Flim-Flam! Psychics, ESP, Unicorns, and Other Delusions - James Randi [128]
Professor G. E. Hutchinson of Yale declared the system "the most carefully conducted investigations of the kind ever to have been made." Professor R. A. McConnell of the University of Pittsburgh said of one of Soal’s published works, "As a report to scientists, this is the most important book on parapsychology since... 1940... If scientists will read it carefully, the 'ESP controversy' will be ended." C. D. Broad, the philosopher, said that the work was "outstanding.... The precautions taken to prevent deliberate fraud [were] absolutely watertight." Even parapsychologist J. B. Rhine gave glowing approval of Soal's design and results. These results were truly fantastic—on the order of billions to one against mere chance. The accolades continued to come in from all corners of the globe. Referring to some of Soal's work, Sir Cyril Burt, the grand old man of British science, said, "It must, I think, be agreed that the Soal experiments are unrivaled in the whole corpus of psychological research." Sir Cyril was recently discovered to have faked extensive data himself in research on heredity and even to have invented witnesses and authorities for his reports. A prominent parapsychologist, Professor Beloff of the University of Edinburgh, called Soal's reports "The most impressive evidence we have for the reality of ESP." Of late, he has not repeated that opinion.
Not long after all this acclaim, it began to appear that there might have been some hanky-panky at work. An observer reported that she had seen Soal altering some l's into 4's and 5's—his l's were written very short and therefore were easily altered to produce the desired figures. When notified, Soal decided it was not "important enough" to report officially. But in 1973, when Christopher Scott and P. Haskell investigated, the case for Soal's deception was very strong. There were too few l's and too many 4's and 5's in the target numbers. Many of the 4's and 5's in the list turned out to be "hits" in Shackleton's tests. Apparently, when the target was a 1, and the subject called out "four," it was simple and tempting for Soal to "correct" the 1 to a 4. But—and it is a huge "but"—even with these digits accounted for, the results of the tests were far better than mere chance would have them, and so the tests, though shadowed, stood as the best example ever of ESP proof.
In Proceedings of the Society for Psychical Research, Betty Markwick, a statistician, revealed early in 1979 the damning facts about Soal that had not been suspected. Aside from changing a few digits when the opportunity arose, it seems he also cleverly managed another simple ruse. Markwick found—after much labor—the places in the logarithm tables where Soal had chosen his digits. Not only had he gotten unforgivably lazy and repeated some series on the lists without properly arriving at them; he also had left spaces in his target list every few digits, into which he inserted "winning" target digits as the tests were conducted. No one had thought to observe him, and in fact they could not, since according to the rules his list was supposed to be secret until presented for checking. But the evidence was there, for the "E.D.s" (extra digits) that had been discovered were "hits" that agreed with Shackleton's guesses. Suddenly, there was no longer any mystery about where these results had come from.
Soal was down and out for good, and the last bigshot in the business was discredited. But it remained to J. G. Pratt, a parapsychologist at the University of Virginia Medical Center, to provide the most astonishing bit of rationalization in defense of Soal that has ever been heard in this field—a field long famous for its Catch-22s and superb alibis. Although Pratt admitted that he "must put all of this work aside marked to go to the dump heap," he could not refrain from the ingrained tendency to excuse the obvious peccadilloes of his former colleague. The work of Miss Markwick, said Pratt,
does not provide an unambiguous