Benveniste’s “high-dilution” experiments are not the first to raise concern about science journals’ proper response to unconventional results. Twice before, Nature published papers dubious enough to warrant accompanying editorials questioning the results. And in one eerily parallel precursor incident, Nature’s then editor actually swooped down on a yet another Paris lab with “The Amazing” Randi and a third party to debunk unorthodox results—and decided not to publish. How do the sharp eyes of hindsight view these decisions now?
The first case was a paper from Hungarian-born neurobiologist Georges Ungar, then working at Baylor College of Medicine, on an alleged memory molecule. According the report, which was co-authored by D.M. Desiderio and W Parr, a 15-amino acid peptide was extracted from mice trained to choose a light chamber over a dark one. When the substance, dubbed scotophobin, was injected into untrained mice, these new rodents made the same choice (Nature, volume 238, page 198, 1972).
After Ungar submitted the report in February, 1971, Nature asked him to extend it considerably to include details of analytical procedures. But the research and its conclusions remained contentious, so Nature editor John Maddox decided to run extensive criticisms of the work along with the paper. Ungar, much like Benveniste in the current flap, was infuriated by the journal’s treatment. “When the article finally came out in July 1972, it was followed by the referee’s comments, almost three times longer than our paper,” Ungar recalled five years later in Discovery Processes in Modern Biology (edited by WR. Klemm, published by Robert E. Krieger, 1977). “We were allowed a short reply that had to be put together in five days. Judging by the remarks that came to our attention, this unusual procedure had the unexpected effect of gaining supporters to our side. The referee’s bias was so glaring that it blunted the edge of his criticisms to all but the most obdurate opponents of our approach.”
Yet even with the “sympathy” support generated by Nature’s harsh treatment of the paper, Ungar’s unusual claims were unable to withstand the hard scrutiny of science. As Maddox says now, “Nothing much has been heard of scotophobin since that time.”
The second case was in October 1974. Nature carried a paper in which Russell Targ and Harold Puthoff of Stanford Research Institute, Menlo Park Calif., claimed that Uri Gefler had psychic powers. While locked up in an electronically shielded room, Geller purportedly reproduced target pictures drawn by experimenters at remote locations.
As with Benveniste’s work, one reason why Nature accepted the paper was that it, as the accompanying editorial explained, “presented as a scientific document by two qualified scientists from a major research establishment appar- ently with the unqualified backing of the research institute itself." Moreover, there had also been “very considerable advance publicity.”
Although one referee opposed publication, the two others agreed that the paper should be published “because it would allow parapsychologists, and other scientists interested in researching this arguable field, to gauge the quality of the Stanford research and assess how much it is contributing to parapsychology.” In addition, Nature’s then editor, David Davies (John Maddox left in 1973 and took over again in 1980), argued that “our readers expect us to be a home for the occasional high-risk type of paper.”
Did the publication of the paper in such a prestigious journal (Nature, volume 251, page 602) lend unwarranted credibility to the questionable claims of parapsychology? In some respects, the answer is “yes.”
Certainly, parapsychologists have had a field day ever since, frequently announcing that the paper heralded their acceptance by mainstream science. Twelve years later, in The Geller Effect (Jonathan Cape, 1986), for example, Geller and co-autbor Guy Lyon Playfair refer six times to Targ and Puthoff's report in Nature. Released this year in paperback, the book is now receiving wide distribution. here, Geller writes, “I find it quite amusing that those same critics who reject what they read in Nature will accept the second-hand libels they read in the popular press.” Of course, Geller and many other citers of the Targ and Puthoff paper fail to mention the damning editorial in the same issue of Nature (page 559).
But if the Nature article has fanned the flames of the paranormal among the general public, it has had little impact in the scientific community (see graph). “Whatever Geller may believe, the fact is that the appearance of the SRI paper was not an authentication of the work,” says Davis, who is now director of the Houndhill Trust, an educational foundation. “What has Geller done since? He has made no headway whatever in the scientific community. The whole thing has been forgotten and buried.”
Given his experience with Targ, Puthoff, and Geller, how would Davies have handled the Benveniste affair? “I’m sorry I am not going to tell John Maddox how to run his journal, any more than he told me at the time,” Davies says. “But to be honest I must say that if I were running Nature now, and if I had had the benefit of Randi’s expertise and known what I know now about Geller, I would not have published the Targ/Puthoff paper. Indeed, we were under a great deal of pressure about 10 years ago to run a report concerning a French mental bender called Ranky. Just as John has done this time, I decided to investigate at first hand. Randi Chris Evans [a late artificial intelligence researcher at the National Physical Laboratory], and I went to Paris, had several meetings, evaluated Ranky, concluded that he was doing nothing paranormal, and decided not to publish anything.” If there is a lesson from these tales, it may be that journals should actively investigate bizarre claims before deciding to publish. In some cases, as with Ungar, editors may be justified in believing that erroneous work will quickly wither under the stern gaze of science. But in others, the general public may still view the paper’s publication as an official imprimatur years after the fact.