Never make up your mind about someone's work until you've heard them under fire on a platform," an old university mentor, Alan Emslie-Smith, said to me many years ago. By stressing the importance of seeing scientists in the flesh, he was not criticizing the learned journals, their editors or their refereeing procedures. He was simply suggesting that the intuitive judgments we all make when reacting to politicians and automobile salesmen were equally appropriate in reacting to physicists and microbiologists. Personal impressions, he insisted, were often more reliable than the printed word.
Alan didn't even mention the possibilities of cheating, "data massage," and wholesale fabrication of laboratory results, about which we have heard so much in recent years. He did impress upon me the danger that we may sometimes be misled by the apparent exactitude and rationality of the scientific literature. The time-honored format of methods, results and discussion can indeed conceal much that is relevant to the proper evaluation of a piece of research. In 1974, for example, when two Stanford Research Institute researchers reported their singular investigations into entertainer Uri Gellen (Nature, vol. 251, p. 602), New Scientist (vol. 64, p. 170) was able to reveal that one of them was extremely nearsighted and the other a Scientologist. Such information has no place in a research paper-perhaps quite properly. But it could hardly be considered irrelevant on that occasion.
Emslie-Smith's warnings came to mind as I read the latest disclosure abut a researcher who has confessed to having published fictional "findings." As reported in The Sunday Times on September 20, 1986, p. 27, Professor Michael Briggs, working most recently at Deakin University, Australia, but now living on Spain's Costa del Sol, has admitted that he invented a considerable quantity of data on the relationship between the contraceptive pill and cardiovascular disease in women. He told the London newspaper that he had simply collected figures from other studies and generalized them into apparently large and convincing trials.
Yet Brigg's work had attracted skepticism and disbelief as early as 1982. In the conference corridors-but not in public or in print-experts were questioning how he could possibly organize and complete his huge studies in so little time. He appeared able to recruit suitable women into trials far more quickly than anyone else. There were also blatant internal contradictions in some of his papers (carried mostly in symposium volumes rather than in refereed journals), and he recorded sophisticated analyses for which Deakin University did not even possess the requisite equipment.
Eventually, complaints were laid, Briggs exploited a procedural point to block a formal investigation, and he left Deakin a year ago accompanied by a statement from the chancellor that nothing adverse should be concluded from his premature retirement. Some months later, a cabal of experts meeting in Berlin concluded that they should no longer use his research "findings."
It's a sad story, of the sort that would be unwelcome to any profession. That said, the Briggs affair is another reminder of the accuracy of the claim by William Broad and Nicholas Wade in Betrayers of the Truth (Simon and Schuster, 1982), that fraud is "a small, but not insignificant, endemic feature of the scientific enterprise." One does not have to accept their more exaggerated estimate that every such revelation is accompanied by another 100,000 undisclosed cheats buried in the literature to take their side against those reviewers who dismissed their book as a figment of journalistic imagination. It would be naive to expect that science should differ from other crafts in not including its aliquot of bent practitioners.
The Briggs affair does, however, prompt again the question of why colleagues and peers of the scientific community's bad apples are often remarkably slow to take tangible action. Not infrequently, they suspect that so-and-so's work is dodgy long before the revelation or confession of dishonesty. Correction: they know that there is something seriously amiss. The work smells. It cannot be independently verified, and the individuals concerned seem evasive and shifty in discussion. Yet, apart from raising eyebrows and gossiping in conference bars, they do little to bring the offender to book.
Why? Could it be that scientists themselves are so preoccupied by the rationality and honesty upon which their distinguished craft depends that, even in the face of compelling evidence of chicanery, they cannot quite believe what is happening when someone deviates from those uniquely high standards? They should, I suggest, put more trust in those gut instincts that serve them so well during their lives outside the laboratory, library and conference hail.
A microbiologist, Dixon is European editor of THE SCIENTIST.