December 1, 1994
Thank you for inviting me. I make a request of this commission. Please reaffirm that scientists must report data with accuracy and care. I ask this because what has happened in the case of the 1986 Cell article shows that there is a fundamental disagreement among scientists about whether the principle of honesty is critical to the advancement of science. I do not contend that any scientist would maintain that honesty is irrelevant in science, only that many scientists act and talk as if we scientists had many concerns that can take precedence over accuracy and honesty. Let us start with a common understanding that scientific progress is driven by scientists taking it upon themselves to report observations with care and accuracy.
In 1986, while I was a post-doctoral fellow at M.I.T. I discovered that results as recorded in laboratory records were seriously discordant with the results as published. Acting as I believe a scientist must, I brought the inaccuracies to the attention of the authors, a few of their colleagues, and to university officials. I asked that the false claims be retracted. The authors refused, and their colleagues backed them up. For those unfamiliar with the evidence in this case let me say that the fact that published results were not obtained has actually never been in dispute. The dispute of the past eight years has centered on whether it matters that the results were not obtained, and on whether the authors had done any unpublished experiments which tended to support their published conclusions. When investigated by the N.I.H., the authors presented laboratory records for unpublished experiments. They asked the investigators to accept these experiments in place of the published values which, they openly admitted, they had not obtained. Numerous investigators have now found all these "replacement" laboratory records to be fabricated. But the point I am making here is that even if the "replacement" results were authentic, which they are not, they would not approach in strength and persuasiveness those claimed in the original article.
When I persisted in my request for a retraction, I was accused of vindictiveness. I was told that any effort I made to publish a correction would be thwarted. I was threatened with legal action. My colleagues were told I had made a fuss because I did not understand the nature of scientific investigation. At least two scientists who openly expressed an interest in allowing me to work with them were pressured not to do so. They were told that giving me a job would cause tension within the community. I went to other senior scientists for help. None would agree to examine the evidence. I was ridiculed, and in some quarters, still am ridiculed for what I had done. In the years that followed the authors and their supporters conducted a wide-ranging public relations campaign attacking my competence and integrity. This is the reaction I encountered despite the facts that 1) my evidence was unassailable, I went through the proper channels as described to me by the university, and that I made it clear that I sought only a correction of the scientific record.
Since coming forward I have learned of the experienced of others who have given evidence in the interests of science and scholarship. Virtually all of these people have fared worse than I. Scientists call me saying that they are in the process of deciding what to do about what they perceive to be scientific misconduct. In my current job I can take no time to evaluate the merits of their cases, or to offer any help. Moreover, beginning in the Spring of 1993, I was no longer able to refer them to the witness aid program at N.I.H.. This program was forcibly shut down and its two staff members, Walter Stewart and Ned Feder, were barred from offering services to witnesses.
I learned of the work of Walter Stewart and Ned Feder when Charles Maplethorpe, a graduate student in the lab where I was a post-doc, contacted them and told them how university officials were behaving. Walter and Ned examined the evidence for themselves. Six months after my request for a retraction, they started a discourse, first with the authors and later with the scientific community as a whole, on the merits of my challenge and whether there were important principles at stake. A great number of scientific leaders put themselves on the record.
In the years before N.I.H. decided to step in an deal with the matter through misconduct investigations, only three of our scientific leaders took the position that the merits of my challenge were germane. John Edsall, Linus Pauling, and Bentley Glass, all of whom were, even at that time, well over eighty, said that they had not done an analysis of the evidence, but the facts mattered and principles were at stake. You may ask whether we should be concerned that the only people who took the position that accurate reporting of data was a principle worth protecting were either young, and without influence, or old enough to remember the place of accuracy in research. My answer is: "YES you should be very concerned." What does it say about primacy of factual accuracy when the only leaders who remember this primacy are so advanced in years? Linus Pauling has since died, but up until he died he made it his business to give me support and encouragement. John Edsall is now a friend and mentor. I spoke with Bentley Glass for the first time earlier this week.
Except for these three exceptions, the attitude of senior level scientists who took a position before the official misconduct investigation began was that discrepancies between results as obtained, and results as reported, mattered either not at all, or very little. It is instructive to review what reputable scientists thought did matter. 1) Research funding mattered at great deal, and revealing misrepresentation would jeopardize it. 2) The intent of the misrepresenter mattered, because, as long as it was asserted that all misrepresentations were the result of careless errors, the university officials maintained that there was no obligation to correct false claims. 3)The image of science mattered. By speaking out I was tarnishing this image. I could engender public skepticism, which in turn could translate into a reduced willingness to trust scientists. This in turn would hamper "scientific freedom". 4) Careers and reputations mattered, (but not the careers and reputations of those of us who had sought corrections). 5) The relative perceived potential of the parties involved to contribute to the advancement of science mattered. My potential to contribute was, and is, perceived as much lower than that of the authors of the article. Therefore, for the good of science, we should protect those at the top. 6) I was told that because there were relatively few scientists likely to rely on the false claims for their own work, keeping quiet would not waste a significant proportion of scientific investigation. The fact that as many as ten scientist were likely to be misled certainly did not matter. Maybe if one or two hundred scientists were likely to be misled it would have mattered, I don't know.
Most of the junior scientists with whom I discussed the matter did not relegate concern for truthfulness of scientific claims below these other concerns of career and funding. One M.I.T. student told me he had arrived at the university under the impression that science has something to do with truth, but with time he realized that scientists did not expect honesty from one another. He discovered that oral presentations should not be regarded as strictly truthful. Then he found out that abstracts submitted for meeting should always be taken with a grain of salt. Then he found out that putting the "best spin" on the data, rather than a candid presentation of the data, was part of "grantsmanship".
He said to me "We end up trying to figure out when it is not O.K. to lie." Until he saw how my case was dealt with, he had believed that the much touted honesty in science began at the point an article was submitted for publication in a reviewed journal. After the way senior M.I.T. scientists dealt with my evidence, this student concluded that misrepresentation was not deemed unacceptable, even in a published article.
The scientist in charge of the investigation at M.I.T. wrote to me saying that he objected to "....pernicious attempts to purge the scientific literature of all errors..." In truth, M.I.T. rode with such rough-shot derisiveness over my efforts to set the record straight, I found myself struggling to reassert that honesty was a cornerstone of the scientific method.
I ask you to consider that at the time of these events I was a fully qualified scientist. I had my Ph.D. and several years post-doctoral experience. I ask you to consider that if I was reduced to this state of debility by the disregard for accuracy in science I witnessed, how much worse is it for graduate students and technicians who witness such behavior. These graduate students go on to careers in science. These technicians often go on to jobs in industries where they are responsible for ensuring the quality of drugs to treat very sick people.
How do you think these junior people respond to seeing the truth being stretched by their seniors? Do you think they leave in disgust. Yes, some of them do. Do you think they decide to stay straight and honest despite the corruption they see seeping into research? Yes, many of them make this valiant commitment. Do you think they decide to speak out as witnesses. Hardly any of them do, in part because they know what happens to those who do. Do you think that some of them catch on to how the game is played and begin to stretch the truth themselves. I know that this does happen, because I have seen it happen.
Walter Stewart and Ned Feder brought matters to the point where it was abundantly clear to non-scientists that the scientific community was incapable of dealing with evidence of data misrepresentation. At that point, and I emphasize not until that point, Congressman Dingell and his staff started to look into the matter. This resulted in the most extraordinary circling of the scientific wagons. Congressman Dingell was pilloried as an enemy of science, subjected to baseless accusations that he wanted to introduce laws and regulations that would "cripple American science." I point out that it was the non-scientists like the Congressmen and the newspaper reporters who understood that science requires accurate reporting. They knew, even if we seemed to have forgotten, that the public funds us to do honest research. By enlarge, scientists behaved just like a "special interest group", seeking to preserve perks. The congressman received mountains of indignant letters from publicly funded scientists predicated on the assumption that the facts of case were irrelevant, and/or the accuracy of the article in question was of no concern whatsoever.
At this time I would like to take a moment to acknowledge the work of Congressman Dingell and his staff, especially Peter Stockton. I note that other representatives of the public also realized the importance of accuracy in publicly funded research, and took steps to promote it. These include Richard Roe, the late Ted Weiss and then Senator Albert Gore. If we as a community, and if you as a commission, are finally coming to a common understanding of our responsibilities as professional scientists, it is in large part because these representatives of the public took the trouble to point out our failings to us.
I believe we need a common understanding of proper conduct in science; our understanding of what constitutes misconduct can then follow. In this context, I define proper conduct as adherence to the principles that advance scientific knowledge. We must create an environment where we can review our choices of conduct within the framework of our professional purpose. In broad terms I define the principles of science as 1) honesty and candor, 2)accountability, 3) skepticism, and 4) collegiality. My premise is that if these principles, which are nothing more than our traditional principles, had been commonly held as essential elements for the practice of science and the advancement of knowledge, we would have been spared the disgrace of this and other mishandled cases. If we can agree that these traditional principles serve us well, and should guide our professional behavior, we will emerge from this ordeal with pride and strength.
The first of these principles, honesty and candor, I have already asked you to reaffirm as the bedrock of our profession. I also ask you to make it clear that the obligation to correct misrepresented results exists independent of the intent of the misrepresenter. "This is error, not fraud" was one of the excuses I was given for the refusal to retract. But guesses and judgements about intent should have no bearing on our responsibility to correct false claims. Surely it is obvious that if we allow misrepresentations asserted to be due to error do go uncorrected, we have set a standard that does not require care and accountability. Those caught misrepresenting can plead lack of intent, and never correct what they know to be wrong.
I also recommend that scientists be advised that there is no such thing as a "trivial" scientific lie. We must adopt the Feynman standard of honesty. I think it likely that virtually all scientific deceits start with "trivial" misrepresentations. Once a scientist starts to cook data, objectivity, which is essential to science, is lost. The scientist has started down the road towards fraud. So if you catch a colleague in an seemingly "immaterial" lie, the red flags should go up, and almost always they do not. If a scientist lies when we can see no reason for the lie, why should we assume this person will not also lie when the reason is obvious?
In 1986, a colleague told me that his position was that scientists should not tell any lies, even small lies. He added that when he discovers scientists engaged in "minor fudging" he immediately becomes suspicious of all the work done by that scientist. At the time I thought his position radical. Now that I have learned more about the nature and origins of deceit in science, I am asking you to endorse his position. Let the M.I.T. student, and all the other scientists and aspiring scientists know, that in science our standard is "utter honesty".
When I discuss the issue of scientific integrity with my colleagues, most take the position that dishonesty in science is really a trivial concern when viewed in light of the damage caused by sloppiness and incompetence. My best guess is that this is not the case. If a scientist is sloppy and incompetent, but is honest enough to accurately report how the experiments were done and what results were obtained, reviewers are unlikely to fund continued work. Even if such rubbish does get published, a rigorous scientist should be able to avoid being misled. It is only when scientists misrepresent, or fail to disclose relevant information, that true damage is done. I think that most of the problems attributed to sloppiness and incompetence would disappear if scientists fulfilled the professional obligation of reporting methods, controls and results with due care and candor.
Another reaction those of us concerned with integrity in science often encounter is that scientific progress is proceeding at a more than acceptable pace. So dishonesty, to whatever extent it occurs, is of little or no consequences, and should basically be ignored. This attitude is conducive to corruption, and I am exasperated by it. Imagine a banker explaining that the embezzlement at his bank should be ignored because commerce is proceeding despite it. Imagine a lawyer explaining that a betrayal of the interests of a client is of no concern because the legal profession as a whole is doing a good job. Well, in effect, this is what has happened when witnesses to scientific misconduct come forward. When presented with actual evidence in a specific incident, the response has often been to assert that fraud is rare, to invoke the achievements of science, and to set up a worried lament about the vulnerability of the creative process to "regulation".
The second principle I mentioned is accountability. A scientist is one who stands ready to substantiate claims, one who says: "This is what I did, this is how I did it. These are my results." I have found an appalling lack of consciousness of this principle among scientists. For instance, when a funding agency asked a grant recipient who was under investigation to help them index her laboratory records, she refused. Subsequently the agency suspended her funding. Her colleagues protested vigorously, saying that no action should have been taken against her until she had been proven guilty. There was no recognition that scientists are accountable for their professional claims. Certainly the accused had the legal right to remain silent. The questions is why other scientists thought she should be able to keep her funds while refusing to answer the questions of the funding agency.
Another example of this lack of realization of the function of accountability in science, comes from my dealings with N.I.H. Prior to the initiation of the O.S.I investigation in 1989, N.I.H. mishandled my case in a way that damaged me. When I protested, I was told that I could have brought my allegations anonymously. I, standing accountable for what I said, was told in effect by N.I.H. that my accountability was something they neither valued nor wished to encourage.
If the principle of accountability were widely understood and supported, the authors would have had no choice but to respond to the critique of their work presented by Stewart and Feder. Instead they asserted that their work fell within "the norms", and refused to answer a single objection.
The third principle is skepticism. A scientist weighs the evidence before taking a position. The case of the 1986 Cell article shows us what happens when scientists forget this aspect of their work. Numerous scientists joined in the controversy without examining a single shred of evidence, and these were the scientists who attacked me and those who had stood with me. In some cases I contacted them after their public attacks, and asked them to please examine the evidence, but they would not agree to do so. In one case an eminent scientist told me that, because his attack had appeared on the essay page of a major newspaper, he was under no obligation to ensure that his assertion of fact were true. In another case, I heard from a friend that a scientists was writing a piece which would be critical of me. I contacted this scientist and asked him to let me go over the evidence before he submitted his piece. He told me that he did not feel like going over the evidence, and now that I was insisting that he do so, he would not submit his article at all.
On the other hand, with the exclusion of the scientists initially involved in the university reviews, all of the scientists who examined the evidence before taking a stand confirmed that the article contained serious misrepresentations. As I have already said, the first who did this were Charlie Maplethorpe, Walter Stewart and Ned Feder. Since then there have been two categories of scientists who have carefully examined the evidence. In the first category are those who, as a matter of principle and professional responsibility, went over the data before taking a position. The earliest of these were Mark Ptashne, Walter Gilbert, Paul Doty, John Cairns, Serge Lang and Charles McCutchen. In more recent years the American Institute of Chemists made their position known by honoring me with their ethics award. Others have made their positions clear by inviting me to give talks on scientific integrity.
In the second category of scientists who examined the evidence and then confirmed the presence of serious misrepresentations are all the scientists who, as the result of an officially assigned duty, have taken part in the official misconduct investigations. These are Joe Davie, Hugh McDevitt and Ursala Storb, members of the first N.I.H. advisory panel, James Wyngaarden former director of the N.I.H., William Raub, Acting Director of the N.I.H., Brian Kimes, the first acting director of the OSI, the scientists who investigated at OSI and later at ORI including Suzanne Hadley, James Mossiman, Barbara Williams, John Dahlberg, John Sogn, special consultant from N.C.I., Stewart Sell and William McClure, members of a second and third scientific advisory panels.
The fourth principle is collegiality. Some scientists seem to think that collegiality is our responsibility to be nice to one another. Those of us who were bad mannered enough to point out the misrepresentations have been accused of violating the tenets of collegiality. But as I define collegiality, it is that special responsibility scientists have to help each other search for the truth. Collegiality calls on us to do responsible reviews and to never knowingly allow another scientist to be misled. This is the principle that requires us to bear witness in cases of dishonesty and intellectual theft. Collegiality means that our commitment to helping each other advance knowledge should never be displaced by loyalty to friends.
Last Friday, more than eight years after I first asked for a retraction of the false claims, the ORI released a report finding that everything I have said about data misrepresentation and fabrication is true. An extraordinary series of events have preceded the release of this report. Even with incontrovertible evidence in hand, it took a series of highly improbable events before the truth could emerge. And none of the extraordinary people that forced it to emerge are in a position to do the same thing today. Walter Stewart and Ned Feder have been barred from helping witnesses. Congressman Dingell no longer has the staff or authority to ensure fair treatment of witnesses and thorough examination of the facts. Peter Stockton, his chief investigator, will lose his job at the end of this month. Suzanne Hadley, the government investigator of extraordinary skill, courage and integrity, the first government investigator to chart the course through the web of fabricated laboratory records, is not in a position to do so again.
Soon after Suzanne Hadley handed in her findings at N.I.H., she was investigated by N.I.H. to determine if she had become too close to me, the witness. Despite the fact that all her finding were subsequently upheld by another investigation outside of N.I.H., and despite the fact that N.I.H.'s own investigation of her did not turn up the slightest error of fact or judgement, HHS has refused to place Suzanne Hadley in a position where she could again apply her considerable investigative talents to misconduct investigations.
I am a survivor, albeit a battered and permanently changed survivor, of this experience. It is imperative that you note, and take steps to correct, the fact that the N.I.H. careers of Walter Stewart, Ned Feder and Suzanne Hadley, have not survived.
If things are left like this, do you really think witnesses to scientific misconduct will believe you think they should come forward?
In conclusion, I again emphasize that we need a common understanding of proper scientific conduct. Let us start by agreeing that methods and results be reported with accuracy and candor, that investigators stand ready to explain the basis of scientific claims, that scientists examine the evidence before taking a stand, and that we have a collegial responsibility not to mislead or abuse each other. These principles did not arise to ensure that scientists meet the minimum standards to avoid being found guilty of misconduct. Nor did these principles develop out of an a false belief that scientists were closer to perfection than other humans. These principles became the foundation of our profession because the knowledge we accumulated throughout our history taught us that these were the principles that allowed us to unlock the secrets of nature.
Your task as I see it is a monumental one. You must create an atmosphere where scientists who take a stand in defense of these principles will be valued.