The following is an abridged version of a paper presented at Skeptics 2000, Dunedin, New Zealand. The author would like to thank NZCSICOP and NZARH for sponsoring this visit to New Zealand.
In this paper I discuss some major categories of “exceptional” or anomalous experiences and how to interpret them. I am particularly interested in the kinds of experiences studied by parapsychologists (ie ESP, PK, and precognition). An important assumption is that the description of an exceptional experience needs to be kept separate from any of its possible interpretations. The brain/mind is structured in such a way that exceptional experiences and their interpretations easily become conflated.
Interpretations of exceptional experience take principally one of two forms: normal or “N” theory accounts (NIEs) or a paranormal or “P” theory accounts (PIEs). Analysis of lay accounts of exceptional experience suggests an acceptance of and belief in paranormal claims ahead of the evidence that is necessary to warrant these claims.
Evidence supporting PIEs appears to be conflicting, inconsistent and, in other respects, seriously wanting. I argue that PIEs often result from eight special circumstances that pass unnoticed by the original investigators or PIE claimants. These are methodological flaws, sensory cues, imperfect randomisation, selection of best cases, subjective validation, inappropriate statistics, and outright deception, trickery or fraud.
The issue will be illustrated with seven specific examples.
1. Remote viewing ability
(1A) SRI series
This series of experiments on a form of ESP called remote viewing ran from 1972 to 1985 under the direction of Dr Harold Puthoff and Russell Targ at the Stanford Research Institute (SRI), and created a huge international interest in the paranormal. Unfortunately, the studies were very badly flawed.
Science separates itself from pseudo-science along a number of dimensions. One of these is accessibility of the data. Following publication of interesting and significant observations it is an accepted scientific practice for researchers to allow colleagues who are doing serious research in the same field to have access to their original data. When researchers consistently refuse to allow colleagues such access, something important is being signalled. Of course data may get lost or destroyed or be difficult or costly to retrieve in the form required. Or they may be classified information or have commercial value that a scientist may wish to exploit prior to their general release.
However, when none of these considerations is applicable, refusal to supply a copy of a data-set leads to the unpleasant inference that something is wrong, that the data do not support what is claimed for them, that the data are an embarrassment following an extravagant claim that cannot be substantiated.
During the 1970s I made frequent requests to Puthoff and Targ for copies of their remote viewing transcripts. Targ and Puthoff consistently refused to supply this information, as they did to others I know who have made this request. Their only concession was to supply me with a single transcript from the Price series (Experiment 7) published in Mind-Reach.
Following publication of The Psychology of the Psychic, I conducted a “remote judging” exercise with the Hammid series of remote viewing experiments. The three remote judges had access only to the cues provided in the Hammid transcripts together with the target lists and the map provided to the SRI judge. No site visits were possible and none of the descriptive material from the SRI transcripts was available. A total of 24 cues was found in six Hammid transcripts.
The claim that the Hammid target list given to Arthur Hastings was randomised is also doubtful. It actually depends on which list one is talking about, because, although the SRI researchers were unwilling to admit this, no less than three listings of targets in the Hammid series were given to the judge. “I received three target lists” (letter from Hastings, May 26, 1977). One of these lists was randomised; this is the one cited by Puthoff as the (implying only) target list given to the judge. The other two lists provided by SRI (described in detail by Hastings in his letter to me of May 26, 1977) were not random.
Contrary to Targ and Puthoff’s claims, the quality of the subjects’ descriptions is extremely poor and, without the cues, cannot be matched against the targets. Had cue-less matching been possible, T and P would not have needed to leave these obvious clues in the transcripts in the first place.
Tart, Puthoff and Targ (1980) claimed to have conducted a re-judging of the Price series with all of the cues removed from the transcripts. After asking Puthoff for a set of the edited transcripts on multiple occasions over three years, they were eventually sent to a colleague Dr Chris Scott. On inspecting these supposedly “edited” transcripts, it was readily apparent that many obvious cues were still present (Marks & Scott, 1986). The little remaining credibility for Targ and Puthoff had now been absolutely and irrevocably destroyed.
(1B) SAIC or Star Gate series.
This series of remote viewing studies ran from 1985 until 1995 and was directed by Dr Edwin May. The Star Gate RV series consisted of three projects:
- “Operations” using remote viewers stationed at Fort Meade, Maryland, to collect intelligence;
- “Research and Development”, the laboratory research conducted at SRI and later at SAIC;
- “Foreign Assessment” focused on gathering intelligence on what potential enemies were doing in the area of parapsychology.
Much of the information about Operations and Foreign Assessment remains classified although many ex-military remote viewers have established private businesses that offer remote viewing services. They are responsible for a lot of hype in the media, in their books and on their WebPages.
The CIA carried out a review of the SAIC program in 1995 which was aimed at determining: (a) whether Star Gate had any long-term practical value for the intelligence community, and (b) if it did, what changes should be made to enhance the value of remote viewing research. The review reached the following conclusions:
- “A statistically significant effect has been observed in the recent laboratory experiments of remote viewing. However, the existence of a statistically significant effect did not lead both reviewers to the conclusion that this research program has provided an unequivocal demonstration that remote viewing exists. A statistically significant effect might result either from the existence of the phenomenon, or, alternatively, to methodological artefacts or other alternative explanations for the observed effects.”
- “The experimental research conducted as part of the current program does not unambiguously support the interpretation of the results in terms of a paranormal phenomenon.”
The principal reason for this conclusion is that only one judge, who happened to be the Principal Investigator, was used in assessing matches throughout these experimental studies. The CIA report concluded:
“As a consequence, there is no evidence for agreement across independent judges as to the accuracy of the remote viewings. Failure to provide evidence that independent judges arrive at similar conclusions makes it difficult to unambiguously determine whether the observed effects can be attributed to the remote viewers’ (paranormal) ability, to the ability of the judge to interpret ambiguous information, or to the combination or interaction of the viewers and the judge. Furthermore, given the Principal Investigator’s familiarity with the viewers, the target set, and the experimental procedures, it is possible that subtle, unintentional factors may have influenced the results obtained in these studies.”
2. Ganzfeld ESP ability
This field is a core part of the parapsychology literature and has been intensively researched in at least ten different laboratories in the USA and Europe for about the same length of time as remote viewing. The ganzfeld research has gained an almost symbolic importance for the parapsychology field as the best case for the existence of psi. The results obtained in the ganzfeld allegedly appeared in multiple studies by different investigators with subjects who were not especially selected or gifted as psychics. It is the main focus for discussions of parapsychology in a leading psychology journal (Psychological Bulletin) and it has increasingly been viewed as a genuine and uncontroversial finding. Until now, that is.
Ganzfeld experiments involve two participants, a sender and a receiver, located in separate rooms. The receiver is in a ganzfeld that is usually created by wearing translucent Ping-Pong ball halves taped over the eyes and a red floodlight directed towards the eyes, producing an undifferentiated visual field. White noise is played through headphones. To reduce internal “noise” in the form of somatic sensations, the receiver may be taken through a series of progressive relaxation exercises. The term “ganzfeld”, a German word meaning “total field”, refers to the mild sensory habituation created by the environment described.
The sender is shown a target picture or video clip that has been randomly selected from a large pool of possible targets. The sender is asked to try to send the information about the picture or video to the receiver by psychic means (telepathy). The receiver is asked to receive this information and to report any images, thoughts or feelings that occur during the trial. The receiver is then given a randomly ordered set of four stimuli, the target plus three decoys. If the receiver chooses the correct target it is recorded as a hit. The mean chance expectation (MCE) is 25 percent: a statistically significant deviation above MCE is suggestive of an anomalous effect consistent with psi.
The ganzfeld research is unique in parapsychology for the way in which believers and sceptics have worked together to agree a protocol for properly controlled investigation (Hyman & Honorton, 1986). This resulted from the leadership of two principal figures, the late Charles Honorton and Ray Hyman. One of the progressive outcomes of Hyman and Honorton’s joint communiqué was a set of guidelines concerning methodology. Hopefully these guidelines for ganzfeld research were implemented and the quality of studies improved as a consequence.
Not only should the methodology have improved beyond the earlier studies, which were heavily flawed (Hyman, 1985), but the results should not be dependent on a small band of investigators: when a small number of well-insulated people and organisations are responsible for a program, the research can go badly off the rails.
Milton and Wiseman (1999) systematically reviewed the ganzfeld literature since the Hyman/Honorton joint communiqué, 1987 to February 1997. They found 30 studies in 14 papers by 10 principal authors from seven laboratories; the database included 1,198 ganzfeld trials. They calculated a probability score across all 30 studies of only .24, meaning that the ganzfeld effect was so small it did not differ statistically from chance. Milton and Wiseman’s study, published in the prestigious Psychological Bulletin, has already created a few waves in the parapsychology pond.
As if this result was not damaging enough, the investigators went on to critically examine three of five other claims made by Bem and Honorton (1994) and to determine their methodological rigour. What they found is even more startling. The new ganzfeld studies, 1987-97, examined three out of five variables that Bem and Honorton had suggested were statistically related to high psi scoring rates in the autoganzfeld studies. These three variables were:
- Trials with dynamic targets (videos) had been more successful than trials with static targets.
- Novices who reported prior psi experiences in everyday life scored more highly than those who did not.
- Novices who reported studying a mental discipline such as meditation or yoga scored more highly than people who did not.
Milton and Wiseman’s analysis found only one of these variables to be significant of the new studies (number 2). However when they looked into the original studies examined by Bem and Honorton (1994) they discovered that there appeared to be no good evidence to support this in the first place. The two papers in which this ‘novices with mental training’ effect was allegedly found contained one non-significant effect (Honorton & Schecter, 1986) and one reversed effect (Honorton, 1997). The same problem appeared in respect of another of Bem and Honorton’s “discoveries” about psi, that high scoring novices were also high on Feeling and Perception on the Myers-Briggs Type Indicator (Briggs & Myers, 1957). This was another “discovery” that appeared in the first study and then disappeared in the second.
Milton (1999) extended the meta-analysis to included nine further well-controlled ganzfeld studies run between February 1997 to March 1999. This longer series of 39 studies managed to reach significance (p= .011) but, when a single highly significant study by Dalton (1997) was excluded, the Stouffer z score was only 1.45 and p=.074 (not significant). Thus, there is no evidence of a consistently replicable ganzfeld effect across a 12-year period of well-controlled ganzfeld research. One or two strikingly significant studies appear now and then but this does not constitute replication. The occurrence of one highly significant study in a general run of non-significant studies in fact sets alarm bells ringing. Dalton’s (1997) study with an effect size that is significantly higher than the general run suggests the need for an in-depth investigation of Dalton’s protocols.
In addition to the non-significant meta-analyses, criticisms of studies included by Bem and Honorton have also arisen. These are:
- Inadequate randomisation of targets and judging sets (Hyman, 1994)
- Possible sensory leakage of target information (Wiseman et al, 1996)
- Lack of replication (Milton & Wiseman, 1999)
3. Ability to detect unseen staring
Rupert Sheldrake (1994) proposed An Alice Through the Looking Glass vision of things that might be so but probably are not. He advocates the collective participation of non-scientists who have the “freedom to explore new areas of research”, and promotes a radically new theory of perception. We do not see images of things inside our brains, he maintains, the images may be outside us: “Vision may involve a two-way process, an inward movement of light and an outward projection of mental images.”
This process of outward projection has some interesting implications. If our minds reach out and “touch” things, then we may affect what we look at. For example, when we stare at somebody from behind s/he may be able to feel that we are staring on the back of his/her neck. Titchener (1898) described the feeling as “a state of unpleasant tingling, which gathers in volume and intensity until a movement which shall relieve it becomes inevitable” (p. 895). Colwell, Schroeder and Sladen (2000) have recently reviewed the literature on psychic staring and carried out some empirical tests.
The idea that “unseen” staring can be detected has been supported in the following research with incidence rates as high as 68-86% (Coover, 1913), 74% (Williams, 1983) and 92% (Braud, Shafer & Andrews, 1993). Titchener rejected the idea that the staring effect was based on telepathy and suggested the hypothesis that the eye is attracted to movement and the starer’s gaze is therefore attracted to the staree’s head turning in his direction.
Sheldrake (1994) has conducted new experiments on the staring phenomenon and encouraged school children and other members of the public to participate in his research program. Experimental kits can be downloaded from the New Scientist Web Site including an interesting list of 24 “random” sequences for use in experimental trials. Sheldrake suggests that each child in a group is tested with a different sequence or use sequences determined by tosses of a coin. The results are being compiled by Sheldrake into a pooled data set.
A colleague, John Colwell, decided to put the Sheldrake findings to rigorous test under controlled laboratory conditions (Colwell et al, 2000). On the basis of Sheldrake’s observations, it was decided to investigate the staring effect both with and without feedback. Colwell’s team carried out two experiments. The results of the first experiment suggested that the subjects in the staring research are able to score above chance as a consequence of being able to learn the non-random patterns in the sequences using the feedback. The tendency of the participants to show negative recency by avoiding multiple repetitions was well matched by Sheldrake’s sequences that showed exactly the same property. The fact that starees can guess when staring is occurring at above chance levels therefore demonstrates nothing other than an ability to notice patterns. This is a low-level ability that even a mouse could manage.
John Colwell and his team repeated the experiment using 10 properly randomised sequences taken from random number tables instead of Sheldrake’s non-random sets. The results support the hypothesis that the improvement in accuracy during staring episodes observed in Experiment One was due to pattern learning. When no feedback was provided and pattern learning was blocked, no ability to detect staring was observed and also no learning.
4. Pets’ ESP ability
Are animals psychic? Since time immemorial human beings have attributed supernatural powers to animals. The latest example of a long tradition of paranormal claims on behalf of our animal friends is the “psychic pet”. For example, a pet dog is claimed to be able to use psychic powers to detect when its owner is returning home. This has been the subject of Sheldrake’s (1999) Dogs that Know When Their Owners Are Coming Home. Sheldrake believes that a dog called Jaytee uses its “sixth sense” of telepathy to determine its owner Pamela Smart’s decision to return home.
According to Sheldrake, many pet owners claim the ability in their pets to know when a member of the household is about to come home. The dog goes and waits for the owner at a door or window, in a driveway, or even at a bus stop.
Sheldrake claims that on 100 different occasions between May 1994 and February 1995 when Pamela left Jaytee with her parents and went out, 85 times Jaytee reacted by going to the French window before Pamela returned, usually at least 10 minutes in advance of Pamela’s decision to set off for home. The anticipatory behaviour occurred regardless of distance or vehicle used. However, as Sheldrake acknowledges, the “anticipatory signalling” behaviour of Jaytee could have been cued by the expectations of Pam’s parents William and Muriel Smart as they consciously or unconsciously cued the dog that Pam would be home soon. It was necessary to conduct trials in which Pam set off for home at randomly selected times that were unknown to William and Muriel.
The story took an interesting twist in 1995 when Rupert Sheldrake (RS) invited Richard Wiseman to investigate Jaytee (Wiseman, Smith and Milton, 1998). Richard Wiseman’s team proposed eight normal explanations for the “psychic pet” phenomenon that controlled studies would need to take into account, including response to routine, sensory cueing, and selective memory.
Wiseman conducted four studies with the full co-operation of Pamela Smart and Sheldrake, based on the above safeguards and precautions. In none of these studies did Jaytee detect accurately when PS set off to return home. If this pet dog had any psychic ability at all, it did not appear in this study.
Rupert Sheldrake (1999) reports a series of observations carried out in a pre-planned series of 12 “experiments” in which Jaytee’s behaviour was recorded throughout Pam’s absences on time-recorded videotape. In these trials Pam came home at randomly selected times that were not known to her in advance. A third party (usually Sheldrake) selected the return time and bleeped Pam on a pager.
The resulting observations were analysed in two ways. First, by plotting the percentage of time that JT spent by the window for three periods:
- 1. The first ten minutes following the bleep: 55%
- The 10-minute period prior to PS’s return: 23%
- The main period when PS was absent prior to the pre-return period that varied between 110 and 150 minutes: 4%.
It would be a common sense interpretation of Sheldrake’s data to assume that JT could learn the timing of PS’s returns. This is exactly what happens. The results of Sheldrake’s tests are therefore not convincing. Why Sheldrake chose to use a pre-arranged bleep period that started between 80 and 170 minutes after PS had left is unclear. This restricted range for the bleep means that the return is more predictable. John Colwell estimated the return periods following the bleep by examining Sheldrake’s plots of the data. He found that PS always returned within a period of 110-200 minutes following her departure, 10 of the returns (83%) occurring during a 40-minute period between 120 and 160 minutes after departure. This means that JT may have learned when PS could be expected home and signalled accordingly. This hypothesis assumes no psychic powers, only the power of memory. The procedures used by Wiseman had allowed the return to occur at any time following PS’s departure. This procedure stopped JT from learning a simple routine based on timing similar to the situation that pertained when PS had followed a daily routine of reliable departures and returns.
5. Uri Geller’s ESP ability
In turning to claims 5-7 by self-proclaimed psychic Uri Geller, the fact that his claims are unproven is so well known among parapsychologists that we do not need to dwell on them for very long. However they need to be discussed because Geller is still actively involved as a paranormalist and many lay people appear to believe that his claims are genuine. There is only one peer-reviewed article on his ESP abilities; this is the same paper that reported the remote viewing studies carried out at the Stanford Research Institute by Russell Targ and Harold Puthoff (1974).
Like the remote viewing studies, Targ and Puthoff’s studies with Geller were very badly flawed. As in the case of remote viewing, however, the flaws were not apparent until in-depth investigations could be carried out into the experimental conditions. Following visits to the SRI laboratory Dick Kammann and I were able to reveal that the investigators Targ and Puthoff had left Geller unobserved in a chamber which had a hole in the wall stuffed only with cotton wool. An intercom was also available for Geller to listen to the investigators as they chose the targets and produced the stimulus drawings. At least one of the drawings (a bunch of grapes) was placed on a wall in the adjoining room opposite the access hole.
There has never been a single replication of this study. Repeated attempts by the author and others to persuade Geller to participate in laboratory experiments have been rejected. Field observations of Geller’s performance have revealed his use of sensory cues available from signals sent by accomplices.
6. Uri Geller’s PK ability
Geller’s claim to be able to bend metal without touching it is well known. However in spite of his claims, there are no peer-reviewed reports of his alleged PK ability (metal bending, influencing objects at a distance). Field observation shows that Geller can only bend metal in his hands.
7. Uri Geller’s clairvoyant ability
A set of 10 trials of a clairvoyance experiment was carried at SRI using a dice in a locked box (Targ & Puthoff, 1974). This study was also seriously flawed because Geller had the time and opportunity to flip open the lid of the box and see the target information. Again there is a lack of replication and field observations of his alleged clairvoyance ability suggest the use of normal perceptual-motor abilities in the form of signals from an accomplice and other means.
The genesis of P Theory
In the seven examples of Exceptional Experience discussed above, a NIE has proved to be a perfectly adequate explanation making any form of PIE redundant or superfluous. This rather pessimistic conclusion about the validity of PIEs is not purely a negative exercise however. Psychological and statistical studies of Exceptional Experience have yielded an interesting account of how the everyday operation of the processes of attention, perception and decision making promote PIE-thinking even when the alternative and more rational NIE-thinking can do perfectly well with the same experiential data. These analyses have revealed processes that make the genesis and high prevalence of PIEs understandable from a psychological viewpoint.
This is a powerful effect of belief and selective attention. Subjective validation occurs when support for one’s beliefs is found in a piece of evidence independently of any objective support. This process is also known under the term, “confirmation bias”.
Coincidences as “odd matches”.
There is a compelling and widespread tendency to believe that coincidences cannot occur purely by chance. An “odd match” is an association between two events that appears to lack a causal explanation. In The Psychology of the Psychic (1st ed.) Kammann and I referred to the belief that such odd matches cannot arise by chance as Koestler’s Fallacy after the most famous of its proponents. In fact, odd matches can and do occur by chance. “One-in-a-million” odd matches occur with a probability of precisely one in a million. The problem is that you and I are unaware of the million-minus-one combinations that do not strike us as vivid odd matches.
Assume that at the end of an ordinary day a person can recall 100 distinct events. This gives 4950 pairs of events. In 10 years and 1000 people we have 18 billion pairs of events. This generates 18,000 “one-in-a-million” events, some of which will be very striking.
The numbers of “one-in-a-million” experiences over the entire human population become impressively large. From the statistical viewpoint, that these experiences happen is inevitable. From a psychological viewpoint, it is equally as inevitable that the individuals concerned will have difficulty dealing with the experience without a fatalistic or paranormal interpretation. If a few exceptional experiences inspire their authors to write about them in their full paranormal regalia (e.g. Koestler, 1972) we have discovered the genesis of parapsychology itself. The reason parapsychologists continue to work in their chosen field is not the often disappointing results they obtain from their formal studies but their compelling personal experiences that have a PIE attached.
References available on request from the editor