Sciences USA, 82, 901-904. Gent, M., & Carter, C. J. (1982). In Barnett, H. J. M., et al. (Eds.), Acetylsalicylic acid: New uses for an old drug (pp. 251-252). New York: Raven Press. GISSI-P Investigators (1999). Dietary supplementation with N-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: Results of the GISSI-Prevenzione trial. Lancet, 354, 447-455. Guallar, E., Aro, A., Jimenez, F. J., Martin-Moreno, J. M., Salminen, I., van't Veer, P., Kardinaal, A. F., Gomez-Aracena, J., Martin, B. C., Kohlmeier, L., Kark, J. D., Mazaev, V. P., Ringstad, J., Guillen, J., Riemersma, R. A., Huttunen, J. K., Thamm, M., Fok, F. J. (1999). Omega-3 fatty acids in adipose tissue and risk of myocardial infarction: The EURAMIC study. Arteriosclerosis, Thrombosis, and Vascular Biology, 19, 1 111- 1 118. ISIS-2 Collaborative Group. (1988). Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of suspected acute myocardial infarction. Lancet, 2, 349-360. ISIS-4 Collaborative Group. (1995). A randomised factorial trial assessing early oral captopril, oral mononitrate, and intravenous magnesium sulphate in 58,050 patients with suspected acute myocardial infarction. Lancet, 345, 669-682. Iso, H., Hennekens, C. H., Stampfer, M. J., Rexrode, K. M., Colditz, G. A., Speizer, F. E., Willett, W. C., & Manson, J. E. (1999). Prospective study of aspirin use and risk of stroke in women. Stroke, 30, 1764-1771. Kmietowicz, Z. (1998). Aspirin and warfarin best for primary prevention of heart attacks. British Medical Journal, 316, 327. Knorr, J. (2000). Aspirin, the multi-purpose compound, not just for headaches anymore. Life Extension, 2, 50-55. Kronmal, R. A., Hart, R. G., Manolio, T. A., Talbert, R. L., Beauchamp, N. J., & Newman, A. (1998). Aspirin use and incident stoke in the cardiovascular health study. Stroke, 29,887-894. Lauler, D. P. (1989). A symposium: Magnesium deficiency-Pathogenesis, prevalence, and strategies for repletion. American Journal of Cardiology, 63, 1-43G. Lieberman, S., & Bruning, N. (1990). The real vitamin & mineral book (pp. 141-142). Garden City Park, NY: Avery. Mann, C. (1990) Meta-Analysis in the Breech. Science, 249,476-479. Mann, C. C., & Plummer, M. L. (1991). The aspirin wars: Money, medicine, and 100 years of rampant competition. New York: Knopf (Wars). Does Aspirin Prevent Heart Attack? 64 1 McCully, K. & McCully, M. (1999). The Heart Revolution-The B vitamin breakthrough that lowers homocysteine, cuts your risk of heart disease, andprotects your health. New York: HarperCollins. McCully, K. & McCully, M. (2000). The Heart Revolution-The extraordinary discovery that finally laid the cholesterol myth to rest and put good food back on the table. New York: HarperCollins. McDonald, J. W. D. (1982). In Barnett, H. J. M. et al. (Eds.), Acetylsalicylic acid: New uses for an old drug (p. 89). New York: Raven Press. McTaggart, L. (Ed.) (1995) Medicine: What works and what doesn't (pp. 52-54). London: The Wallace Press. Meade T. W., with The Medical Research Council's General Practice Research Framework. (1998). Thrombosis prevention trial: Randomised trial of low-intensity oral anticoagulation with warfarin and low dose aspirin in the primary prevention of ischaemic heart disease in men at inreased risk. Lancet, 351, 233-241. Miller, T. (1999). www.execpc.com, last update July 23, 1999. Morisco, C., Trimarco, B., & Condorelli, M. (1993). Effect of coenzyme Q10 therapy in patients with congestive heart failure: A long-term multicenter randomized study. Clinical Investigations, 71, S134-136. Mustard, J. F. (1982). In Barnett, H. J. M., et al. (Eds.), Acetylsalicylic acid: New uses for an old drug (p. 1). New York: Raven Press. Overvad, K., Diamant, B., Holm, L., Holmer, G., Mortensen, S. A., & Stender, S. (1999). Coenzyme Q10 in health and disease. European Journal of Clinical Nutrition, 53,764-770. Patki, P. S., Singh, J., Gokhale, S. V., Bulakh, P. M., Shrotri, D. S., Patwandhari, B. (1990). Efficacy of potassium and magnesium in essential hypertension: A double-blind, placebo-controlled crossover study. British Medical Journal, 301,521-523. PHs: Steering Committee of the Physicians Health Study Research Group. (1989). Final report on the aspirin component of the ongoing physicians' health study. New England Journal of Medicine, 321, 129-135. Purvis, J. R., & Mohaved, A. (1999). Magnesium disorders and cardiovascular diseases. Clinical Cardiology, 5, 566-568. Radin, D. (1997). The conscious universe (pp. 55-56). San Francisco: Harper Edge. Rasmussen, H. S., Norregard, P., Lindeneg, O., McNair, P., Backer, V., & Balslev, S. (1986). Intravenous magnesium in acute myocardial infarction. Lancet, February, 234-235. Rimm, E. B., Stampfer, M. J., Ascherio, A., Giovannucci, E., Colditz, G. A., & Willett, W. C. (1993). Vitamin E consumption and the risk of coronary disease in men. New England Journal of Medicine, 328, 1450-1496. Seelig, M. S. (1980). Magnesium deficiency in the pathogenesis of disease (pp. 8, 10, 169). New York: Plenum. Singh, R. B. (1990). Effect of dietary magnesium supplementation in the prevention of coronary heart disease and sudden cardiac death. Magnesium & Trace Elements, 9, 143-151. Stampfer, M. J., Hennekens, C. H., Manson, J. E., Colditz, G. A., Rosner, B., & Willett, W. C. (1993). Vitamin E consumption and the risk of coronary disease in women. New England Journal of Medicine, 328, 1444-1449. Stephens, N. G., Parsons, A., Schofield, P. M., Kelly, F., Cheeseman, K., & Mitchinson, M. J. (1996). Randomised controlled trial of vitamin E in patients with coronary disease: Cambridge heart antioxidant study (CHAOS). Lancet, 347, 78 1-786. Vane, J. R., & Botting, R. M. (Eds.) (1992). Aspirin and other salicylates (p. 10). London: Chapman & Hall Medical. Weil, J., Colin-Jones, D., Langman, M., Lawson, D., Logan, R., Murphy, M., Rawlins, M., Vessey, M., & Wainwright, P. (1995). Prophylactic aspirin and risk of peptic ulcer bleeding. British Medical Journal, 310, 827-829. Whitaker, J. (1996). How aspirin could save your life. Health & Healing, 6, 1-3. Windholz, M. (Ed.) (1976). The Merck Index (9th ed., p. 114). Rahway, NJ: Merck. Woods, K. L., Fletcher, S., Roffe, C., & Haider, Y. (1992). Intravenous magnesium sulphate in suspected acute myocardial infarction: Results of the second Leicester intravenous magnesium interventional trial (LIMIT-2). Lancet, 339, 1553-1558. Yudkin, J. (1995). Editorial: Which diabetic patients should be taking aspirin? British Medical Journal, 311,641-642. Journal of Scientific Exploration, Vol. 14, No. 4, pp. 643-648,2000 0892-3310100 02000 Society for Scientific Exploration LETTERS TO THE EDITOR A Further Note on Walach and Schmidt's "Empirical Evidence for a Non-Classical Experimenter Effect" In Vol. 11, No. 3, of the Journal of Scientific Exploration two of us (J.M.H. and D.V.) criticized an article by Walach and Schmidt (Journal of Scientific Exploration, Vol. 11, No. 1) as they did not correct their statistical evaluation for possible dependencies caused by a "stacking effect" and for the increased alpha-risk caused by multiple analysis. In their rejoinder, Walach & Schmidt argued that the "Wilcoxon test for paired data" that was used by them was "applicable to dependent data" and therefore correct. After that, J.M.H. and D.V. consulted with U.T., who had long dealt with these problems. U.T. argued that the Wilcoxon test is not applicable and that the well-known method by Greville (1944) should be used, also for the correlation coefficient. Thus a reanalysis of the original data was called for and performed by U.T., who obtained the data from Walach and Schmidt. The necessary corrections for multiple analysis were incorporated in the new analysis. The result was that Walach and Schmidt's two best p-values were reduced from .O1 to .05 (experimenter 3) and from .002 to .O8 (correlation with paranormal belief), respectively. The new analysis has been published by U.T. in detail in the Zeitschrift fur Parapsychologie und Grenzgebiete der Psychologie (Timm, 199712000). Joop M. Houtkooper and Dieter Vaitl Center for Psychobiology and Behavioral Medicine Justus Liebig University of Giessen Otto-Behaghel-Strasse 10 0-35394 Giessen Germany Ulrich Timm Auwaldstrasse 98 0-79110 Freiburg im Breisgau Germany References Greville, T.N.E. (1944). On multiple matching with one variable deck. Annals of Mathematical Statistics, 15, 432-434. Houtkooper, J. M., & Vaitl, D. (1997). Comments on Walach and Schmidt's "Empirical Evidence for a Non-Classical Experimenter Effect." Letter to the editor. Journal of Scientific Exploration, 11, 395-396. 644 Letters to the Editor Timm, U. (1997). Zum methodischen Kontroverse um das Wiinschelrutenexperiment von Schmidt und Walach [On the methodological controversy about the dowsing experiment by Schmidt and Walach]. Zeitschrift fur Parapsychologie und Grenzgebiete der Psychologie, 39, 92-101. (This issue, designated 1997, actually appeared recently, in spring 2000.) Walach, H., & Schmidt, S. (1997). Empirical evidence for a non-classical experiment effect: An experimental, double-blind investigation of unconventional information transfer. Journal of Scientific Exploration, 11, 59-68. Reply to ''A Further Note on Walach and Schmidt's 'Empirical Evidence for a Non-Classical Experimenter Effect"' After the discussion of our paper in the Journal of Scientific Exploration, we agreed to the idea proposed by Ulrich Timm to do a reanalysis of the data for the control of a stacking effect. This new analysis is, as stated by J.M.H. and D.V., published in detail in the Zeitschrift fur Parapsychologie und Grenzgebiete der Psychologie. We agree with most of the issues it deals with but do find the way it is presented in this letter misleading. To clarify we would like to quote Ulrich Timm (1997) himself with some additional comments (translation into English by ourselves). With the abbreviation MCICD and CD, Timm refers to different designs in parapsychological experiments. CD stands for a design that does not need correction for stacking effect; MCICD, on the other hand, stands for a design that needs correction. 1. Experimenter Effect There is no obvious "stacking effect" and the MCICD evaluation leads in this case (italics by U.T.) more or less to the same results as the (nevertheless inadmissible) CD evaluation. Compared to the authors' Wilcoxon test the MC evaluation even scores clearly better. For the key result of overall performance of experimenter 3 one obtains now a one-tailedp-value of p = .0059 (was .01). (p. 96-97) 2. Correction for multiple analysis I agree with Houtkooper and Vaitl to multiply the p-value of experimenter 3 of p = .0059 with a factor for the correction of selection. I roughly estimate this factor by 3 x 3 - 1 = 8, because of post hoc [italics by Timm] selection of one of 3 experimenters and there would have been also the possibility of a selection on an equal base of one of the 3 experiments and the overall results. ...The correctedp-value is p = .047 and even doubles, if one thinks of a 2-tailed test to be appropriate. Our comment: The analysis was not post hoc. In the original paper (Journal of Scientific Exploration, Vol. 11, No. 1) we wrote ". .. however, subjects in- Letters to the Editor 645 structed by experimenter 3 had a hit rate significantly above chance @ = 0.04, Wilcoxon, two sided) for experiments 1 and 2 together. We subsequently checked whether this significance would be repeatable by testing an additional hypothesis introduced after experiment 2, stating there will be no differential effects of experimenters" (p. 63). Moreover, "Our third experiment can be seen as first positive experimental a priori test of the hypothesis of a non-classical experimenter effect" (p. 66). So for the last experiment, the analysis was not post hoc, although for the first two it was. We disagree furthermore with the correction factor. It is not sensible to take all possible selections (like the number of experiments we have conducted!) into account. One might multiply our result by factor 3 for three experimenters if one likes to be very straight or educative (we would not do that for the reason stated above-our introduced hypothesis focused only on experimenter 3), but all these corrections leave us with the same important fact: something was different for experimenter 3 because, with all correction factors applied so far in this long-lasting debate, the results remain significant. 3. Correlation Coefficient J.M.H., D.V. and U.T. state in their letter that "the well-known method by Greville (1944) should be used, also for the correlation coefficient." Timm wrote in his German reanalysis of our data: "It shows that the normal stacking effect is inoperative in this case. Instead one has to claim a differential stacking effect" (p. 97). Timm then develops a new model for this new differential stacking effect. So one can hardly speak of "well known method." This model includes a lot of preconditions that remain unproven, so we cannot decide whether it is appropriate for our data or not. With this model the correlation is corrected from an initial value of r = -0.29 to r' = -.22 or r" = -.265 (p.98). The p-value is corrected from initial p = .002 to p'= .0236. This correlation differs not much from our initial value and is for itself a significant finding. 4. Correction for multiple analysis Timm states: The new calculatedp-value for the correlation with paranormal belief can only be interpreted in an isolated way after multiplication with a correction factor of approximately 3.5 (see above) as this variable was selected post hoc [Timm's italics] out of a group of four (slightly correlated) prediction variables. (p. 99) Our comment: The correlation for paranormal belief is indeed a post hoc finding. As such it would be a useless procedure to correct thep-value for multiple testing. Explorative methods are used to find what's interesting in the 646 Letters to the Editor data; their aim is to generate future hypotheses not to proof or disproof facts within the data they are calculated on (Bender & Langer, 1999). As any reader of our paper can find out, we have presented this correlation only to find out whether experimenter 3 by chance had most of the subjects low in paranormal belief. If so the experimenter effect could have been due to this fact. Anyway, this was not the case. Timm himself states: Nevertheless this [the corrected p-value, S.S. & H.W.] undermines statistically the (similarly corrected) p'for the MC evaluation (.047 respectively .094), as a true correlation of any external variable with the hit rate is only possible if true hits are present at least for some of the participants. (p. 100) 5. Conclusion From the analyses presented by Timm and by us, one has to draw the same conclusion. Something unconventional happened in our experiments that needs explanation. After all, statistical analysis is an aid for decision, not a lead stick of the blind. Everyone looking at our data with open eyes can see the effect. Stefan Schmidt and Harald Walach Institute for Environmental Medicine and Hospital Hygiene University Hospital Freiburg Hugstetter Street 55 0-79106 Freiburg Germany References Bender, R. & Langer, S. (1999). Multiple test procedures other than Bonferroni's deserve wider use. British Medical Journal, 318, 600. Timm, U. (1997). Zur methodischen Kontroverse urn das Wiinschelrutenexperiment von Schmidt und Walach. Zeitschrift fur Parapsychologie und Grenzgebiete der Psychologie, 39, 92-101. Response to "Valentich Disappearance: New Evidence and a New Conclusion," by Richard E Haines and Paul Norman In "Valentich Disappearance: New Evidence and a New Conclusion," (JSE, Vol. 14, No. I), Richard F. Haines and Paul Norman report on new (alleged) witnesses and the disappearance of a young, relatively inexperienced Australian general aviation pilot on a night-flight on October 21, 1978, to King Island. When Valentich failed to report landing at King Island, Australian aviation Letters to the Editor 647 authorities immediately launched a week-long intensive search for the missing pilot, which was widely publicized in the Australian media. But "Ken Hansen" (pseudonym) and his two nieces, who now claim to have seen Valentich's aircraft (and a green 1ightlUFO) far away from his intended flight path, did not promptly report their sighting information to aviation authorities. According to the authors, none of the (alleged) witnesses ever reported the incident to Australian aviation authorities. For many years, according to the authors, the only person Hansen informed of the (alleged) incident was his wife, who reportedly "laughed at his story.. . ." Years later, he happened to discuss his sighting with a local policeman, who later mentioned the story to.. . the father of the missing pilot.. . [who] then told author [Norrnan], who interviewed Hansen and his two nieces. The published transcript of pilotlair-traffic controller radio communication suggests that the aircraft might have crashed into the ocean because of a malfunction of its single engine. But the authors offer no potential explanation for why none of the three witnesses ever reported what they (allegedly) saw to Australian aviation authorities. One possible explanation for their strange behavior: their tale is concocted. Philip J. Klass 404 N Street S W Washington, DC 20024 Authors' reply to Philip Klass We wish to thank Mr. Klass for reading our article ("Valentich Disappearance: New Evidence and a New Conclusion, JSE, Vol 14, No. 1,2000). He was correct in mentioning that Frederick Valentich disappeared while on a night flight to King Island on Saturday, October 21, 1978. However, we have several comments in reply to his suggestion that because the three eye witnesses didn't immediately report their sighting to aviation authorities they probably "concocted" their report. We would like to remind Mr. Klass of the following: Mr. Hansen (pseudonym) didn't know that an airplane and pilot had gone missing until some time later, and even then didn't interpret that what he and his nieces had seen that night was related to Valentich's disappearance. Perhaps we should have said this in our article. Mr. Hansen's reticence to make an immediate and official report about what he and his nieces had seen is not at all unusual. Why should he risk all of the bureaucratic paperwork, notoriety, and possible derision without having any firm proof of his sighting? Mr. Klass's comments only point to the practical difficulties that arise when someone would like to report an anomalous phenomenon but doesn't do so because of the stigma of 648 Letters to the Editor ridicule which skeptics and other have heaped up on all sides of the UFO issue. 2. One also should consider the issue of motive. Why would Mr. Hansen, who still wants no publicity whatsoever, make up this story with all of its internal consistency and "apparent" accuracy? Why wouldn't he just keep quiet? He impressed us with his mature, no-nonsense responses to all of our questions, which seems so typical of Australians, and also with his excellent knowledge of the appearance of aircraft in flight. If he did "concoct" this story, he was very convincing indeed. 3. Finally, it is interesting to us that, given the continued well-known attitude of negativity that Mr. Klass shows toward UFO phenomena, he only mentions in passing Mr. Hansen's statement that they saw a light aircraft in descending flight toward the ocean with a "large greenish circular light 'like it was riding on top of the airplane."' In fact the light was so strange looking he said he pulled off the road just to watch it. Although it was not the focus of our article, we hold firm to our concluding statement in our abstract that "The nature of the large object with green lights that accompanied the airplane during its steep descent remains to be identified." Interestingly, Mr. Klass adroitly avoided this detail, preferring to label the witnesses as possible liars. Richard I? Haines and Paul Norman 325 Langton Avenue Los Altos. CA 94022 Journal of Scientific Exploration, Vol. 14, No. 4, pp. 649-664,2000 0892-3310100 O 2000 Society for Scientific Exploration BOOK REVIEWS Mysterious Flame: Conscious Minds in a Material World by Colin McGinn. New York: Basic Books, 1999.224 pp. $24. ISBN: 0-465-014-224. The Mysterious Flame is Colin McGinn's 1999 addition to the plethora of books addressing the relationship between mind and matter, and the nature of consciousness. Among other features, what makes this book interesting is that it is not a defense of a particular model or theory of consciousness. The approach is at a more general level and asks whether we can ever understand the relationship of our inner mental lives and the objective, third-person world that science describes. To that end, McGinn is "concerned.. .with the geometry of human ignorance, with providing a map of human knowledge and its limits" (p. xii). Most authors who tackle theories of mind begin by pointing out what a daunting task this is. "What is it about consciousness that makes it so elusive to theoretical understanding?" (p. 6). McGinn spends the first section of the book pointing out the challenges that face any account of consciousness. His discussion is lucid and powerful, using intuition pumps about "conscious meat" and the lack of uniformity of nature that belies interaction between entities as disparate as mind and matter. The problem, as he rightly points out (p. 61), is that we do not even have an idea of what a possible solution to the problem might be like. A large part of the book is spent discussing various contenders for models of consciousness in the physical world and pointing out their problems. The main thrust of this book is "Mysterianism." McGinn is of the opinion that we are simply not up to the challenge of understanding consciousness. Not just now, but in principle, "our intelligence is wrongly designed for understanding consciousness" (p. xi). In chapter 2, he develops the idea that our intelligence is an evolutionary offshoot of other cognitive processes and thus will have important limits in explanatory power. To bolster the case for our never understanding consciousness, McGinn spends most of chapter 5 deconstructing our naive, pretheoretic ideas of the self. This pessimistic opinion runs counter to the dominant rationalist paradigm. The scientific revolution and most science since then has been based on the assumption that given sufficient man-hours, energy, and computational power, humanity will eventually understand every feature of the universe. Interestingly, McGinn focuses fairly narrowly on the possibility of understanding consciousness. In fact, limits on understanding and the nature of scientific impossibility is a topic addressed by the philosophy of science, and some of those issues would surely be relevant here. Limits set by quantum theory, relativity, basic logic, and theory of computation are related and are fascinating topics for 650 Book Reviews discussion. Interested readers can see Chaitin (1987), Davis (1986), and Penrose (1996). All of these issues are relevant because they bear on the plausibility of natural ceilings to our understanding. Some philosophers have taken the "unreasonable effectiveness of mathematics in describing the natural world" (Wigner, 1979) to be evidence for the contrary view. If the universe is indeed constructed according to rational principles, and if our minds are complex enough to understand the basic principles, then given enough time and computation, perhaps we can indeed understand everything there is. In the final analysis, the "Mysterian" viewpoint may be correct. There is no a priori way to rule it out. Our success with understanding features of the natural world may lead one to infer by induction that the problem of consciousness will likewise fall eventually, but perhaps the problems with which we have done well are precisely the ones that are tractable for us. Sooner or later we might come to those we simply cannot understand, due to biases and limitations built into our cognitive apparatus by evolution. Perhaps the longevity of the problem (thousands of years!), and the lack of meaningful progress on it, points to precisely this conclusion: "Philosophy marks the limit of human theoretical intelligence" (p. 209). It should be noted that McGinn isn't arguing that consciousness is necessarily in principle mysterious: While we cannot understand it, given our evolutionary history, while a creature with a different makeup might grasp consciousness perfectly but not be able to understand something that is obvious to us (such as clockwork, p. 66). McGinn believes that a specific "cognitive faculty" is necessary for understanding consciousness, and we might not have it (p. 52). This radical claim might be true, but I don't see any reason to believe it. A major problem with Mysterianism, as I see it, is that it is methodologically pernicious-a research stopper. If we assume that we cannot possibly understand consciousness, we run the risk of making a self-fulfilling prophecy. McGinn sees these objections and lists 10 reasons to adopt his viewpoint, to make it "more palatable, even liberating" (p. 69). The first is so that we do not waste time on hopeless pursuits and not use methods from other domains that don't work here. I find this unconvincing to the extent that the history of science is unequivocal in the verdict against those who say "X cannot be understood," but it is reasonable in that it keeps cognitive scientists honest about what their work is and is not accomplishing. The second reason is that facing the limitations can remove religious mysticism about consciousness because the limits are seen as due to our evolutionary past. This is quite reasonable, and it is important to know that mysteries are not necessarily a sign of magic. The third reason McGinn gives is that complete knowledge of ourselves might not be a good thing. I find this completely unconvincing. Aside from antitechnology arguments, it is unreasonable to avoid knowledge. The fourth reason is a sociological one: A mystery will allow us to keep our "sense of awe about the universe" (p. 71). I also find this unconvincing. This sort of argument has been used by various fundamentalists and antiprogress factions to Book Reviews 651 I 1 stop progress in many areas. Our sense of awe is better found in the profound beauty of the laws governing the universe and in our understanding of them. The fifth reason concerns the curbing of the hubris of science. This is reasonable in that it forces us to confront the possibility that other forms of knowledge about the universe exist, aside from objective third-person science (e.g., meditative experiential traditions). The sixth reason McGinn gives concerns trade-offs: He postulates that the changes that would be necessary to the human mind to enable it to understand consciousness would make us other than human and perhaps involve the loss of ability to understand other things. I see no reason to believe that this is the case, as we simply have no evidence that understanding consciousness is incompatible with any other faculty of our mentality. One might have argued that the ability to understand the relativity theory was incompatible with literature. Of course, the problem of consciousness is fundamentally different from problems in physics, but this reason is simply a restatement of the original difficulty (i.e., "we have no idea what a theory of consciousness would look like"). Reason number seven points out that accepting Mysterianism does not preclude us from doing useful phenomenology and neurophysiology. This is true, as far as it goes. Reason number eight basically urges us not to waste time on hopeless projects. This is a good point, but given the importance of the problem of consciousness, what is the strategic costlbenefit of ignoring it on the off chance that it is unsolvable? Reason number nine reasserts that this cognitive blind spot is a consequence of talents in other areas, so we should just accept the trade-off. This is perhaps true, but again, I see no reason to think so. The last reason points out that this viewpoint opens a whole new area of research: our cognitive strengths and talents. This is true, but as it stands now, we already have such a science: cognitive psychology. As for a study of fundamental limits on what is in principle unknowable, one might perversely argue to McGinn (in a Godel-like move) that this pursuit is hopeless because we don't even have an idea of what such a theory would look like. I would argue that things we could never understand would be beyond our even posing questions about them, thus foiling attempts to study such limits. In discussing previously proposed contenders for a theory of consciousness, McGinn goes over several popular views. One of these is dualism (chapter 3), the view that the mind is a separate, nonphysical entity and interacts with the physical body and brain. Toward the goal of showing how no available approaches treat the problem of consciousness appropriately, McGinn adduces a number of arguments against dualism. This section is clearly written for the layman, since most working cognitive scientists dismiss dualism out of hand anyway. Personally, I think that dualism will, in the end, be the only framework within which consciousness can be understood (Polten, 1973; Robinson, 1993; Warner, 1986). Most of McGinn's arguments against dualism can be circumvented by more sophisticated theories (Beloff, 1989; Dilley, 1988; Foster, 1991; Smythies, 1994; Smythies and Beloff, 1989). For example, it is true that 652 Book Reviews naive dualism can underestimate the dependence of consciousness on brain state (p. 25). Given how certain specific modulations of the brain change consciousness, it is untenable to think of the mind as a pure Platonic substance only transmitted by the brain. A workable dualist theory has to describe consciousness as the product of a tightly coupled, bidirectional mind-brain system. Certain medical cases (Lorber, 1981) underlie the danger of the opposite extreme: assuming that we understand exactly what is necessary for a body to exhibit consciousness. Another argument McGinn brings up is the old (and significant) problem of causality and energetic closure of the physical world. What variety of the laws of thermodynamics would allow the mind, as a nonphysical system, to interact with the physical world? The reason for the omission may have been in the interest of space, but in fact this issue has been discussed thoroughly in the past and shown not to be a fatal shortcoming for a sophisticated dualist theory (Burns, 1999; Larmer, 1986; Mohrhoff, 1999). The most significant problem McGinn brings out is the relationship of mental states to immaterial substance; that is, given "ghostly mind," how does it have indexicality, intentionality, qualia, and all the other aspects of consciousness that are so problematic for physicalist theories of mind? This is a crucial point and points to work that must be done by proponents of this theoryMcGinn also discusses panpsychism, a nonmonadic version of dualism. He asks the appropriate question: If all matter has some degree of sentience, why are nervous systems necessary for organisms to think and feel at higher levels (p. 97)? Of course, one could argue that nervous systems allow a better degree of coherence and information flow between biophysical systems at different scales of organization, and thus allow the development of more sophisticated mentalities than embodiment in "simple" kinds of matter. Panpsychism is a viable theory; interested readers are invited to look at a most curious and interesting book by Jean Charon (1984). A couple minor points in the book are interesting and bear comment. For example, it seems that McGinn believes that sentience is a discrete property: He likes "to imagine that day, many millions of years ago, when the first sentient organism came along, and a brand new ingredient entered the cosmos" (p. 63). This is possible but needs argument. It is at least as likely that consciousness, if it is even quantifiable, is a gradual, not an all-or-none, property. Another cognitive gem is the idea (p. 86) that Paley's famous design argument fails because although it explains the origins of order in nature, it begs the question of where the initial order (that which is surely necessary for a powerful deity to have) came from. Of course, this problem is commonly discussed in the context of entropy in cosmology: If the universe is winding down according to the laws of thermodynamics, how was the initial ordered state achieved? That is, who wound it up? By analogy with various cosmological Book Reviews 653 arguments, Paley's reasoning may not be as hopeless as McGinn makes it seem. In arguing that we lack the facilities to understand consciousness, McGinn says that "consciousness is necessarily not perceptible. We cannot even in principle conceive of a type of sense organ that would enable us to perceive consciousness." It should be pointed out that we likewise cannot conceive of a type of sense organ that would enable us to perceive entropy, but our understanding of entropy is proceeding well. More importantly, while McGinn's statement is true for a physicalist theory, it is not at all necessary for a dualist one. Most esoteric traditions hold that a mind can directly perceive the consciousness of another mind (Besant, 1994; Steiner, 1993). In a surprising move (given that this book is supposed to be about why models of consciousness are doomed), McGinn attempts (pp. 118-1 23) to resolve the tension between the nonspatial mind and the spatial brain. He speculates that consciousness is intrinsically linked with nonspatial properties that spacetime has (connected with events at the singularity before the big bang). Of course, he heads off possible models of exactly what about space-time makes consciousness possible by "sounding [a] customary note of pessimism" (p. 130). In fact, by saying that the universe must "have two compartments-one for matter, the other for mind-but these two compartments have to be joined together some way" (p. 128), McGinn seems to have come full circle to a dualist conception of consciousness. And at the end of the book, we see "mind and brain form an indissoluble unity at the level of objective reality," which slants powerfully toward panpsychism. In conclusion, this is a thought-provoking book, readable by the nonexpert in cognitive science or philosophy of mind. It points out the difficulties facing theories of consciousness embedded in a physical world. In the end, I find McGinn's reasons for adopting Mysterianism unconvincing and his acceptance of Mysterianism a potential brake to further progress. My take on the job of philosophers of mind and cognitive scientists is that they should follow the strategy of all scientists: Assume that the answer exists, and do their best to find it. The train of thought expounded in this book best serves the dual purpose of keeping such theorists honest and of pointing out that alternative approaches might be needed. Its most profound contribution is the idea that certain features of the world are cognitively closed to our intellects. This topic deserves much more development, but its application to consciousness is a fascinating topic, admirably dealt with in the Mysterious Flame. Michael Levin Department of Cell Biology Haward Medical School 240 Longwood Avenue, Boston, MA 021 15 levin @ whiz.med. haward.edu 654 Book Reviews References Beloff, J. (1989). The case for dualism. Chalottesville, VA: University Press of Virginia. Besant, A. (1994). A study in consciousness. Adyar, Madras: Theosophical Publishing House. Burns, J. (1999). Volition and physical laws. Journal of Consciousness Studies, 6, 27-47. Chaitin, G. (1987). Information, randomness & incompleteness. Singapore: World Scientific. Charon, J. (1984). The unknown spirit. London: Newcastle Publishing. Davis, P., (1986). No way: Essays on the nature of the impossible. New York: W. H. Freeman & Company. Dilley, F. (1988). Mind-brain interaction and PSI. Southern Journal of Philosophy, XXVZ, 469-480. Foster, J. (1991). The immaterial self. London: Routledge. Larmer, R. (1986). Mind-body interaction and the conservation of energy. International Philosophical Quarterly, 26, 277-285. Lorber, J. (1981). Is your brain really necessary? Nursing Mirror, 152, 29-30. Mohrhoff, U. (1999). The physics of interactionism. Journal of Consciousness Studies, 6, 165-184. Penrose, R. (1996). Shadows of the mind: A search for the missing science of consciousness. Oxford, UK: Oxford University Press. Polten, E. (1973). Critique of the psycho-physical identity theory. The Hague: Mouton. Robinson, H. (1993). Objections to physicalism. New York: Clarendon Press. Smythies, J. (1994). The walls of Pluto's cave: The science &philosophy of consciousness. Aldershot, UK: Avebury. Smythies, J., & Beloff, J. (1989). The case for dualism. Charlottesville, VA: University Press of Virginia. Steiner, R. (1993). Outline of occult science. Spring Valley, VA: Anthroposophic Press. Warner, R. (1986). A challenge to physicalism. Australasian Journal of Philosophy, 64, 249-265. Wigner, E. (1979). Symmetries & rejZections. Cambridge, UK: Ox Bow Press. Dancing Naked in the Mind Field by Kary Mullis. New York: Vintage Books, 2000 (first published 1998, Pantheon). 222 pp. $13 (paperback). ISBN 0-679-774-009. Where you agree with Kary Mullis, you find him delightful. Where you disagree, he seems naive and simpleminded. In either case, it will be delightfully plain what he means (or thinks he means). About how science works, about bureaucracy, about the corruption of Big Science, about the lack of evidence that HIV causes AIDS, I found Mullis delightful. About his personal experiences of telepathy and astral projection and about the validity of astrology, I found him naive. Mullis tells it like it is. That is to say, pulling no punches he tells it as it seems to him. On AIDS, for example, he says, A segment of our society was experimenting with a life style and it didn't work. They got sick. Another segment of our pluralistic society, call them doctor/scientist refugees from the failed War on Cancer, or just call them professional jackals, discovered that it did work. It worked for them. They are still making payments on their new BMWs out of your pocket. (p. 182) The book is accurately described on its cover as "Memoir/Science." It is autobiographical but not an autobiography, selections from personal experience, Book Reviews 655 but not a full account. Thus, Mullis's mother is mentioned a number of times, respectfully, while his father gets no mention. Chapter 1, "The Invention of PCR," recounts the discovery that was worth a Nobel Prize and that is now instantiated in virtually all work on DNA. In much of the book, what Mullis says about his sex life might better have been left out; but here in chapter 1, it serves a purpose: The Big Idea evidently drove everything else out of his mind, which is an authentic description of creative scientific discovery. Sex becomes a preoccupation only when there's nothing better to do. Authentic, too, in chapter 1 are the periodic doubts that so simple an idea could be new and original; the lack of imagination of the colleagues who are told of the idea; and that Mullis's employer did not adequately share with him the profits from commercial exploitation of PCR. (They sold the rights for $300 million and gave him $10,000, according to Greenberg [1998].) Chapter 2 is about getting the Nobel Prize and other awards. In rather bad taste, Mullis reports his private words with the Empress of Japan, with Hillary Clinton, with the Swedish Royal Family. The genre is reminiscent of "Lucky Jim" Watson's The Double Helix and Richard Feynman's Surely You're Joking, Mr. Feynman-what my mentor called the "watch-me-mama" syndrome, "look how clever I am." It illustrates Mullis's self-described (Lang, 1998, p. 628) tendency to behave like a jerk. Becoming a scientist is described in chapter 3. Those of Mullis's generation and earlier will sigh with nostalgia here as they relive those innocent days when they could experiment in ways that nowadays are deemed far too dangerous. And all working scientists will empathize with his account of the bureaucratic idiocies of the likes of "safety officers" at laboratories supposedly for grown-ups. In a somewhat similar vein, chapter 6 is sheer fun, a quite believable account involving a novel application of biofeedback. Is there anything new to be said about the 0. J. Simpson case? Yes, by Mullis in chapter 4. And chapter 5 gives food for thought in several further directions. The chief point is one I agree with and have not seen made elsewhere, at least not in so plainly understandable a fashion: that humankind should concentrate primarily on coping with human-scale problems. In physics and space science, for example, we should put 90% of our effort into looking for and neutralizing asteroids that might hit Earth, leaving the other 10% to "the trivial [problems] ...that involve our insatiable need to think that we can understand the fundamental nature of the universe" (p. 74). But chapter 5 also illustrates Mullis's penchant for oversimplifying. He asserts that the wavelengths of light that we can see are determined by the fact that "we developed our vision while we were living under the canopies of giant trees that let in only green light" (p. 65). A nice speculation, but it may not be so, in at least two ways: Our vision may not have developed while we were so living, and more generally, evolution doesn't work so predictably. Again, "How do I know when someone is standing behind me?" (p. 67) asserts what may not be a fact. In chapter 7, Mullis describes how he and a friend 656 Book Reviews "were being scientists" as they tested their ability to communicate telepathically; but parapsychologists and stage performers could offer explanations other than telepathy for the described events. Chapter 8 goes even further out: Mullis states as fact that his life was saved in 1974 by someone traveling on the astral plane, who-four years before they had met in the flesh-physically removed something from his mouth. Chapter 13 tells of an abduction experience that made Whitley Streiber's Communion resonate not only for Mullis but for his daughter as well, who had a similar experience quite independently. And in chapter 15, Mullis appears to subscribe to the validity of quite a simpleminded version of sun-sign astrology. Chapters 10 and 11, on the other hand, are trenchant and accurate critiques of contemporary science. While Enlightenment science was worthy of respect, modern Big Science has become commercial and corrupt, replete with professional administrators and other parasites. One recognizes the revamped versions of Science and Nature magazines quite readily from Mullis's description, even though he names no names: slick glossy paper with pictures on the front and lots of advertisements, a lot of editorial material by people who are professional journalists, and a few pictures of girls selling you things you might want to buy for your lab. The advertisers are the companies who make things for scientists to buy and the companies who make drugs for doctors to sell. (P. 172) Chapter 16 should be required reading for everyone, cold common sense about nutrition fads; and chapter 22 is only slightly extrapolated common sense about global warming. Chapters 18 and 19 detail why Mullis does not believe that HIV causes AIDS: because for more than 15 years, he has asked everyone including Gallo and Montagnier, the discoverers of HIV, for references to the published articles that demonstrate the connection and has never been given one: This would be the last time I would ask my question without showing anger. In response, Dr. Montagnier suggested, "Why don't you reference the CDC report?" "I read it," I said. "That doesn't really address the issue of whether or not HIV is the probable cause of AIDS, does it?" He agreed with me. It was damned irritating. If Montagnier didn't know the answer, who the hell did? (p.174) As to some of the major players, "In spite of his lack of luster as a scientist, Gallo had worked his way up in the power structure. Peter Duesberg, despite his brilliance, worked his way down" (p. 176). We're reminded (p. 177) that Gallo, at the 1984 press conference announcing the discovery of HIV, predicted a vaccine and cure within a couple of years. "Years from now, people looking back at us will find our acceptance of the HIV theory of AIDS as silly as we find the leaders who excommunicated Book Reviews 657 Galileo" (p. 180). "HIV didn't suddenly pop out of the rain forest or Haiti. It just popped into Bob Gallo's hands at a time when he needed a new career" (p. 178). This book, then, has something to please everyone and something to aggravate everyone. In my view, Mullis suffers from an excess of sanity and an inability to keep his views to himself or to mask them in polite banalities. Carl Djerassi's (1998) review was scathing, but others found some things to like and some to dislike. And of course different reviewers liked different things. While Greenberg (1998) and I think Mullis is accurate about the state of Big Science, for instance, to Kirkus Reviews (1998), "what really irks is when Mullis pontificates on.. .science." Read the book and decide for yourself. Henry H. Bauer Professor Emeritus of Chemistry & Science Studies Virginia Polytechnic Institute & State University hhbauer@ vt.edu References Djerassi, C. (1998, November 21). Surely you're joking. New Scientist, p. 51. Greenberg, D. S. (1998, November 5). Nature, 396, 38-39. Kirkus Reviews (1998, June 15). Lang, S. (1998). Challenges. New York: Springer-Verlag. Consciousness by D. Rakovic and D. Koruga (Eds.). Belgrade, Terezije: The Center for Peace and Development, United Nations University for Peace, 1996. $20. ISBN 86-7236-005-2. This book is unusual in its breadth and variety of coverage, looking at Consciousness from many points of view, but it comes from a place that does not have the most positive image in the view of American readers: Belgrade, Yugoslavia. To those who know the city, the fact that all the book's authors come from research institutes or universities is no surprise. Since World War 11, Belgrade has become the center in southeastern Europe for intellectual activities. The present book consists of four parts. The first, "The Phenomenology of Consciousness," deals with the relationship of disciplines such as philosophy and psychology with Consciousness. Part two goes straight to the heart of things, covering "altered states of Consciousness," such as "psychotherapeutic rituals in Amazon tribal societies." Part three details physiological and instrumental studies such as electroencephalography (EEG). An account is given of the EEG study of the effects of drugs, of sleep disorders, and the methods of EEG analysis. Part four, titled "The Search for the New Paradigm," has a varied content and approaches Consciousness in terms of "information physics," 658 Book Reviews "the brain and its physiology," and finally "a biophysical model for Consciousness." I found the quality of the chapters variable, with some simple articles on brain physiology, with a slant toward Consciousness studies. Nevertheless, there is a broad scope among the articles, and one chapter contains the most complete coverage I have ever encountered of the Sidhis reported by advanced practitioners of Yoga. The material stresses the need for self-discipline and training in terms of meditation before such states can be achieved, but it makes it clear that there is a second path, however sporadic and dangerous it may be, and that is the path through psychedelic drugs. The chapter titled "The Biological Basis of Consciousness" is strong. The approach is largely through brain physiology and the chapter is well illustrated with instrumentation related to the parts of the brain that are accessed. In this kind of study, Consciousness is reduced to studies of perception and volition. The article speaks of the rhythm in encephalography and how it is stimulated, for example, by darkness in which one attempts to strain to see or by what is called negative experience. The articles on "Altered States of Consciousness" are very detailed. The first is largely concerned with practices of oriental religions and the disciplines by which such religions access aspects of Consciousness. The author stresses the value of mystical states and explains the confusion in which the word mystical is mistakenly taken to be "mysterious." In fact, the article seems to suggest that studies of mystics are the royal road to knowledge of Consciousness. Mysticism involves simply the direct perception of Reality. The author stresses the eastern discipline in which the attitude toward the study of Consciousness is by means of experience while in the West, it is characterized by intellectual study of the phenomena and laboratory work. The eastern practice is broad and deep, unlimited in scope. The western instrumentalist approach is always concerned with what is intellectually comprehendible at the time and therefore always changes. However, it is admitted that the training of a medical scientist (maybe 8 to 9 years from high school) is relatively easy and can be attained by many, whereas the learning of sufficient self-discipline and mind control to attain higher states of Consciousness is something that can be undertaken only by those of very strong character with a great deal of patience and time. Chapter 4 treats the Veda's (looking outward) and the Upanishads (looking inward) in great detail, emphasizing that meditation is the means for achievement. The text then turns toward "esoteric" Christianity and the attainment of higher states of Consciousness by disciplinary practices in some Christian monasteries. This leads to the discussion of psychic phenomena, their relation to studies of Consciousness, and the remarkable scientifically established powers of Yogis. The "high standard" of western living (i.e., the soft life) is not conducive to the attainment of higher states. However, as stated above, the attainment of higher states can also be reached by the use of psychedelic drugs, for example, as used by Shamans in Amazon- Book Reviews 659 ian societies. The experience of persons who undergo the rituals prescribed by the Shamans is described. In my opinion, the outstanding chapter in the book is the last of the three articles on "Altered States of Consciousness." It is a powerful survey of the whole field, covering the psychological approach, the superindividual level, the transindividual level, and finally the "universal level." There are phenomena that can only be interpreted if there are pathways in the brain much faster than those that are effectuated by ionic currents. This leads to some elements of the theory of hypnosis and telepathy. The chapter is valuable in the electrochemical part, relating some processes in the brain to the dielectric properties of ionic solutions in contact with membranes. The third part of the book consists of a survey of the EEG approaches to Consciousness studies. What kind of activity increases the rhythm? In what areas of the brain is the rhythm observed? How does meditation affect the EEG? What about the EEG in psychic healing? A study of a Qui Gong master is mentioned and here the rhythm was found in the anterior region of the brain. The last part of the book, the New Paradigm, chooses anesthesia as an important method in the study of Consciousness. It is surmised that anesthetics work on the water structure of cells and affect the structural dynamics of the microtubules, which have been presented more recently by Penrose as a possible locus of Consciousness. The last chapter is reductionistic and deals with EEG patterns, but it tries to begin a model for the functioning of Consciousness. Emphasis is placed on the properties of ionic solutions and ELF effects are discussed. Neuronetworks and information processing are related and compared to ELF phenomena. Finally, in the last part, one comes back to what was mentioned at the beginning of the book, the reduction of the state vector. Wigner's concept is that this is determined by the Consciousness of the experimenter. It is speculated that nonlocality makes Consciousness of a given brain globally related in terms of electromagnetic field to other brains. I consider this to be a valuable book. Consciousness is written about a great deal, but too much in terms of specialist papers. At present, I know of nowhere besides this book that covers so much ground in 300 pages. Indeed, the book could serve as a modern textbook of Consciousness. J. O'M. Bockris Molecular Green Technology 4973 Afton Oaks Drive, College Station, TX 77845 Searching for Eternity: A Scientist's Spiritual Journey to Overcome Death Anxiety by Don Morse. Memphis, TN: Eagle Wing Books, 2000. 428 pp. $19.95. ISBN 0-940829-27-4. Searching for Eternity is a timely overview of the area of spirituality and meta- 660 Book Reviews physics. This particular book is remarkable in its encyclopedic overview of paranormal events and religious/spiritual beliefs that point to the likelihood of an afterlife. The book includes an index 15 pages long while chapter notes and references take 31 pages. There are more than 100 illustrations of medieval paintings, biblical scenes, photographs of authors of spiritual teachings, and paintings done by the author himself. This encyclopedia of dying and afterlife literature should be on the bookshelf of everyone interested in the topic. In the first chapter, Morse discusses the death anxiety experienced by almost all humans. He defines it as "overwhelming fear of death and/or dying that usually involves all-encompassing fear about whether or not one will continue to exist in some form or fashion after physical death" (p. 20). He also describes various techniques to allay death anxiety and candidly recounts his own near-death experience (NDE) and tells how it changed his views on death. In the book, he expands on his own research about afterlife and how his discoveries can help others wrestle with death anxiety. This is an admirable task, as there is certainly a great deal of anxiety in American culture. Dr. Morse succeeds in providing a helpful overview to the concept that there is life after the death of the body. Morse discusses many seers, religions, and metaphysical ideas, as well as paranormal events, without attempting to analyze them and delve for deeper levels of understanding. Morse is Jewish, and his concentration on Jewish views is a welcome change from the common focus on Christian sources more often found in books of this kind. An eclectic range of organizations and individuals is surveyed for their views on the afterlife, and from this, the author draws his conclusions on the nature of the afterlife. In his survey of modern individuals' views on the afterlife, Morse includes such eminences as Emanuel Swedenborg, William Blake, and Edgar Cayce; all of them are certainly important thinkers on this topic. Trying not to miss anybody, he also includes L. Ron Hubbard, who is the founder of Scientology. Scientology is a controversial religion. Although the Internal Revenue Service has recognized Scientology as a religious organization for tax purposes, other countries have done the opposite. In certain countries, such as France, Belgium, and Greece, it is considered a commercial, rather than a religious, enterprise. Many people have publicly stated that in their opinion, Scientology is a destructive cult. It has been the target of much litigation of many kinds. Some of the little-known individuals and groups Morse surveys, such as the Christadelphiand, the Dutchmen Roe1 van der Meullen, and the German Anton Schneiderfranken, who wrote about afterlife in 1920 under the pseudonym Bo Yin Ra. None of them, however, are controversial in the manner of Scientology. The strongest part of the book is the chapter on NDE. It is clear that Morse has done a good deal of research in this area and understands it well. There are many interesting examples of NDEs, as well as a discussion about their common characteristics. These include a tunnel, an experience of light and dark, psychic abilities, a life review, and the experience of some kind of heavenly Book Reviews 661 presence. Morse describes four different kinds of NDEs, including the unpleasant ones, which he calls "hell-like experiences." This is an important addition to the book, because much of the NDE literature includes only descriptions of positive experiences. He also discusses the differences with nearing-death awareness, which often comes to those who are terminally ill. They differ in some important way from NDEs, notably in less often including a life review. The chapter on out-of-body experiences (OBEs) and apparitions is also fairly strong. Although Morse gives some tantalizingly interesting OBE cases, there is not much analysis, unlike in the NDE chapter. In discussing apparitions of various kinds, Morse concentrates on after-death contacts (ADCs). This is a very interesting description. There are visual, auditory, and tactile ADCs, as well as spirit and electronic voice phenomena. This last is when the deceased person's voice is heard through a radio, answering machine, or other medium. He also describes some other kind of visions, including predeath visions, dream visions, and mediumship. This last is a useful discussion, because the role of mediumship in the paranormal has not always been very clear. The idea of reincarnation is all important to the afterlife, and Morse surveys it. His chapter relies on the work of the well-known psychiatrist Brian Weiss, whose books have made reincarnation popular. Immortality, God, and the Origin of Universe and Life are the subjects of the following four chapters, while the topic of afterlife in ancient Egyptian, Greek, Hindu, Zoroastrian and Buddhist, and modern Judeo-Christian, Islamic religions are topics of the next six chapters. One chapter, a welcome and candid addition, is devoted to Morse's personal concepts of the afterlife. From all he has surveyed, Morse is trying to build a case for the idea of an afterlife that is not physical immortality but spiritual. His final chapters are an essential listing of probable characteristics of the afterlife, based on what traditions and recorded experiences tell us. He concludes that the soul survives the death, and one's life is a major determinant of the kind of afterlife one will have. Hateful and angry people will not have a pleasant afterlife, and thus it is much better to be loving, kind, considerate, and righteous. It seems slightly dogmatic, but Morse gives tips on how to order one's life, especially in terms of stress management, to make life quality, and therefore afterlife quality, as high as possible. Morse stresses that the findings of parapsychology, religions, and spirituality, and even science, all point to an afterlife that is based on one's own actions. The advice that is certainly quite sound is to live one's life as respectfully as possible. Alexander Imich President, Anomalous Phenomena Research Center 305 West End Avenue, New York, NY 10023 aimich @juno.corn 662 Book Reviews Gardner's Whys & Wherefores by Martin Gardner. Amherst, NY: Prometheus Books, 1999. xiii + 261 pp. $18.95 (paperback). ISBN 1-57392744-9. Martin Gardner is a fascinating character and the most compelling reason to read this book is to learn more about him. Forty years after first reading his criticism of parapsychology, many of those years enriched by his mathematical games column in Scientific American, I learned that Gardner is an expert on "The Rime of the Ancient Mariner," and "Casey at the Bat." I also learned that his taste in humor is low and bawdy. This paperback is a new edition of a collection first published in 1989. It consists of 16 essays and 20 reviews, plus a new preface to the paperback edition. The essays were originally published between 1965 and 1987; the reviews are dated between 1981 and 1988. Five essays and 16 reviews deal with natural or mathematical sciences or technology. The preface includes brief references to more recent material touching on one of the science essays and five of the reviews, four related to science. So, one would not read the collection to become informed about the latest developments in the sciences. Gardner writes wonderfully and it would probably be a pleasure to read one of his grocery lists. However, two of his reviews in this volume can be recommended for their relevance to the interests of Journal of Scientific Exploration readers. Chapter 17, "Polywater," originally appeared in Science Digest in September 198 1. It is a review of the book Polywater by Felix Franks (1981). Gardner summarizes the properties claimed for this polymeric form of water and charts the rise and demise of interest in polywater, a process that ended with the conclusion that impurities were the source of all interesting results. His last paragraph is a statement of his view that science is open to unorthodox ideas and tests them appropriately: On the positive side is the speed with which the scientific community corrected itself. The establishment erred not on the side of dogmatic rejection but on the side of tolerance. There were no witch-hunts. Deryagin did not settle in Siberia. Perhaps four years of fruitless research and acrimonious debate is a small price to pay for the thorough testing of an honestly claimed anomaly. No one interested in the sociology of science should pass up this absorbing chronicle. (p. 156) Chapter 21, "How Science Self-Corrects," brings Gardner again to his position that science should not be viewed as "a rigid orthodoxy, contemptuous of off-trail thinking." His platform is a review, originally published in Science (March 1984), of Dismantling the Universe by Richard Morris (1983). He notes Morris's examples of truly bizarre theories, such as the many-worlds interpretation of quantum mechanics, which are published in mainstream journals. He also notes that Morris cites the histories of topics such as polywater as reflecting favorably on science. Gardner supports Morris in refusing to apply Book Reviews 663 the term "crackpot" to such ideas after they are found to be mistaken, and he also supports Morris's application of "crackpot" to Velikovsky. Gardner feels that there is a vast difference between the methods of Deryagin and Velikovsky. It is likely that everyone would put some scientific theories into each of these categories: well supported, disproved, unproved but promising, unproved and very unlikely. Drawing boundaries can divide theories in this manner; deciding how to draw such boundaries divides the scientific community. Robert L. Raymond University of St. Thomas St. Paul, MN 55105-1096 References Franks, F. (1981). Polywater. Cambridge, MA: MIT Press. Morris, R. (1983). Dismantling the universe. New York: Simon and Schuster. Paranormal Experience and Survival of Death (Suny Series in Western Esoteric Traditions) by Carl B. Becker. Albany, NY: State University of New York Press, 1993. x + 257 pp. $23.95. ISBN 0-791-414-760. If, like the present reviewer, you have attained your 80th year, you may be forgiven for wanting to know whether you have at most a few more years of existence on this planet or whether you can look forward to an indefinite future life in some other guise in some other realm. Carl Becker does not purport to know the answer and does not seek to proselytize for any particular faith or creed, but he clearly, and rightly in my view, considers that we have been too hasty in letting ourselves be beguiled by the prevailing scientific materialism into abandoning this ancient and widespread belief. His first three chapters are devoted respectively to the idea of reincarnation, to hauntings and apparitions, and to near-death experiences. He then turns to general objections to evidence for the paranormal, be they rational or irrational objections. It is followed by a chapter on the conservatism and resistance to change in the sciences. This brings him to chapter six, the final chapter, in which he asks, "What Will the Next World be Like?" To anyone who still has an open mind on this issue, this book can be strongly recommended. The author does not try to browbeat the reader into accepting any particular position, but reading his book enables us to understand more clearly what is at issue and what we may reasonably believe. The text is fully annotated, and the bibliography runs to some 30 pages. Like any reviewer, I was agog to spot mistakes, but I could spot only two. Michael Polanyi was never a psychiatrist. He was a research physical chemist who, in later years, made a name for himself as a philosopher of science. Richet's first name was Charles, not Claude. I was surprised that there was no reference to Susan Blackmore's Beyond the Body (Heinemann, 1982). She 664 Book Reviews was, after all, something of a pioneer in stimulating interest in out-of-body experiences even if she eventually threw in her lot with the skeptics. But these are all minor points. The book will, I am confident, become essential reading for all who are involved in the survival issue. John Be108 6 Blacket Place, Edinburgh EH9 IRL, Scotland UK Scientific Development and Misconceptions through the Ages, A Reference Guide by Robert E. Krebs. Westport, CT: Greenwood Press, 1999. vii + 286 pp. $49.95 (paperback). ISBN 0-313-30226-X. The announced aim of this book is to present a short history of science and a review of misconceptions through the ages as they have arisen in all the fields of science. Because the author is a retired administrator at a health sciences tenter, it is not surprising that biological and applied areas receive special emphasis. The author insists that deliberate falsifications are beyond the scope of the study, which attempts to be comprehensive in other respects. Classification plays a major part in the author's attempt to distinguish between the sciences and other fields of belief. Most mathematicians will be surprised to find their discipline classified as an applied science. Students of theology will find its inclusion in the class of superstitions, a less than promising starting point for a comprehensive study of intellectual endeavors. Krebs tries to give due credit to philosophy by calling it a "borderline science." Philosophers will be able to forgive administrative misconceptions, which may be fruitful as topics for discussion in the classroom. Problems of classification appear once more when astrology is treated in the same chapter as cosmology. There is a large amount of instructive material in sections on medicine and health, theoretical biology, chemistry and physics, and conservation. The identification of historical disputes and outcomes is in many cases helpful at a certain level of instruction. In my opinion, some of the strongest lines of discussion in this book concern the identification of motivations for less than critical beliefs and belief systems. The errors upon which the author concentrates are of various kinds, ranging from technical mistakes in calculation to flaws in broad conceptual systems. A stronger account of misconceptions through the ages might devote more analytic attention to different levels of concept identification and use in the scientific endeavor. Robert I? Creegan Professor Emeritus, Philosophy The University at Albany, SUNY Journal of Scientific Exploration, Vol. 14, No. 4, pp. 665-667,2000 0892-3310100 6 2000 Society for Scientific Exploration 3 Society for Scientific Exploration 20TH ANNUAL SSE MEETING LA JOLLA RADISSON CONFERENCE CENTER AND HOTEL LA JOLLA (SAN DIEGO), CALIFORNIA PROGRAM Program Committee: Henry Bauer, Wayne Jonas, Beverly Rubik, Peter Sturrock (Chair) Program: Sessions on History and Philosophy of Science, UFOs, HealingIAlternative Medicine, Young Investigators presentations. See ScientificExploration.org for latest information. HOTEL FACILITIES MEETING ROOM La Jolla Radisson Hotel and Conference Center, 3299 Holiday Court, La Jolla, California 92037. SSE has the First Floor. HOTEL ROOMS & AMENITIES The La Jolla Radisson is a small hotel, where our meeting will probably be the main event. Hotel rooms are in three buildings, separate from lobby, meeting rooms, and restaurants. Hotel rooms are all the same price: $1 19 is our group rate if reservations are made by May 4. That rate is available for the three nights before the conference and the three nights after the conference, as well as the four nights of the conference. Most rooms have king-sized beds; a few have two queens. Rooms can be shared for no additional cost. All rooms have big screen TVs, VCRs, refrigerators, computer hookups, newspaper delivery, hair dryers, voice mail telephones, coffee makers, iron, etc. Hotel has an outdoor heated pool, an outdoor jacuzzi, a small garden/lawn/koi pond, an exercise room, and a putting green. An executive services room will have fax, Xerox, computer, and Internet services. CONTACT: La Jolla Radisson at 800-333-3333; Address: 3299 Holiday Court, La Jolla, California 92037; Hotel e-mail: [email protected]. Reserve before May 4, 2001, for the group rate; be sure to mention the SSE conference. For those sharing, reserve early for the few two-queen-bed rooms (most are one king bed). TRANSPORTATION San Diego Airport is 20 minutes away. The La Jolla Radisson Hotel has a free shuttle van that runs continuously to the airport from 7 a.m. to 11 p.m., leaving the hotel every hour. Pickups at the airport can be arranged by calling the La Jolla Radisson courtesy phone in the baggage claim area at the airport. There are two San Diego Radissons, so make sure you tell them you are going to La Jolla. The hotel has an Enterprise car rental agency in the lobby. Parking is free at the hotel. A stop for the public bus to San Diego, which takes 15-20 minutes and runs frequently, is a 10minute walk from the hotel. MEALS There are two restaurants in the hotel, a truly gourmet restaurant and a barlsnack bar where you can get sandwiches and where there is entertainment at night. For $12 you can order breakfast vouchers that include anything on the menu, tax, and tip. These can only be ordered ahead of time from SSE. There are other less expensive restaurants within walking distance, as well as a small shopping center with a deli and 24-hour-drug store. SSE EVENTS WELCOME RECEPTION Wednesday, June 6 , from 7-10 p.m., there will be an outdoor buffet on the lawn near the pool. This is included with your registration fee. LUNCHES You may register for a sit-down buffet lunch on the lawn each day for $10 per lunch per day. There is immediate access to the pool and hotel rooms. There are also refrigerators in the rooms. A shopping center is a five-minute walk (at the edge of the hotel parking lot). Other restaurants are within walking distance. La Jolla village is ten minutes away by car. FIELD TRIP San Diego Harbor cruise on Friday afternoon June 8 from 2-5 p.m. There are more than 50 points of interest in the harbor, including military installations, the national landmark Hotel del Coronado where we will probably dock and walk on their beach, the fishing fleet, etc. The trip is $25 per person, including roundtrip coach bus to the harbor (a 15-20 minute ride each way) and a threehour cruise. Reserve early, as the private boat holds only 50 people. BANQUET Saturday, June 9,7-9 p.m. at the La Jolla Radisson. Dinner is $35 per person and includes wine and entertainment. Society for Scientific Exploration 20th Annual SSE Meeting, June 7-9,2001 LaJolla Radisson Hotel & Conference Center, LaJolla (San Diego), California Please type or print clearly; All amounts are given in U.S. dollars Contact Information Affiliation (or City) for Badge Guest Name Guest Affiliation (or City) for Badge Street Address CityIStatelZIP PhoneFaxIE-mail Number in Party (please copy form and fill out relevant sections for each additional person) Meeting Registration (Per person) Early Registration Fee, $125 Late Registration Fee (after May 4, 2000), $135 Daily Registration Fee, $45 ($55 after May 4, 2001) Student registration (please fax ID card), $50, late fee $60 $ $ $ $ Reaistration Fee includes Wed. evenina outdoor O ~ e n i n a Reception Buffet Supper Gourmet Lunch Buffet on the Lawn Rate per person per day, $10 Indicate which days: Hotel Breakfast vouchers Rate per person per day, $12 (anything on menu, includes tax and tip) Indicate which days: Thursday Friday Saturday $ Thursday Friday Saturday $ Friday afternoon San Diego Harbor Cruise, 2 - 5 p.m. Rate per person, $25 (includes 20-minute coach bus trip) Saturday evening banquet at LaJolla Radisson Rate per person, $35 (includes dinner, wine, band) Choose: meat fish vegetarian TOTAL PAID: Payment Method (check one) Check Enclosed [ ] Charge my credit card Card Number Signature Please return with check payable to SSE o r with VISA/Mastercard information to: Charles Tolbert, SSE Treasurer, Box 3818, University Station, Charlottesville, VA 22903 Fax: 804-924-3104 HOTEL: CONTACT LaJolla Radisson 800-333-3333; [email protected] SSE Group rate $119 per room per night i f reserved by May 4,2001 [ ] Mastercard $ $ $ [ ]VISA Expiration Date -1 - - -- Journal of Scientijic Exploration, Vol. 14, No. 4, pp. 669-680,2000 0892-3310100 O 2000 Society for Scientific Exploration Index for Volume 14 The following pages provide comprehensive name and subject indexes for volume 14 of the Journal. The Name Index includes, by author, all articles and book reviews that have appeared over the last year, as well as books by authors that have been reviewed by someone else. Substantive citations to, and quotes from, an author's work have also been indexed; routine references have not. The Subject Index includes a list, sorted alphabetically by author, of all books reviewed in the Journal, under the header "Book reviews." The name of the reviewer is included in parentheses at the end of each entry. The Editors wish to thank Dawn Hunt and James Matlock for their diligence in compiling and checking the indexes. Name Index Abel, George, 432 n2,435 n9,445 Adamski, George, 292 Ader, Robert, 490 Allison, Ralph B., Minds in Many Pieces: Revealing the Spiritual Side of Multiple Personality (coauthor) (book review), 295-96 Almeder, Robert, book review by, 124-29 Alvarado, Carlos S., 367 Antoniadis, I., 273 Arakni-Hamad, N., 273 Arp, Halton, 304, What Has Science Come to?, 447-54 Atmanspacher, Harald, Investigating Deviations from Dynamical Randomness with Scaling Indices (coauthor), 1-18 Bakalar, James B., Psychedelic Drugs Reconsidered (co-author) (book review), 135-37 Barthes, Roland, 289 Baruss, Imants, Authentic Knowing: The Convergence of Science and Spiritual Aspiration (book review), 287-89 Bauer, Henry H., book reviews by, 485-91,654-57 editorials by, 304-5,497 Beeker, Carl B., Paranormal Experience and Survival of Death (book review), 663-64 Bell, Derrick, 458 Beloff, John, 457, book reviews by, 137-38,663-64 Bern, Daryl J., 55 Bengston, William F., The Effect of the "Laying on of Hands" on Transplanted Breast Cancer in Mice, 353-64 (co-author) Benor, Daniel J., 354 Bergson, Henri, 456 Berliner, Don, 466 Bishai, David, Can Population Growth Rule out Reincarnation? A Model of Circular Migration, 411-20 Blake, William, 660 Blanden, Robert V., Lamarck's Signature (co-author) (book review), 282-87 Blasland, Richard A., The Ordering of Random Events by Emotional Expression, 195-216 Bockris, John O'M., book reviews by, 129-32,657-59 Bohr, Niels, 257 Boller, E.,Mind/Machine Interaction 670 Index: Volume 14 Creegan, Robert F., book reviews by, 295-96,664 Darwin, Charles, 283-87 De Marre', Luc, 443-44,443 n16,445 DeQuincey, C., 214 n3 Dibble, W. E., Jr., 218 Dick, Steven J., Life on Other Worlds: The 20th Century Extraterrestrial Life Debate (book review), 477-80 Dillner, Debra K., Protection of Mice from Tularemia Infection with Ultra-Low, Serial Agitated Dilutions Prepared from Francisella tularensis-Infected Tissue (co-author), 35-52 Dimopoulos, S., 273 Dixon, N. F., 457 Djerassi, Carl, 657 Dobyns, York H. ArtREG: A Random Event Experiment Utilizing Picture-Preference Feedback, 383-409 (co-author) Contributions to Varience in REG Experiments: ANOVA Models and Specialized Subsidiary Analyses (co-author), 383-409 MindIMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 Overview of Several Theoretical Models on PEAR Data, 163-94 Dommanget, Jean, 431-46 Douglass, William Campbell, 630 Druffel, Ann, 291, Reanalysis of the 1965 Heflin UFO Photos, 583-622 (co-author) Dunne, Brenda J., 83,195-96,197,2 12, 215 ArtREG: A Random Event Experiment Utilizing Picture-Preference Feedback, 383-409 (co-author) Contributions to Varience in REG Experiments: ANOVA Models and Specialized Subsidiary Analyses (co-author), 73-89 MindJMachine Interaction Consortium: PortREG Replication Exper- Consortium: PortREG Replication Experiments (co-author), 499-555 Bond, Frederick Bligh, 462 Bosch, H., MindIMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 Bradish, G. J. ArtREG: A Random Event Experiment Utilizing Picture-Preference Feedback, 383-409 (co-author) MindIMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 Braude, Stephen, 462 Broderick, Damien, book review by, 289-93 Bryan, C. D. B., 292 Bryan, Ronald A., What Can Elementary Particles Tell Us About the World in Which We Live?, 257-74 Bucke, Richard, 133 Bullard, Thomas, 291 Burr, Harold, 196-97 Cayce, Edgar, 462,660 Chalmers, David, 139 Chauvin, RCmy, 445 Clark, Jerome, Cryptozoology A to 2: The Encyclopedia of Loch Ness Monsters, Sasquatch, Chupacabras, and Other Authentic Mysteries of Nature (co-author) (book review), 468-71 Clayton, Craig, book review by, 491-93 Clear, Constance, Reaching for Reality-Seven Incredible True Stories of Alien Abduction (book review), 463-65 Coleman, Loren, Cryptozoology A to 2: The Encyclopedia of Loch Ness Monsters, Sasquatch, Chupacabras, and Other Authentic Mysteries of Nature (co-author) (book review), 468-71 Cone, William P., The Abduction Enigma (co-author) (book review), 476-77 Corliss, William, 304, book review by, 480-82 Index: Volume 14 iments (co-author), 499-555 Dvali G.. 273 Eccles, Sir John, 137,138 Edwards, Paul, 411,419 Ehrenwald, Jan, 457 Eichwald, C., 218 Eisenstein, Charles, book review by, 135-37,293-95 Elguin, G., 355 Eliade, Mircea, 288 Emmons, Charles F., At the Threshold (book review), 293-95 Ertel, Suitbert, 304 Bulky Mars Effect Hard to Hide: Comment on Dommanget's Account of the Belgian Skeptics' Research, 431-46 The Mars Effect is Genuine: On Kurtz, Nienhys, and Sandhu's Missing the Evidence, 421-30 (co-author) Estebany, Oskar, 354-55 Estes, Russ, The Abduction Enigma (coauthor) (book review), 476-77 Eysenck, Hans, 456 Faith, Laura V., The Correlation of the Gradient of Shannon Entropy and Anomalous Cognition: Toward an AC Sensory System (co-author), 53-72 Fenwick, Elizabeth, The Truth in the Light (co-author) (book review), 132-35 Fenwick, Peter, The Truth in the Light (co-author) (book review), 132-35 Fowler, Raymond, 290-91,466 Franks, Felix, 662 Freeman, W., 474 Freud, Sigmund, 288,557 Friedman, Stanton, 466 Frye, Northrop, 289 Fuller, John, 290 Gardner, Martin, Gardner's Whys & Wherefores (book review), 661-63 Gauquelin, Michel, 421,422-29,433, 434,438-39,445 Globus, Gordon, 473 67 1 Goldschmidt, Richard, 482 Gould, Stephen J., 447 Grad, Bernard, 354, 355 Greeley, Andrew, 133 Greenberg, D. S., 657 Greenhouse, J. E., 97,99 Greenspoon, Lester, Psychedelic Drugs Reconsidered (co-author) (book review), 135-37 Greville, T. N. E., 643 Grey son, Bruce, 366-67 Grof, Stanislav, 136 Guo-long, Liu, 455 Hahn, Richard S., Understanding the Nature of Racial Prejudice, 458-59 Haines, Richard F., 121,646-47, book review by, 121-24 Reply to Klass, 647-48 (co-author) Valentich Disappearance: New Evidence and a New Conclusion (co-author), 19-33 Hansen, Ken, 22,25-31 Harrington, Anne, The Placebo EfSectAn Interdisciplinary Exploration (ed.) (book review), 485-91 Hastings, Arthur, With the Tongues of Men and Angels: A Study of Channeling (book review), 461-63 Hawking, Stephen, 448 Hayles, Katherine, 137 Heflin, Rex E., 583-622 Heuvelmans, Bernard, 469 Hill, Betty and Barney, 290 Ho, Mae-Wan, 197,218 Honorton, Charles, 55 Hopkins, Budd, 291,466 Houran, James, book review by, 476-77 Houtkooper, Joop M., 644-46 A Further Note on Walach and Schmidt's "Empirical Evidence for a Non-Classical Experimenter Effect," 643-44 (co-author) MindIMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 Hoyle, Fred, 458 Hubbard, L. Ron, 660 672 Index: Volume 14 Keil, Jiirgen, The Stability of Assessments of Paranormal Connections in Reincarnation-Type Cases, 365-82 (co-author) Kelson, Eric, Reanalysis of the 1965 Heflin UFO Photos, 583-622 (co-author) King, Chris, 472-73 Klass, Philip, 647-48, Response to "Valentich Disappearance: New Evidence and a New Conclusion," by Richard F. Haines and Paul Norman, 646-47 Klauber, Robert D., Modern Physics and Subtle Realms: Not Mutually Exclusive, 275-79 Klimo, John, 461 Knight, J. Z., 462 Kohane, Michael J., Energy, Fitness, and Information-Augmented Electromagnetic Fields in Drosophilia melanogaster (co-author), 217-3 1 Koruga, D., Consciousness (co-editor) (book review), 657-59 Kozima, Hideo, The Discovery of the Cold Fusion Phenomenon (book review), 129-32 Krebs, Robert E., ScientiBc Development and Misconceptions through the Ages (book review), 664 Kreiger, Dolores, 354 Krinsley, David, The Effect of the "Laying on of Hands" on Transplanted Breast Cancer in Mice, 353-64 (co-author) Kurtz, Paul, 421,424-29 Lang, Serge, 431 Lantz, N. D., 53-54,61 LaViolette, Paul, Earth Under Fire: Humanity's Survival of the Apocalypse (book review), 296-99 Lazarus, M., 557,558 Leslie, Desmond, 292 Lettieri, Arnold L., Jr., book review by, 281-82 MindJMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 Imich, Alexander, book reviews by, 465-68,659-61 Irving, Kenneth, The Mars Effect is Genuine: On Kurtz, Nienhys, and Sandhu's Missing the Evidence, 421-30 (co-author) Iyengar, S., 97,99 Jablonka, Eva, Epigenetic Inheritance and Evolution (co-author) (book review), 282-87 Jacks, Amanda, 238 Jacobs, David, 291,292,466 Jahn, Robert G., 9,195-96,197,212, 215 Contributions to Varience in REG Experiments: ANOVA Models and Specialized Subsidiary Analyses (co-author), 73-89 ArtREG: A Random Event Experiment Utilizing Picture-Preference Feedback (co-author), 383-409 MindIMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 James, C. L., 54-55 James, S., The Correlation of the Gradient of Shannon Entropy and Anomalous Cognition: Toward an AC Sensory System (co-author), 53-72 James, William, 472,557 Jibu, M., 474 Johnson, Donald, 123 Jonas, Wayne B., Protection of Mice from Tularemia Infection with Ultra-Low, Serial Agitated Dilutions Prepared from Francisella tularensis-Infected Tissue (co-author), 35-52 Josephson, Brian D., 116 Jung, Carl Gustav, 288,291,470 Katra, Jane E., Remote Viewing in a Group Setting (co-author), 107-14 Kauffman, Joel M., Should You Take Aspirin to Prevent Heart Attack?, 623-41 Keel, John, 292 Index: Volume 14 Lkvi-Strauss, Claude, 291 Levin, Michael, book reviews by, 282-87,471-75,649-54 Lindemann, Michael, 466 Lindley, Robyn A., Lamarck's Signature (co-author) (book review), 282-87 Lohrey, Andrew, The Meaning of Consciousness (book review), 137-38 Ludwiger, Illobrand von, Best UFO Cases-Europe, 1 21-24 Luke, W. I. W., 53-55,61 Luria, A. R., 456 Maccabee, Bruce, 121 Mach, Ernst, 430 Mack, John, 291,292,466 Mannion, Michael, Project Mindshift: The Re-Education of the American Public Concerning Extraterrestrial Life, 1947 to the Present (book review), 465-68 Mansfield, Victor, Synchronicity, Science and Soul-Making (book review), 491-93 Margulis, Lynn, 482 Martin, Brian, Supression Stories (book review), 483-85 Maslow, Abram, 288 Matheson, Terry, Alien Abductions: Creating a Modern Phenomenon (book review), 289-93 May, Edwin C., 53-55,61,62,llO, The Correlation of the Gradient of Shannon Entropy and Anomalous Cognition: Toward an AC Sensory System (co-author), 53-72 McConnell, Robert A., The Soverignty of Science, 115-20 McCully, Kilmer, 304 McDonald, James E., 590-604,6 18-20, 621-22 McGinn, Colin, Mysterious Flame: Conscious Minds in a Material World (book review), 649-54 Menendez, R. G., 218 Milton, Julie, 249-50,253 Milton, Richard, Shattering the Myths of Darwinism (book review), 480-82 Mischo, J., MindIMachine Interaction 673 Consortium: PortREG Replication Experiments (co-author), 499-555 Mitchell, Edgar, 466 Moody, Raymond, The Last Laugh (book review), 124-29 Morganato, V., 227 Morris, Richard, 662-63 Morrison, Philip, 479-80 Morse, Don, Searching for Eternity: A Scientist's Spiritual Journey to Overcome Death Anxiety (book review), 659-61 Morse, Melvin, 133 Miiller, A., 421,424,425,439 n13 Mullis, Kary, Dancing Naked in the Mind Field (book review), 654-57 Myers, Frederic W. H., 456-57 Nagel, Thomas, 133 Narlikar, Jayant, 449 Nash, Carroll, 355 Nelson, Roger D., 214 ArtREG: A Random Event Experiment Utilizing Picture-Preference Feedback, 383-409 (co-author) Contributions to Varience in REG Experiments: ANOVA Models and Specialized Subsidiary Analyses (co-author), 383-409 MindIMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 Nienhuys, J. W., 421,423,424-29,439 n13,444,445-46 Nordberg, Robert, book review by, 463-65 Norman, Paul, 21-22,646-47 Reply to Klass, 647-48 (co-author) Valentich Disappearance: New Evidence and a New Conclusion (coauthor), 19-33 Nlbrretranders, Tor, The User Illusion: Cutting Consciousness Down to Size (book review), 281-82 Northrop, F. S. C., 196-97 Nuland, Sherwin, 134 Null, G., 355 Onetto, B., 355 674 Index: Volume 14 Shapiro, Arthur K., The Powerful Placebo-From Ancient Priest to Mode m Physician (co-author) (book review), 485-91 Shapiro, Elaine, The Powe#.d PlaceboFrom Ancient Priest to Modem Physician (co-author) (book review), 485-91 Shaposhnikov, M. E., 269-70 Sheldrake, Rupert, 138, A Dog That Seems to Know When His Owner is Coming Home: Videotaped Experiments and Observations (coauthor), 235-55 Smart, Pamela, A Dog That Seems to Know When His Owner is Coming Home: Videotaped Experiments and Observations (co-author), 235-55 Smith, Matthew, 249-50,253 Snel, F., 354 Spaulding, William, 604,605 Spottiswoode, S. James P., 54-55, The Correlation of the Gradient of Shannon Entropy and Anomalous Cognition: Toward an AC Sensory System (co-author), 53-72 Stapp, H., 472 Steele, Edward J., Lamarck's Signature (co-author) (book review), 282-87 Steiger, Brad, 292 Stevenson, Ian, 304,417,571-72, book review by, 132-35 The Stability of Assessments of Paranormal Connections in Reincarnation-Type Cases, 365-82 (co-author) Unusual Play in Young Children Who Claim to Remember Previous Lives, 557-70 Strasenburgh, Gordon, book review by, 483-85 Streiber, Whitley, 289,292,656 Sturrock, Peter A., 93 Swedenborg, Emanuel, 660 Targ, Russell, Remote Viewing in a Group Setting (co-author), 107-14 Taylor, S. M., 47 Petersen, John L., book review by, 296-99 Pflock, Karl T., book review by, 477-80 Place, U. T., 139 Popp, F. A., 197 Popper, Karl, 1 37, 1 38 Porshnev, Boris F., 470 Pratt, David, 304, Plate Tectonics: A Paradigm Under Threat, 307-52 Pribram, Karl, 456 Radin, Dean I., 1 16,214,239,628 Rakovic, D., Consciousness (co-editor) (book review), 657-59 Randle, Kevin D., The Abduction Enigma (co-author) (book review), 476-77 Raymond, Robert L., book review by, 661-63 Reich, Wilhelm, 196, 197-99,212-13, 215,467,468 Rhine, J. B., 139 Rogo, D. Scott, 291 Rosenthal, Robert, 94-97 Rubakov, V. A., 269-70 Sabom, Michael, Light and Death (book review), 132-35 Sagan, Carl, 136 Sandhu, R., 42 1,424-29Scargle, Jeffrey D., Publication Bias: The "File-Drawer" Problem in Scientific Inference, 91-106 Scheingraber, Herbert, Investigating Deviations from Dynamical Randomness with Scaling Indices (co-author), 1-1 8 Schmidt, Helmut, 8,304 Schmidt, Stefan, Reply to Houtkooper, 644-46 (co-author) Schnabel, Jim, 292 Schneider, Adolf, 123 Schneiderfarnken, Anton, 660 Schuessler, John F., 121 Schwartz, G. E. R., 214 Schwartz, Stephen A., book review by, 461-63 Scott, James A., book review by, 287-89 Index: Volume 14 Tiller, William A., 218, Energy, Fitness, and Information-Augmented Electromagnetic Fields in Drosophilia melanogaster (co-author), 217-31 Timm, Ulrich, 644-46, A Further Note on Walach and Schmidt's "Empirical Evidence for a Non-Classical Experimenter Effect," 643-44 (co-author) Tucker, Jim B., A Scale to Measure the Strength of Children's Claims of Previous Lives: Methodology and Initial Findings, 571-8 1 Turney, Jane, 239 Utts, Jessica, 67 Vaitl, Dieter, 644-46, A Further Note on Walach and Schmidt's "Empirical Evidence for a Non-Classical Experimenter Effect," 643-44 (co-author) Valentich, Frederick, 19-33,646-48 Vallee, Jacques, 291,600 Van der Muellen, Roel, 660 Van Flandern, Tom, Dark Matter, Missing Planets and New Comets (book review), 483-85 Vandervert, Larry, Understanding Tomorrow's Mind (ed.) (book review), 471-75 Veicsteinas, A., 227 Vilenskaya, Larissa, book review by, 468-71 675 Walach, Harald, Reply to Houtkooper, Vaitl and Timm, 644-46 (co-author) Walleczek, J., 218 Walter, B., MindIMachine Interaction Consortium: PortREG Replication Experiments (co-author), 499-555 Walton, Travis, 291 Wegener, Alfred, 307 Weiss, Brian, 661 Welsome, Eileen, 452 West, B., 472 Westman, James, 470 Whitaker, Julian, 628,628 n l Wickramasinghe, Chandra, 458 Wiseman, Richard, 249-50,253 Wolff, Franklin, 288 Wood, Robert M., Reanalysis of the 1965 Heflin UFO Photos, 583-622 (co-author) Wunder, E., 446 Xiong, Jesse Hong, Is the Activity of the Cerebral Cortex a "Brake" on the Functioning of the Brainstem?, 455-457 Yasue, K., 372 Zhang, C., 197 Ziman, John, 304 Zingrone, Nancy L., 367 1 676 Index: Volume 14 Subject Index 1 Acquired Immune Deficiency Syndrome (AIDS), 452,654 Alien abductions, 273,289-93,463-65, 656 Alternative healing, 108, 109 of cancer in mice, 353-64 placebo effect and, 485,486,487,489 Animals dogs as psychic pets, 233-55 fitness experiments with fruit flies (Drosophila melanogaster), 217-31 See also Mice Anomalous cognition (AC) Shannon entropy and, 53-72 Apparitions, 661 ArtREG, 383-409 Aspirin heart attacks and, 623-41 Astrology, 297-98,299,654,656 see also Mars effect Astronomy, 297,448-50,483-85 Bayesian analysis, 193-94 Belgian Committee for the Scientific Investigation of Alleged Paranormal Phenomena (Para Committee) Mars effect research of, 422,429 n10, 431-46 Biology, 452 fitness experiments, 217-31 theory in, 282-87 see also Alternative healing; DNA; Reichian biopsychiatry Book reviews Allison, R., with Schwarz, T., Minds in Many Pieces: Revealing the Spiritual Side of Multiple Personality Disorder, 295-96 (R. F. Creegan) Baruss, I., Authentic Knowing: The Convergence of Science and Spiritual Aspiration, 287-89 (J. A. Scott) Becker, C. B., Paranormal Experience and Survival of Death, 663-64 (J. Beloff) Clear, C., Reaching for RealitySeven Incredible True Stories of Alien Abduction, 463-65 (R. Nordberg) Coleman, L. and Clark, J., Cryptozoology A-Z: The Encyclopedia of Loch Ness Monsters, Sasquatch, Chupacabras, and Other Authentic Mysteries of Nature, 468-71 (L. Vilenskaya) Dick, S. J., Life on Other Worlds: The 20th-Century Extraterrestrial Life Debate, 477-80 ( K . T. Pflock) Emmons, C. F., At the Threshold, 293-95 (C. Eisenstein) Fenwick, P., The Truth in the Light, 132-35 (I. Stevenson) Gardner, M., Gardner 's Whys and Wherefores, 661-63 (R. L. Raymond) Grinspoon, L. and Bakalar, J. B., Psychedelic Drugs Reconsidered, 135-37 (C. Eisenstein) Harrington, A. (ed.), The Placebo Effect-An Interdisciplinary Exploration, 485-91 ( H . H. Bauer) Hastings, A., With the Tongues of Men and Angels: A Study of Channeling, 461-63 (S. A. Schwartz) Jablonka, E. and Lamb, M. J., Epigenetic Inheritance and Evolution, 282-87 (M. Levin) Kozima, H., The Discovery of the Cold Fusion Phenomenon, 129-32 (J. O'M. Bockris) Krebs, R. E., Scientijc Development and Misconception through the Ages, 664 (R. F. Creegan) LaViolette, P., Earth Under Fire: Humanity's Survival of the Apocalypse, 296-99 (J. L. Peterson) Lohrey, A., The Meaning of Consciousness, 137-38 (J. Beloff) Ludwiger, I. v., Best UFO Cases-Europe, 121-24 (R. F. Haines) Mannion, M., Project Mind-shift: The Re-Education of the American Index: Volume 14 Public Concerning Extra-terrestria1 Life, 1947 to the Present, 465-68 (A. Imich) Mansfield, V., Synchronicity, Science and Soul-Making, 491-93 (C. Clayton) Martin, B., Suppression Stories, 483-85 (G. Strasenburgh) Matheson, T., Alien Abductions: Creating a Modern Phenomenon, 289-93 ( D . Broderick) McGinn, C., Mysterious Flame: Conscious Minds in a Material World, 649-53 (M. Levin) Milton, R., Shattering the Myths of Darwinism, 480-82 (W. Corliss) Moody, R., The Last Laugh, 124-29 (R. Almeder) Morse, D., Searching for Eternity: A Scientist's Spiritual Journey to Overcome Death Anxiety, 659-61 (A. Imich) Mullis, K., Dancing Naked in the Mind Field, 654-57 (H. H. Bauer) N~rretranders, The User Illusion: T., Cutting Consciousness Down to Size, 281-82 (A. L. Lettieri, Jr.) Rakovic, D. and Koruga, D. (eds.), Consciousness, 657-59 (J. O'M. Cockris) Randle, K. D., Estes, R., and Cone, W. P. The Abduction Enigma, 476-77 (J. Houran) Sabom, M., Light and Death, 132-35 (I. Stevenson) Shapiro, A. K. and Shapiro, E., The Powerful Placebo-From Ancient Priest to Modem Physician, 485-91 (H. H. Bauer) Steele, E. J., Lindley, R. A., and Blanden, R. V., Lamarck's Signature, 282-87 (M. Levin) Van Flandern, T., Dark Matter, Missing Planets and New Comets (Paradoxes Resolved, Origins Illuminated), 483-85 (G. Strasenburgh) Vandervert, L. (ed.), Understanding Tomorrow 's Mind, 471-75 (M. Levin) Brain psychic abilities and, 456-57 677 Cancer in mice, alternative healing of, 353-64 Channeling, 461-63 Chaos theory, 471 consciousness and, 471 Christadelphiand, 660 Clairvoyance, 114 Climatology, 297,298-99 Cold fusion, 129-32,480 ComitC Franqais pour 1 ' ~ t u d des e PhCnomknes Paranormeaux (CFEPP) astrology study of, 422,423 Comit6 PARA see Belgian Committee for the Scientific Investigation of Alleged Paranormal Phenomena Committee for the Scientific Investigation of Claims of the Paranormal (CSICOP), 116,117,125,126, 290,294 astrology study of, 421-30 Consciousness, 657-59 chaos theory and, 471,472 emotions and, 196 mind-brain relationship question and, 135-37,649-53 philosophy of, 281-82 psi phenomena and, 137 quantum physics and, 278-79,47 1, 472-73 Core1 Stock Photo Library of Professional Photographs used in experiments in anomalous cognition, 56 Cryptozoology, 468-71 Dissociative Identity Disorder (DID) spiritual side of, 295-96 DNA, 282,283,285-86,292 Electromagnetism 678 Index: Volume 14 Immunopharmacology, 35-52 Institut fiir Grenzgebiete der Psychologie und Psychohygiene (IGPP) replication of PEAR studies by, 8-10, 13-15,16 International Astra Meeting remote viewing workshop at, 108 Laying on of hands see Alternative healing Los Angeles NICAP Subsection (LANS), 584-98 Lysergic acid diethylamide (LSD), 136 Mars topography of, 484 Mars effect (Gauquelin effect), 421-30, 431-46 Meta-analysis, 91-106,530-53 Mice alternative healing of cancer in, 353-64 immunization of, 35-52 Mind-brain relationship placebo effect and, 490 Mind-machine interactions theoretical models on, 163-94 Mind-matter research random event generators (REGS) in, 8-16 temporal correlations in, 1-1 8 MindIMachine Interaction Consortium studies by, 499-555 Monte Carlo simulation mindlmachine interactions and, 524-38,530,531 Morphogenetic fields, 482 Multiple Personality Disorder (MPD) spiritual side of, 295-96 Mutual UFO Network, Central European Section (MUFON-CES), 123 see also Society for the Scientific Exploration of Anomalous Atmospheric and Radar Phenomena, MUFON-CES, Inc. Mythology UFO and abduction accounts and, 289,291,293 biological organisms and, 197, 217-31 information-augmented electromagnetic fields Emotions REG output and, 195-216 EmotREG, 195-216 Engineering anomalies research operator effects in, 79-82,87 random event generators (REGS)in, 73-89,499-555 Evolution Darwinian, 284-87,448,480-82 Lamarckian, 282-87,482 Experimenter effect, 643-46 Extrasensory perception (ESP), 110, 119 see also Anomalous cognition FieldREG, 200,201,203,206,208 File-drawer problem see Publication bias Folklore UFO and abduction accounts and, 291,293 Freiburg Anomalous Mind/Machine Interactions Group see MindIMachine Interaction Consortium Fruit flies (Drosophila melanogaster) fitness experiments with, 217-31 Futurology, 296-99 Gender differences random event generators (REGs) and, 87,195 Geology, 297,298-99 plate tectonics, 307-52 Giessen Anomalies Research Project (GARP) see MindIMachine Interaction Consortium Hallucinogens, 135-37 Heart attacks aspirin and, 623-41 Holistic medicine psychedelic experience and, 137 Homeopathy, 35-52,487 Index: Volume 14 National Investigations Committee on Aerial Phenomena (NICAP), 584, 585,588-89,590,596,599 see also Los Angeles NICAP Subsection (LANS) Near-death experiences (NDEs), 132-35,660-61 alternative universes and, 271-72, 278 Orgonomy, 197-99 see also Reichian biopsychiatry Out-of-body experiences (OBEs), 661, 663-64 Para Committee see Belgian Committee for the Scientific Investigation of Alleged Paranormal Phenomena Parapsychology, 480 called pseudoscience, 125,126 gansfeld research in, 54 see also Anomalous cognition; Engineering anomalies research; Mindmachine interactions; Mind-matter research Placebo effect, 485-91 Plate tectonics evidence against, 307-52 Play in reincarnation cases, 557-70 Playful paranormalism, 127-28 Polywater, 662 PortREG, 499-555 Precognition, 114 Princeton Engineering Anomalies Research (PEAR) Laboratory remote viewing studies of, 214 replications of studies by, 499-555 theoretical models on data from, 163-94 see also Engineering anomalies research; Random event generators Psi ability brain and, 456-57 Psi phenomena science and, 115-20 Psychedelic drugs see Hallucinogens Psychic pet phenomenon, 233-55 Psychoanalysis see Reichian biopsychiatry Psychokinesis (PK), 119 Publication bias, 91-106 679 Quantum mechanics, 196, 197,288,296 extra dimensionslother universes, 257-74,275-79,662 subtle realms and, 275-79 virtual and anti-particles (Dirac equation), 257-74 Racial prejudice, 458-59 Random event generators (REGS) emotions and, 195-216 experiments with, 8-16,73-89, 383-409,499-555 parapsychology and, 499-555 see also FieldREG Reichian biopsychiatry emotions and REG outputs in, 195-216 UFOs and, 467-68 Reincarnation, 661 population growth and, 411-20 psychedelic experience and, 136,137 Reincarnation cases scale to measure strength of, 571-81 stability of, 365-82 unusual play in, 557-70 Religion psi and, 116-17 science and, 109,447-54 Remote perceptionlviewing, 61,68, 107-14,197,214 Science history of, 662,664 parapsychology and mainstream, 115-20 religion and, 109,447-54 sociology of, 447-54,662-63 spirituality and, 287-89 suppression of unorthodox views in, 304-5 Shannon entropy 680 Index: Volume 14 Tarot, 207,209 Telepathy, 654,656 human-animal, 233-55 Transpersonal psychology psychedelic experience and, 137 UFO sightings, 121-24 Bass Straight (Australia), 19-33, 646-48 Fort Belvoir (California), 583-622 Vaddo Island (Sweden), 122 UFOs and ufology, 293-95,477-8 1 alternative interpretations, 123 crash investigation, 19-33 European, 121-24 extra-terrestrial hypothesis (ETH), 123,479 folklore and, 291,293 photographic cases, 122 photographs, image analysis, 583-622 psychedelic experience and, 137 quantum particles and, 272-73,278 scholarship on, 121-24 U.S. government and, 465-68 see also Alien abductions anomalous cognition and, 53-74 Society for Scientific Exploration (SSE), 288 Society for the Scientific Exploration of Anomalous Atmospheric and Radar Phenomena, MUFON-CES, Inc., 121 Stanford Research Institute (SRI), 114 Statistics and methodology bitwise effect, 163-94 decision augmentation theory, 163-94 randomness, 1-2,6 scaling indices, 5-7 teleological models, 163-94 theoretical models, 163-94 time series analysis, 1-18 time-normalized models, 163-94 transition probabilities, 1-18 see also Bayesian analysis; Experimenter effect; Meta-analysis; Monte Carlo simulation Subliminal perception, 281,457 Survival after death, 659-61,663-64 impossibility of proving, 124-29 see also Reincarnation Synchronicity, 491-93 Society for Scientific Exploration Prof. Peter A. Sturrock, President Varian 302 Stanford University Stanford. CA 94305-4060 Prof. Robert Jahn, Vice President D334, School of Engineering & Applied Science Princeton University Princeton, NJ 08544-5263 Prof. L. W. Fredrick, Secretary Department of Astronomy P. 0.Box 3818 University of Virginia Charlottesville, VA 22903-0818 Prof. Charles R. Tolbert, Treasurer Department of Astronomy P. 0.Box 3818 University of Virginia Charlottesville, VA 22903-08 18 Ms. Brenda Dunne, Education Officer C 13 1, School of Engineering &Applied Science, Princeton University, Princeton, NJ 08544-5263 Prof. Ezio M. Insinna, European Representative Bioera 18, Allee des Freres Lumiere Bussy St. Georges 77600 France Council Dr. Marsha Adams 1I00 Bear Gulch Rd. Woodside, CA 94062 Dr. Stephan Baumann fMRI Project Director Psychology Software Tools, Inc. 2050 Ardmore Blvd. Pittsburgh, PA 15221 Dr. John S. D e n Albq. Seismology Center Albuquerque, NM 87 1 15 Dr. Thomas M. Dykstra 3499 NW 97th Blvd., Suite 6 Gainesville, FL 32606 Dr. Wayne B. Jonas Dept. of Family Medicine 4301 Jones Bridge Road Bethesda, MD 20814 Dr. Roger D. Nelson C 131, Engineering Quad. Princeton University Princeton, NJ 08544 Dr. Harold E. Puthoff Institute for Advanced Studies-Austin 4030 W. Braker Ln., Suite 300 Austin. TX 78759-5329 Dr. Beverly Rubik Institute for Frontier Science 6114 LaSalle Avenue, #605 Oakland, CA 9461 1 Dr. Robert M. Wood 1727 Candlestick Ln. Newport Beach, CA 92660
Copyright © 2024 DOKUMEN.SITE Inc.