Book Review

Stealing into Print: Fraud, Plagiarism, and Misconduct in Scientific Publishing

N Engl J Med 1993; 328:1648June 3, 1993

Article

Stealing into Print: Fraud, Plagiarism, and Misconduct in Scientific Publishing
By Marcel C. Lafollette. 293 pp. Berkeley, Calif., University of California Press, 1992. $30. ISBN: 0-520-07831-4

Like a priest betraying a confession, a journalist revealing his or her source, or a doctor killing a patient, a scientist who commits fraud violates a central principle of the perpetrator's profession. Despite the intellectual enormity of the crime, there has been a steady trickle of cases over the past decade, yet the academic community has still not completely learned how to cope with the problem or devised meaningful prophylactic reforms.

Stealing into Print offers a clear and smoothly written description of many of these cases, combined with discussions of various institutional aspects, such as the role of journal editors, the duties of referees, and the fate of whistle-blowers. The author's practical introduction to this vexing issue came about when, as editor of the journal Science, Technology & Human Values, she joined the list of editors asked by Walter Stewart and Ned Feder of the National Institutes of Health to publish their manuscript on the Darsee affair. John Darsee was a personable and energetic young man who spread joy among academic cardiologists for a number of years by inviting them to be coauthors of his many papers. They were too busy to notice that the data were forged. Stewart and Feder's point -- that those accepting kudos should also accept responsibility -- was too hot for most editors to handle, and only Nature, despite England's harsh libel laws, had the courage to print their article.

In contrast is the often remarkable boldness of those who draw attention to fraud. In case after case, the whistle-blowers turn out to be young researchers, too principled to connive at a superior's legerdemain and naive enough to assume that university authorities will welcome the pointing out of a wrong. Almost invariably, however, the first response of universities is to investigate the whistle-blower, not the accused. Bruce Hollis, a whistle-blower at Case Western Reserve University, is quoted here with the bitter reflection: “I cannot recommend that junior scientists who discover scientific misconduct blow the whistle unless they want to experience immense personal suffering and a possible end to their scientific careers.”

Since universities have generally been so expert in suppressing charges of fraud, the press has had a necessary compensatory role. Though many scientists wish reporters had done less, Marcel LaFollette believes their record has been less than uniformly diligent. “With several notable exceptions,” she says, doubtless referring to Daniel Greenberg's writings in Science & Government Report, “if one looks closely at the role of the press in the most sensational cases, science journalists as a whole appear relatively tame.”

My chief criticism concerns the book's basic thesis, that fraud is an important problem for journal editors. It hardly seems cost effective for editors and referees to try and root out fraud, especially when the occasional paper with forged data is forgotten just as quickly as the vast bulk of published research articles.

Fraud is important, however, in ways the author does not explore. It is a gnawing public-relations problem for academic science. In addition, fraud is methodologically interesting in that it shows how easily the checking mechanisms of science can be bypassed. If the elaborate apparatus of peer review and referees cannot routinely detect fraud, how can we be confident that it detects honest error? I suspect we cannot, and that science is credible for quite different reasons.

Nicholas Wade
New York Times, New York, NY 10036