The Soal-Goldney Shackleton Experiments

S. G. Soal was a mathematician at Queen Mary College who came to psychical research as a sceptic, ran years of card tests without result, and said so publicly. He was persuaded to re-examine his old records for displacement, that is, for hits on the card before or after the target rather than the target itself, and reported finding it in the data of two subjects.

One was Basil Shackleton, a photographer, and from 1941 Soal ran a long new series with him in London, with K. M. Goldney as co-experimenter. The reported result was extraordinary: Shackleton scored not on the card the agent was looking at but on the one the agent would look at next, at odds against chance so large that the figures ceased to be meaningful as probabilities. The series was conducted in wartime, with witnesses invited from outside, and it was published and defended by the Society for Psychical Research. For three decades it was cited as the answer to the objection that psi results came only from sloppy laboratories, because here was a hostile mathematician who had produced them under conditions he himself had designed. Displacement scoring is worth pausing on, because it is the feature that later made the results checkable. A subject scoring on the next card rather than the target generates a very specific pattern in the record, and a pattern is something an auditor can look for thirty years afterwards.

Suspicion had circulated for years, including a claim by one witness that she had seen Soal altering figures, which he denied and which could not be substantiated. What settled the matter was arithmetic on the surviving records. Betty Markwick, working in the 1970s, set out to test whether Soal had generated his target sequences by the method he described, and searched the sheets for repeated runs of digits. She found them, and she found something more damaging: where a sequence recurred, it recurred with extra digits inserted, and the inserted digits fell overwhelmingly at the positions where Shackleton had been credited with a hit. That is not a pattern any honest procedure produces. Her paper appeared in 1978 in the Proceedings of the Society for Psychical Research, the same body that had published and defended the original work, and the field accepted the conclusion.

Two things about this episode deserve emphasis. The exposure came from inside parapsychology and was published by it, which is a point in the field's favour and is rarely credited. And it was only possible because the original score sheets had been kept, thirty-five years after the sessions. A summary statistic cannot be audited this way, and this case is the standing argument for archiving raw data.