by Guy Lyon Playfair
Samuel G. Soal (1889-1975) was one of the highest-profile researchers of his day. His book Modern Experiments in Telepathy (1954) earned him a degree from London University – only the second such honour to be awarded in Britain for a parapsychology-related thesis – and the long series of card guessing tests he carried out with ‘star performer’ Basil Shackleton soon became regarded as unequalled, both for their strong positive results and the rigour of the controls involved. To many, Soal had proved telepathy to be real beyond any reasonable doubt, and for good measure it seemed he had proved precognition as well, for Shackleton was found to have a way-above-chance skill at guessing the next card to be viewed. Soal’s reputation as a meticulous researcher was rock-solid.
Until 1960, that is, when the first small crack appeared. It was made by one of Soal’s subjects, Mrs. Gretl Albert, who claimed to have spotted him on more than one occasion altering his score sheets after a test. Soal’s colleagues were horrified, and rushed to his defence in a series of letters to the Journal of the Society for Psychical Research (JSPR) Soal, of all people, cheating? Out of the question!
Yet the crack widened. In a posthumous article in the JSPR for March 1971, George Medhurst announced that his computer search for the source of the random numbers that Soal had claimed to use for deciding which of five picture cards to transmit had been unsuccessful. This was generally seen as a sign of carelessness rather than fakery, but two mathematicians, Christopher Scott and Philip Haskell, widened the crack even further by claiming to have found evidence suggesting that Mrs. Albert had been right.
However, as a young computer expert named Betty Markwick pointed out, their work was impressive but not conclusive, and ‘one longed for a conclusive settlement of the matter – either way’, as she put it, and after some ingenious mathematical detective work she reckoned she had found it. Soal’s random number sequences, she claimed in the JSPR (May 1978) had to have been altered, as is now generally admitted. The rock had come tumbling down.
An Element of ESP?
She also revealed that what had first got her interested in the matter was ‘a dream of a most intense quality’ she had in March 1971 in which George Medhuirst, who had just died and whom she had met only once, was urging her work on some mathematical problem or other. Just five days later, her copy of the JSPR arrived containing Medhurst’s posthumous article mentioned above, and she could see all too clearly what the problem was. ‘While shunning a survivalist interpretation,’ she wrote, ‘it was difficult to resist the feeling that an element of ESP might nevertheless be involved, impelling me to follow up certain ideas suggested by the dream.’
The Levy Affair
The question of cheating by researchers was discussed at length by the doyen of parapsychologists, J.B.Rhine, in the March 1974 issue of the Journal of Parapsychology. He revealed that there had been twelve cases of ‘experimenter unreliability’ in his own lab, nearly all of them apparently detected without much difficulty, and he suggested ways of ensuring that there would not be any more. Then, in the very next issue (June), he had to announce that there had been another one, the guilty party being his own assistant and heir apparent, Walter J. Levy. He had been caught red-handed by his own colleagues fiddling with a piece of automatic recording equipment in order to make the result of a rat experiment look better than it was. Rhine was immediately informed, and within minutes Levy was out of the door and out of the parapsychology community. None of his many published experiments, Rhine said, should be considered reliable unless they had been replicated by somebody else.
Sceptics refer to these two cases whenever possible, implying that if such high-level researchers as Soal and Levy were known to have cheated, then others probably did as well. Yet these are the only cases on record in which published experimental results are known to have been based on fake data. In the other cases mentioned by Rhine, at least eight were detected before the experiment concerned reached publication stage and even if any of the other four contained any unreliable evidence, this would not undermine Rhine’s overall success record. (See my appeal below).
Sargent – Not Guilty
Mention should be made of two other cases often cited in order to discredit parapsychology, in each case without justification. The first, from the early 1980s, involved Carl Sargent of Cambridge University, a pioneer in the use of the Ganzfeld procedure for experiments in telepathy and clairvoyance. A fellow researcher, Susan Blackmore (later to become a prominent sceptic) announced, after a good look at Sargent’s experimental setup, that she had found shortcomings, such as poorly selected target material and not enough of it, which would have made cheating possible. Yet though it is still occasionally alleged that she ‘caught him cheating’, she has assured me that she did not and never said she had, She merely showed, quite plausibly, that his protocol could have been tighter. Those were the early days of Ganzfeld work, and subsequent researchers have used much stricter controls and protocols while still obtaining results as positive as Sargent’s, some of them indeed more so.
Wirth – Guilty
The other case is very different indeed. The researcher concerned, Daniel P. Wirth, is a convicted criminal, sentenced in 2004 to five years in prison for a whole string of fraud and felony charges. He was author or co-author of twenty papers published between 1987 and 2001, chiefly in journals dealing with alternative and complementary medicine. He ran into trouble when serious doubts were cast on the report which he co-authored on the now notorious Columbia University ‘miracle study’ published in the peer-reviewed Journal of Reproductive Medicine (46, 781-7, 2001). This purported to show that distant prayer can help infertile women become pregnant, but it has been suggested, apparently with some justification, (Skeptical Inquirer, Sept./Oct. 2004, p.31) that ‘the study may never have been conducted at all’.
Although having a master’s degree in parapsychology from John F. Kennedy University, Wirth has never been considered to belong to the mainstream parapsychology community, his publications being mainly devoted to marginal areas of healing. One should naturally regard any work by a convicted swindler with suspicion, especially since efforts to locate several of Wirth’s co-authors have failed, suggesting that they may not exist. (Full details of the Wirth saga can be found here).
Betraying the Truth
Scientific fraud has been going on since at least the 2nd century BC when the Greek astronomer Hipparchus tried to pass off a Babylonian star chart as his own. Noted scientists who have resorted to data-fudging, plagiarism or outright invention, or who have been plausibly accused of them, include Ptolemy, Galileo, Bernoulli, Mendel and even Newton. More recently, physicist Robert Millikan is now known to have ‘selected’ the data that helped him win a Nobel prize (1923) although he claimed to have reported all his results.
This and many other scientific malpractices and scandals are recounted in William Broad and Nicholas Wade’s Betrayers of the Truth (1982) which contains dozens of cases from astronomy, biology, chemistry, geology and several areas of medical research, notably immunology, yet only one (Levy) from parapsychology. The authors appealed for any cases they had missed (I told them about Soal) and I do the same here.
Broad and Wade make an important point: ‘Because parapsychology is still regarded as a fringe subject not properly part of science, its practitioners have striven to be more than usually rigorous in following correct scientific methodology’. That was written 25 years ago, and they are even more rigorous today. They are also getting much better at catching fakers who try to deceive them. I once attended a lecture by a young magician who had hoped to persuade Edinburgh parapsychologists that he could bend spoons paranormally, but failed. ‘They were very good,’ he admitted ruefully.*
Such would-be impostors, of whom there have been several, may have done psi researchers a favour by forcing them to tighten up their controls against fraud, not only by putative spoon-benders but also by rogue researchers – and even their own colleagues.
*For the details, see the Journal of the Society for Psychical Research, October 1987, p. 247-56.