Stewart-Feder: Reassignment Is A Moral--Not An Administrative--Matter

Editor's Note: Scientists Walter Stewart, 48, and Ned Feder, 65, have been working side by side at the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) in Bethesda, Md., since the late 1960s. Their employment agreements with NIDDK once required that they spend 80 percent of their time on health- related lab research and 20 percent on their investigations of scientific fraud and misconduct. However, their renown in the international science community stems virtually 100 percent from their work in the latter category. Most prominent among their forays into what's become known as fraud-busting are cases that arose in the early and mid-1980s involving scientists John Darsee, a Harvard University cardiologist, and David Baltimore, a Nobel Prize-winning immunologist. In the first instance, Stewart and Feder found--and subsequently documented in a published article (Nature, 325:207, Jan. 15, 1987)--that some of Darsee's coauthors on a paper had, to use Stewart and Feder's phrase, "not been candid" in their research. The two misconduct sleuths were widely praised in the science community for their efforts in the case. Likewise, their persistent efforts in the Baltimore affair elicited strong praise for the NIDDK team from many scientists. It also drew harsh criticism. In that instance, Stewart and Feder launched an examination of work coauthored by Baltimore and published in Cell (45[2]:247, April 25, 1986) concerning gene expression in transgenic mice. Stewart and Feder's findings supported accusations made by Massachusetts Institute of Technology postdoc Margot O'Toole that some of the data in the Cell paper had been incorrectly reported. To some minds, the fraud-busters had gone too far in their detective work by taking Baltimore on and sullying his reputation. Hostility mounted against them when they became advisers on scientific misconduct to the Subcommittee on Oversight and Investigations of the House Committee on Energy and Commerce, chaired by Rep. John Dingell (D-Mich.), which many scientists regard as an intrusive body of outsiders poised to police a science community that should be able to police itself. For the past five years or so, Stewart and Feder have become targets of widespread disapproval in the research community. Despite calls for their dismissal from NIDDK, however, they have persisted in their investigations--with the official approval of their NIDDK superiors. As part of their misconduct-hunting weaponry, they developed what they call "a plagiarism machine"--a computer system whose software rapidly scans and compares texts, identifying identical word groupings. Earlier this year, they used this device in examining the work of a nonscientist--University of Massachusetts historian Stephen Oates, author of a 1977 book titled With Malice Toward None: The Life of Abraham Lincoln (New York, Harper & Row) as well as biographies of the writer William Faulkner and Martin Luther King, Jr. Stewart and Feder concluded that Oates had plagiarized the work of previous authors in creating his biographies. And not long after they made their opinions known, their operations at NIDDK were officially shut down. In notifying them of this move, the institute expressed its intention to impound all of their files--including documents obtained by Stewart and Feder from misconduct witnesses to whom they had vowed confidentiality. In a phone interview, Stewart did not conceal his indignation over his and Feder's ouster and reassignment to other institute positions. Particularly painful and alarming to both is, in Stewart's words, "the realization that the shutdown of our laboratory is another step toward the policing of science by a bureaucracy." In the accompanying essay, Stewart and Feder reflect on various aspects of what they see as their mission--a morally imperative mission that, in their view, should be shared by all in the science profession. Scientists--not bureaucrats or politicians--should, in our opinion, be the ones to set the standards of the scientific profession. This means that scientists have to pay attention to the problem of scientific misconduct. The National Institutes of Health's recent termination of our work on scientific misconduct and our forced reassignment to other work is a step backward for scientists who care about maintaining the standards of the profession. Our reassignment is bound to discourage those who have seen instances of scientific misconduct, such as data falsification or plagiarism, and want to report them. For years, it has been common knowledge that reports of scientific misconduct are often poorly investigated and those who make the reports are often exposed to retaliation. We are attempting to apply the scientific method to understand the origin of this disturbing problem. Our work has involved the use of case studies. As part of our work, we have talked to a large number of individuals who believe they have witnessed misconduct, but who are reluctant to report it because they fear retaliation. In 1983, we set out to learn about scientific misconduct by delving into a prime example of scientific fraud, the Darsee case. What we found was not reassuring: About a third of Darsee's coauthors had not been candid in reporting their work, which is one of our criteria for misconduct. Because this sample was special, our findings did not imply that a third of all scientists are guilty of misconduct--and we still do not know what fraction of scientists are. But the Darsee case presented evidence to us that the occurrence may not be as rare as one would like it to be, and it inspired us to move forward in our study of misconduct. A disturbing aspect of the Darsee case was that the scientists who initially reported wrongdoing to their staff superiors were poorly treated. This proved to be a general feature of scientific misconduct cases. We are able to recount many cases of men and women who were shabbily treated after reporting their evidence of misconduct. How important is the problem of scientific misconduct? At first, we thought its practical importance was small. After all, it is widely believed that a high percentage of research is not of much use. Indeed, we believe that most of the hundreds of thousands of articles produced each year by federally funded science agencies are ultimately useless. Occasional fraudsters, it stands to reason, add only marginally to the pile of useless articles. There is no way of measuring the amount of direct damage caused by fraud. However, real damage is done for certain when scientists respond--as they commonly do--with indifference or hostility to particular allegations of fraud. We have often used the following analogy: An individual burglary causes no significant damage to the gross national product. But if society does not respond to particular cases, things will get worse. Cases of fraud that are mishandled within the science community cause, in our view, three kinds of damage: The person reporting misconduct suffers by loss of position or professional standing. This is ethically unacceptable and is a reproach to a profession whose goal is truth. The injustice leads to the intervention of bureaucrats designated to deal with science fraud. Under the usual political pressures, this bureaucracy can hardly avoid becoming ineffectual or worse. Most important, when misconduct or fraud is condoned or not handled properly, it cheapens the pro- fession. When some scientists profit from misconduct, the resolve of others to compete honestly suffers. Over the years, we have sometimes been accused of favoring the presence of "science police"--outsiders charged with monitoring the practices of researchers. The truth is the opposite. We favor professional self-regulation. We advocate methods that are an integral part of science, such as free and open public debate, the sharing of data used in published papers, and the assuming, personally and professionally, of responsibility for inaccuracies in published papers. These methods are already a part of science, and they are effective at getting at the truth: Had they been followed, we would not have felt compelled to become involved in the Darsee, Baltimore, or Oates cases. Despite our aversion to the notion of "science police," we occasionally have been accused of being just that. But this is not accurate. With the important exception of our work for the Dingell subcommittee, we have never had any power beyond the weight that our scientific colleagues have chosen to give the evidence we present. We wish no power, only a chance to speak. We hope scientists will address the question of misconduct using the methods of science--and we have always wanted to be part of that process. Our use of case studies to understand these problems has stirred controversy. When we began our work on the Darsee case, the dean of Emory University, where Darsee had worked, asked James Wyngaarden, then director of NIH, to "influence" us to "cease [our] bothersome activities." Three weeks later, we received a letter from a superior directing us to stop doing what we were doing with NIH resources. We appealed the directive, and six months later, NIH took the courageous step of reversing its decision, thereby allowing us to resume our work on the Darsee case. Subsequently, NIH actually mandated that we proceed with our studies of misconduct. It is a tribute to NIH that it has generally supported our research. However, the conviction behind that support has not been unwavering, with the result that NIH repeatedly has been thrown into crisis by the findings of work it had approved. For example, when we pointed out problems in Baltimore's Cell paper, we received two written directives from our superiors at NIH forbidding us from talking with the authors. But then, soon after, when the manuscript of our report on the matter was refereed by NIH personnel, the referees complained that we had not talked with the authors--after which we received written directives saying that we must discuss the problems with the authors. NIH's unsteady behavior didn't stop there: After we complied and had our discussions with the authors, we were forbidden by our supervisors to submit the resulting manuscript for publication. Then, after the American Civil Liberties Union asked NIH about the basis for this decision, we were allowed to submit the manuscript because of, in the words of an NIH memorandum to us, "the overriding importance of permitting free and untrammeled investigation and reporting thereof in any research institution." Moreover, on numerous occasions, when we were invited to speak about the Baltimore case at universities and professional meetings, we were forbidden by NIH officials to attend. On several occasions, the disapproval was subsequently--and sometimes mysteriously--changed to approval. At one point, we were forbidden to participate in any activities off the NIH campus without the permission of our supervisors--a ruling that prevented our giving an invited talk just a few blocks south of the campus. Once, just before we were to deliver an invited talk on the Baltimore case, we were handed a set of NIH guidelines on permissible speech that contained a prohibition on speaking about "information that should not have been disclosed to the public." We asked about procedures for determining which information that is public should have been disclosed to the public and which information that is public should not have been disclosed. Following our question, the guidelines were declared a "draft," and in the replacement guidelines the prohibition had vanished. Despite its frequent ambivalence, we feel that NIH's long-term support of our difficult but necessary work on developing methods for the study of scientific misconduct should be a matter of pride to the scientific community. But our current forced reassignment appears to have been prompted by political pressure applied as a result of our using a newly developed technique to explore an academic controversy. In caving in to this pressure, NIH, in our opinion, is manifesting a very troubling brand of timidity. As has been widely reported in the media, with the help of our plagiarism machine, we found in biographies of Abraham Lincoln, William Faulkner, and Martin Luther King, Jr. written by Oates hundreds of previously undetected examples of what we consider plagiarism. After consultation with our NIDDK supervisor, and after giving the biographer numerous opportunities to respond to our allegations, we submitted the evidence to the American Historical Association (AHA), a government-chartered body that judges such matters. Oates reacted by attacking our credentials and motives. He also launched a public relations campaign on Capitol Hill that prompted inquiries from senators and congressmen. It was at this point that NIH took the position that our work had "progressively moved outside the mission" of the institute. Our files were to be packed by professional movers and shipped to dead storage, our work was to be terminated, our documents were to be searched and then warehoused, our computers were to be given to other scientists, and we were to take on new NIDDK jobs, unrelated to our past experience--all of this despite the fact that our work had received ratings of "Excellent" for the past three years, most recently on March 26, 1993, from the very same NIDDK official who subsequently directed us to terminate our work. We were directed by our supervisors to turn over allegations of scientific misconduct to the Office of Research Integrity (ORI) or to the Office of the Inspector General. The problem is that we had received this sensitive information from people on our promise that this would not ever happen. As a matter of conscience, we do not see how we can ethically comply with this directive to breach faith with our sources. During the first week of May, the plans for our termination took a bizarre turn. We were told that our rooms would be filled with file cabinets and documents. We were notified that, under police guard, the locks would be changed, and neither we nor any NIH administrators would have keys to the room; only the NIH police force will hold keys. Our plagiarism machine and all our computers and data will remain locked in this specially secured room. Neither we nor anyone else will have access to the machine. NIH, which prides itself on being like a university, appears to have forgotten how academic institutions are meant to behave. Our supervisors, citing simply the progressive movement of our work "outside the mission" of the institute, did not tell us that their decision had anything to do with our work on Oates's biography, but an NIDDK spokeswoman repeatedly gave a different story to the press. She said that the complaint by Oates was "one of several factors" in the decision to terminate our work. When we called to ask what the "several factors" were, she abruptly ended the call. To the argument that our plagiarism machine should be used only on science cases, we make three points. First, we make our discoveries available to all. Second, experts frequently consult with colleagues in other fields. Third, slipshod scholarship and academic misconduct should be of concern to all scholars. We asked our future supervisors what we would be allowed to do if we saw published errors in a scientific paper. The answer was simple: Notify your supervisor and ORI. This directive amounts to the repeal of the system of scientific self-regulation and its replacement by a system of regulation by authority. This view of science--at the world's largest biomedical research institution-- is disappointing. NIH's shutdown of our activities raises a crucial question: Should anyone be doing research on scientific misconduct? If yes, who? All scientists have an interest in this question. The job of a scientist is to find the truth and make it known, and to correct error. The first part is fun and results in enormous progress; the second part is essential for the health of science. The seizure of our files, the termination of our work, and the prohibition against public discussion of scientific error is not an administrative issue. It is a moral issue. Walter Stewart and Ned Feder are scientists employed by the National Institute of Diabetes and Digestive and Kidney Diseases in Bethesda, Md.

Stewart-Feder: Reassignment Is A Moral--Not An Administrative--Matter

Interested in reading more?

Become a Member of