The venerable Nature, which published the paper in its January 15 issue, has once again served science well. The publication process was certainly protracted; various versions of the paper have been under consideration there and elsewhere off and on since 1983. But drawn-out negotiations and repeated reviews resulted finally in an article presumably purged of potential grounds for libel suits. It was accompanied by a thoughtful two-page editorial and a commentary by Eugene Braunwald, one of Darsee's Harvard co-authors, whose attorney had been among the most importunate in the array of legal talent that came close to preventing publication altogether.
Darsee, you'll remember, was the Harvard postdoc who was caught making up data in 1981. Vast amounts of it, as it turned out, in scores of papers he published while at Harvard and at Emory University in Atlanta. Until its publication, the Stewart-Feder paper may well have been the best-known "in press" study ever. As the Nature editorial observed, "the document by Stewart and Feder has by now acquired a notoriety comparable with the Darsee-affair itself."
Two things made it such a hot topic. First it focused attention, not on Darsee's fraud, but on whether Darsee's papers presented such obvious problems that at least some of his 47 coauthors should have spotted something fishy. This implied accusation raised the ire of those coauthors, thus contributing handsomely to the coffers of a number of law firms.
Second was the fact that Stewart and Feder, exasperated by the unwillingness of several journals to publish their findings, took their case public. They testified before Congress. They sent copies of the paper to everyone who might matter, including all members of the National Academy of Sciences. They gave interviews to the press, from The New York Times to The Scientist. They even appeared on talk shows. This tactic, violating as it does the unwritten canons of science, annoyed a great many people. The Nature editorial, which on the whole was friendly, still chided Stewart and Feder for behaving "injudiciously."
One cannot defend pre-publication publicity as a satisfactory practice in general; it is too subject to abuse. The brief history of biotechnology, for example, is littered with "breakthroughs" that burst upon the world at press conferences instead of in refereed journals, and whose performance was about as long-lived as the copies of the newspapers that reported them. But the Stewart-Feder case may well be an exception. It is a reasonable question whether their paper ever would have been published had they been good boys and not made such a fuss.
The issue that they raise—"the co-author problem"—is not going to go away. Elsewhere in these pages (see p. 10) my colleague Bernard Dixon takes an amused approach to the proliferating co-author syndrome, but he make a serious point and charges journal editors with responsibility for doing something about it.
Does the responsibility lie with journal editors? The editors disagree among themselves, as The Scientist recently reported (January 12, 1987, p. 1). Does it lie elsewhere, with academic departments or university deans or funding agencies or even, perish the thought, regulatory bodies? Who knows, and that's part of the problem. The topic has fallen between the stools. Despite clear need, no coherent, consistent approach to defining the co-author role, and co-author responsibilities, yet exists. Bodies such as the Institute of Medicine are thinking about thinking about co-authorship. But it's clear there will be no resolution soon.
If people do not agree on the solutions, at least they mostly agree on the cause. The simple, decent impulse to acknowledge the help of a colleague has been transmuted into a desperate grab at any smidgen of credit that will help advance a career in these competitive times. Liveithoods and lives depend on the size of a CV.
It is not hard to get scientists to agree in principle that what matters is quality of work. But that's lip service. In practice, what mostly counts is quantity, and everyone knows it. The publish-or-perish mentality is central not only to the co-author problem, but to other problems that reduce the utility of the scientific literature. These include outright fraud, but also much more frequent difficulties that range from honest error to endemic unintelligibility.
Calls to do something about the pressure to publish have been heard for some time. It is one of the solutions posed by Nature's editorial, which champions "decoupling the literature from promotion prospects."
A worthy goal. But is it a realistic one? Is there an individual or group that can wave a magic wand and make the decoupling happen? Or is science going to have to come up with some other ways of making sure that a co-authored paper is a genuine group effort?