Parascience has so far failed to produce a single repeatable finding and, until it does, will continue to be viewed as an incoherent collection of belief systems steeped in fantasy, illusion and error.
Originally appeared in Nature Vol. 320, 13 March 1986, pp. 119-124.
Many factors of a psychological nature foster paranormal beliefs and make them a common feature of human thinking and behaviour. Our cultural traditions are steeped in religion and magic, many features of which lend themselves to belief in supernatural agencies. Scientific thinking is a recent departure in human history and scientific ideas have had little time to affect the magical thinking from which science itself evolved.
Sociologist D. O’Keefe argues that paranormal research has evolved from within the traditions of magic which themselves evolved from religion51. The current occult revival is seen as a reaction to the excessive rationalism which many perceive in science. O’Keefe argues that religion created the ‘cloud-cuckoo land’ in which magic, and thence the paranormal, can flourish. Yet scientists are often ill-prepared to provide the necessary counterbalancing rational account of the paranormal. Against this background of magi co-religious entrenchment, there are some extra psychological processes that make paranormal beliefs an inevitable characteristic of human consciousness and thinking.
A mental image is a quasi-perceptual experience in the absence of an objective stimulus. There are huge individual differences in the reported vividness and controllability of images. In Western cultures 1-5% of the population appears regularly to experience fantasies which seem as real as actual events even though they are entirely fictional52. Such individuals often experience vivid, uncontrollable ‘eidetic’ images of almost hallucinatory quality53, are highly suggestible and can be easily hypnotized52. They report more putatively paranormal experiences, such as telepathy, precognition, ghosts and out-of-the-body experiences. While mental imagery has a large number of practical uses in thinking, memory and problem solving, it can also occur in altered states of consciousness in which the normal level of lucidity is no longer present53.
Research conducted a century ago by E. Gurney and F. W. H. Myers described 27 cases of ‘spirit communication’ from deceased persons54. Eighteen of the apparitions occurred in sleep-related states normally associated with highly vivid and autonomous images which are easily mistaken for reality. The remaining cases occurred in subjects who were fully awake and these could easily have been structural eidetic images stimulated by thought-processes of the daydreaming kind53. H. Sidgwick noted that 9.9% of 17,000 subjects had experienced at least one vivid visual, auditory or tactile image of a living being or object while completely awake55. The appearance of ghosts is shaped by cultural expectancies and beliefs about what a ghost should look like56. Mental images can be easily misinterpreted in terms of pre-existing beliefs57.
Otherwise known as mental set, expectancy provides the framework within which we organize new experience. Human cognition is not a simple copying process but entails a constructive striving or ‘effort after meaning’. What we experience is often more a confirmation of belief than a matter of plain fact. Beliefs are not automatically updated by the best evidence available, but have an active life of their own and fight tenaciously for their own survival. They tell us what to read, what to listen to, who to trust and how to rationalize contrary information4,5,57.
Selective exposure protects beliefs from more dramatic forms of contradiction. When the mentalist U. Geller visited the city of Dunedin in New Zealand there were seven different opportunities to obtain information abut his alleged psychic abilities: four media interviews, two newspaper stories and one stage performance. Of 17 subjects who, before Geller’s visit, were already ‘believers’ 15 selected three or more of the available exposures. Of 20 ‘non-believers’, only 10 selected as many as three exposures (X2(1) = 6.13; P<0.02).
A further problem is that when we are exposed to relevant information, our opinion revisions are often less than optimal, and we act like conservative Bayesians58, with a confirmation bias59. In a recent ‘ESP’ demonstration to a class of 226 psychology students, presented as an exercise in observation, I performed five mentalists’s tricks consisting of: (I) correctly naming a colour written out of sight; (2) correctly transmitting a colour name to a volunteer who, like me, had not previously seen it; (3) helping a volunteer correctly to read messages sealed inside envelopes or to appear to transmit messages to me; (4) producing bent keys which I had not previously touched; and (5) moving or stopping the hands of a watch in a mysterious manner.
Although at no time did I claim to be psychic, 90% of the class stated that I had demonstrated psychic ability. When the results from subjects who had previously been classified as ‘believers’ and ‘sceptics’ were analysed separately, 79% of believers thought at least three of the five effects were psychic compared with only 43% of sceptics (P < 0.001).
Naturally, we often encounter information that is unexpected or ambiguous. In such instances, there is a second line of defence: the data can be selectively perceived or even misperceived so that they still appear to support our beliefs by ‘subjective validation’4. One illustration of this powerful cognitive defence in the context of ESP research is the strong conviction that one has successfully viewed a complex target site by ESP in a remote-viewing experiment even when one is completely wrong (Fig. 2).
There are many now-classic examples of subjective validation: the prophecies of the Delphic Oracle and Nostradamus60, the discovery of N-rays61, phlogiston, Vulcan, the canals on Mars, flying saucers, Freud’s interpretation of dreams, prejudice, faith-healing, the placebo effect, bone pointing and the ‘evil eye’. Beliefs of all kinds tend to be self-perpetuating.
Psi phenomena consist of an experience, image or thought matched by some other similar experience, image or thought. Collections of such coincidences have been published by A. Koestler62, L. Rhine,63 and others based on the assumption that odd-matches of events cannot occur purely by chance.
Probability theory shows that an event which is improbable over a short run can become highly probable over the long run. If five coins are tossed all at once on a single occasion the probability of obtaining five heads is 2 -5 or approximately 0.03. If the coin tossing is repeated 100 times the probability of five heads somewhere in the series is approximately 0.96.
The principle of the long run is easy to grasp in simple situations but much less visible in the more chaotic world of spontaneous human experience. Calculation shows how easily Koestler could obtain his 40-plus odd-match anecdotes. Assuming that in an ordinary day a person can recall 100 distinct events, there are 100C2 or 4,950 pairs of events per day. Odd-matches can be remembered for years, perhaps 10 yr or 3,650 days. If Koestler knew 1,000 people, he could draw upon a total pool of 4,950 x 3,650 x 1,000, or more than 18 x 109 pairs of events. That Koestler obtained 40 striking odd-matches seems hardly surprising.
Koestler’s fallacy (see ref. 4) is certainly not unique to him, although he was one of a small group of analysts who wanted to make a scientific revolution out of it. The fallacy is widespread and several biases contribute to it. First, we notice and remember odd-matches. Second, we do not notice non-matches. This triggers the short-run illusion that makes the oddmatch seem improbable. Third, we are normally poor estimators of probabilities, especially for combinations of events.
Another class of psychic-looking experiences is generated by invisible chains of cause and effect which bias the probabilities away from chance levels. Failure to randomize target stimuli properly in ESP experiments is a good example of this. Thus, Tart reported a successful ESP experiment in which his subjects learned to score above chance in guessing which of 10 digits was displayed by an apparatus in another room following the presentation of feedback)4. The random number generator mistakenly avoided using the same digit twice in succession, a bias which is matched by the pervasive ‘gambler’s fallacy’. When Tart removed this bias, the ‘ESP’ also disappeared65.
Another unseen factor, used by illusionists, is the ‘population stereotype’. The performer ‘sends a message to the audience, saying “I am thinking of a number between 1 and 50, both digits are odd, and different“. Controlled experiments show that the most common response for the 1-50 problem is 37, which accounts for 30-35% of all responses, and the second most common response is 35 (20- 25%)4. If the performer always says he had been thinking of 35 and then changed his mind to 37, at least 50% of the audience will be thinking of the ‘correct’ number.
Human beings never behave randomly. Our experiences contain many culturally shared elements such that particular items are associated with particular verbal contexts. This causes associative networks to be set up and a tendency towards nonrandom, stereotypical responses even when there is freedom to choose.
Other unnoticed causes of putatively psychic effects include subliminal and non-verbal sensory cues66 which may lead to common thought patterns in different people, presenting the illusion of telepathy.
The ‘Will to Believe’
What factors differentiate believer from sceptic? Psychologists down the ages have puzzled over the question of what motivates different world-views and the so-called will to believe. Research conducted by J. Waugh used Kelly’s personal construct theory. In this framework67, people vary in the quality and extent of their investigatory procedures so that, while some may be working to establish an ordered and meaningful world which is not highly predictable or readily explained, others may be content that they already have all the necessary explanatory constructs.
In Kelly’s theory, each individual deals with the world in terms of a hierarchial system of constructs with which people, objects and events are compared, contrasted and predicted. Core constructs have relatively superordinate positions and a large range of convenience while peripheral constructs are relatively subordinate and more easily altered. Waugh compared the personal construct systems of sceptics and believers in the paranormal using a belief questionnaire. Ten subordinate and ten superordinate67constructs were generated using standard procedures and each subject’s constructs were tested for their relative resistance to change and the number of implications entailed by changing the subject’s preferred pole on the 20 constructs and 10 paranormal beliefs (Fig. 3).
Believers’ core constructs were significantly more resistant to change and there was a parallel difference in the number of implications resulting from changes at the superordinate level. Compared with sceptics, believers seem to possess much tighter construct systems in which any change at the core level implies a significantly greater upheaval or threat. Waugh also found that believers had significantly higher neuroticism scores than sceptics (see also ref. 68). These data are congruent with those reported by Zusne and Jones57 who found that believers are less flexible than sceptics when confronted with disconfirming evidence. Content analyses of believers’ construct systems indicate the presence of spiritual, non-materialist constructs at superordinate level. Such core constructs are not easily shaken because they are closed off from empirical considerations and appear to be impermeable to rational persuasion. Hence the feeling of futility experienced in trying to hold rational discussion between believer and sceptic; one could well be arguing about the existence of God. Belief in the paranormal is metaphysical and therefore not subject to the constraints of empirically based science.
Parascience has all the qualities of a magical system while wearing the mantle of science. Until any significant discoveries are made, science can justifiably ignore it, but it is important to say why: parascience is a pseudo-scientific system of untestable beliefs steeped in illusion, error and fraud.
I thank Jerry Andrus, Bob Audley, Ray Hyman, A. R. Jonckheere, Peter McKellar, J. Randi, Christopher Scott, Jean Waugh and many colleagues in CSICOP for useful discussions and information. The late Richard Kammann contributed substantially in the earlier stages of this research.
- Kurtz, P. Skeptical Inquirer 3, 14-32 (1978).
- Diaconis. P. Science 201, 131-136 (1978).
- Hansel. C. E. M. ESP and Parapsychology: A Critical Reevaluation(Prometheus, Buffalo. 1980).
- Marks. D. & Kammann, R. The Psychology of the Psychic (Prometheus, Buffalo, 1980).
- Alcock. J. E. Parapsychology: Science of Magic? (Pergamon, Oxford, 1981).
- Frazier. K. (ed.). Paranormal Borderlands of Science(Prometheus, Buffalo. 1981).
- Gardner, M. Science Good Bad and Bogus (Prometheus, Buffalo. 1981).
- Randi. J. Flim-Flam! Psychics, ESP. Unicorns. and Other Delusions(Prometheus. Buffalo. 1982).
- Kurtz. P. Skeptical Inquirer 8, 239-246 (1984).
- Frazier. K. (ed.) Science Confronts the Paranormal (Prometheus. Buffalo, 1985).
- Kurtz, P. (ed.) A Skeptics Handbook of Parapsychology (Prometheus. Buffalo, 1985).
- Bunge, M. Skeptical Inquirer 9, 36-46 (1984).
- Tiller, W. A. New Scient. 62, 160-163 (1974).
- Pehek, J.O.,Kyler, H.J. & Faust. D. L. Science 194, 263-270 (1976).
- Houdini Miracle Mongers and Their Methods (Prometheus, Buffalo, 1981).
- Leikind, B. J. & McCarthy, W. J. Skeptical Inquirer 10, 23-34 (1985).
- Vogt, E. H. & Hyman, R. Waterwitching USA 2nd edn (Chicago University Press. 1979).
- Randi, J Skeptical Inquirer 8, 329-333 (1984).
- Martin. M. Skeptical Inquirer 8, 138-140 (1983).
- Randi, J. The Magic of Uri Geller (Ballantine, New York, 1975).
- Fuller, U. Confessions of a Psychic (Karl Fulves, Box 433, Teaneck, New Jersey, 1975).
- Fuller, U. Further Confessions of a Psychic (Karl Fulves, New Jersey, 1980),
- Marks, D. &Kammann. R. Zetetic 1(2),9-17 (1977).
- Hyman, R. Zeteric 1(2). 18-37 (1977).
- Soal, S. G. & Goldney, K. M. Proc.. Soc. psychical Res. 47, 21-150 (1943).
- Scott, C. & Haskell, P. Nature 245, 52-54 (1974).
- Marwick, B. Proc.. Soc. psychical Res.56, 250-281 (1978).
- Hoebens, P. H. Skeptical Inquirer 6, 32-40 (1981).
- De Mille, R. Castandeda’s Journey2nd ed (Capra. Santa Barbara, 1978).
- Skinner, B. F. Science and Human Behavior (Macmillan, New York, 1953),
- Bunge, M. Method, Model and Matter (Reidel, Dordrecht, 1973).
- Schmeidler, G. R. & McConnell, R. A. ESP and Personality Patterns (Yale University Press, 1958).
- Collins, H. M. & Pinch. T. J. Frames of Meaning: The Social Construction of Extraordinary Science (Routledge & Kegan Paul, London, 1982).
- Taylor, J. Superminds: An Inquiry into the Paranormal (Macmillan, London, 1975).
- Barratt, W. Proc.. Soc. psychical Res34, 275–297 (1924)..
- Rhine, J. B. The Reach of Mind, 209-214 (Sloane. New York, 1947).
- Tart, C. T. Psi: Scientific Studies of the Psychic Realm, vii-viii (Dutton. New York, 1977).
- Flew. A. in Science, Pseudo-Science and Society (edsHanen, M. P., Osler, M. J. & Weyant, R. G.) 55-75 (Wilfrid Laurier University Press, Waterloo, 1980).
- Cooper. D. E. in Philosophy and Psychical Research (ed. Thakur, S. C.) 59-80 (Allen & Unwin. London, 1976).
- Scriven, M. in Philosophy and Psychical Research (ed. Thakur, S. C.) 181-194 (Allen & Unwin. London. 1976).
- Wolman, B. J. (ed.) Handbook of Parapsychology (Van Nostrand. New York. 1977).
- Beloff, J. Zetetic Scholar 6, 90-94 (1980).
- Morris. R. L. J. Am. Soc. psychical Res. 74, 425-443 (1980).
- Akers, C. in Advances in Parapsychological Research Vol. 4 (ed. Krippner, S.) 112-164 (McFarland, Jefferson, North Carolina, 1984).
- Hyman, R. J. Parapsychol. 49, 3-49 (1985).
- Targ, R. & Puthoff, H. E. Nature 252, 602-607 (1974).
- Marks, D: F. & Kammann, R. Nature 274. 680-681 (1978).
- Hyman, R. Skeptical Inquirer 9, 125~ 145 (1984-5).
- Tart, C. T.. Puthoff, H. E. & Targ, R. Nature 284. 191 (1980).
- Marks, D. F. Skeptical Inquirer 6, 18-29 (1982).
- O’ Keefe, D. Stolen Lightning (Robertson, Oxford, 1982).
- Wilson, S. C. & Barber, T. X. in Imagery: Current Theory, Research and Application (ed. Sheikh, A. A.) 340-387 (Wiley. New York, 1983).
- Marks, D. & McKellar, P. J. mental imagery 6, 1-124 (1982).
- Gurney, E. &Myers, F. W. H. Proc. Soc. psychical Res. 5, 403-485 (1889).
- Sidgwick, H. Proc. Soc. psychical Res. 10, 25-422 (1894).
- Finucane, R. C. Appearances of the Dead (Prometheus, Buffalo. 1985).
- Zusne, L. & Jones, W. H. Anomalistic Psychology (Erlbaum, Hillsdale, New Jersey, 1982).
- Edwards, W. in Formal Representation of Human Judgement (ed. Kleinmuntz, B.) 17-52 (Wiley, New York, 1968).
- Nisbett, R. & Ross, L. Human Inference: Strategies and Shortcomings of Social Judgment (Prentice-Hall, Englewood Clills. 1980).
- Hoebens, P. H. Skeptical Inquirer 7, 38-45 (1982).
- Klass, P. J. Zetetic 2, 57-61 (1977).
- Koestler, A. The Roots of Coincidence (Hutchinson, London, 1972).
- Rhine, L E. J. Parapsychol. 15, 164-190 (1951).
- Tart, C. T. Learning to Use Extrasensory Perception (University of Chicago Press, 1976).
- Tart, C. T., Palmer, J. & Redington, D. J. J. Am. Soc. psychical Res. 73, 151-165 (1978).
- Dixon, N. F. Preconscious Processing (Wiley, Chichester, 1981 ).
- Franselle. F. & Bannister. D. A Manual for Repertory Grid Technique (Academic, London, 1977).
- Windholz, G. & Diamant. L. Bull. psycnon. Sci. 3, 125-126 (1974).