Integrity, inquiry and inquisition

David Weatherall considers a troubling tale of alleged scientific fraud

Despite the best efforts of Hollywood and the British tabloid press to persuade us otherwise, scientists are human beings. Hence it should have come as no surprise to learn that they make mistakes or are occasionally dishonest in pursuing their activities. Yet in the early 1980s the spate of press reports that followed the disclosure of a handful of cases of scientific fraud in the United States was enough to generate rumours of a crimewave in science. After the first of several congressional hearings to investigate these concerns, the chairman, congressman Albert Gore, concluded that these cases were only the tip of the iceberg, a view that was also expressed in the book Betrayers of the Truth, published in 1983. This claimed that crookedness is virtually endemic in science. And, even worse, it always has been; even Galileo, Newton and Mendel were not blameless.

For a long time this growing suspicion of science and scientists remained a peculiarly American phenomenon, possibly as the result of a political culture that had generated a deep mistrust of public authority and institutions following the Vietnam war and Watergate. But whatever the reason it was only later that similar concerns started to be aired in Europe and elsewhere.

While evidence that scientific misconduct has always been rife or has become more common in recent years is extremely flimsy, there is no doubt that the environment in which scientists work is becoming increasingly conducive to cutting corners or even to dishonest practice. Funding has become highly competitive, career posts in science are scarce, and pressures on heads of research teams and young scientists to be productive and publish with minimal delay are intense; the days when a piece of research could be polished and reflected on at leisure are over. And these tensions are not confined to the laboratory. The establishment of the research assessment exercise in this country, with the large amounts of money that accompany five-star gradings, has changed the role of deans and departmental heads into that of football managers, touting around to tempt away stars from other institutions or from overseas to help them to take their universities to the top of the Premier League. There are growing possibilities for conflicts of interest among those on bodies that fund science, and enormous pressures on scientists who receive support from the pharmaceuticals and tobacco industries to obtain results compatible with the goals of their sponsors. Similarly, government is encouraging scientists to work on problems that will provide quick answers for industrial exploitation, and many university scientists are setting up their own companies or are so closely involved with industry that it is often difficult for them to communicate freely with others in their laboratories.

Another challenge that faces the scientific community is the changing pattern of research itself. Many important research projects, particularly in the biomedical sciences, require the collaboration of very large teams of scientists. It is sometimes quite impossible for their leaders or any individual research worker to have sufficient technical knowledge to appreciate fully all the nuances of the work. This difficulty is compounded by the wide diversity and different levels of sophistication of the tools of modern biology, particularly projects that require the input of research methods from widely disparate fields. While the technologies of molecular biology may have introduced a new level of precision into the study of living organisms, many of the other approaches that are equally important - those required to study the products of genes and their interplay one with another for example - are still based on analytical techniques that are subjective, do not always give clear-cut "yes" or "no" answers and require a great deal of skill and experience for their interpretation. Even in the most obsessional hands, the results may not be entirely unambiguous.

Many of these tensions in the biological sciences are reflected in Daniel Kevles's excellent and carefully documented account of the prolonged inquisition that followed the accusation of scientific fraud against Thereza Imanishi-Kari, based on her input into a collaborative study with the Nobel laureate David Baltimore, work that led to a paper reported in the scientific journal Cell in December 1985. The extremely complex experiment described in this paper required expertise in molecular biology, supplied by Baltimore's group, and a series of extremely difficult serological analyses that were carried out by Imanishi-Kari. Following the publication of the paper a young post-doctoral scientist in Imanishi-Kari's laboratory, Margot O'Toole, blew the whistle, raising questions about the validity of some of the serological aspects of the study. While it subsequently became clear that some errors had been made, the central question of whether these were genuine mistakes or whether data had been fabricated, unbelievably, took over ten years to answer.

In the climate of intense mistrust of science that had been gathering strength in the US during the early 1980s, and now that a Nobel prizewinner appeared to be involved, it was inevitable the anti-science lobby would go for the jugular. In his remarkably objective and balanced account of the events that followed this accusation, Kevles paints a frightening picture of the prolonged interrogation to which the scientists, including O'Toole, were subjected. There were numerous institutional inquiries, congressional hearings led by ambitious congressmen who were clearly disenchanted with the scientific scene, detailed analyses of laboratory notebooks by forensic experts who seemed to have little understanding of what they were looking at, harassment by some unproductive scientists who found fulfilling new careers as self-appointed hunters of scientific misconduct and continuous trial by media. Some extremely distinguished American scientists took sides against Baltimore, not always, it appears from Kevles's account, with a clear understanding of the science and other issues involved. Indeed, one of the main reasons the work of this team was eventually cleared of scientific fraud was because a lawyer took the trouble to become properly informed about the extreme complexities of some of its serological aspects.

Although there was never any suggestion that Baltimore had behaved improperly in his input into this research, both he and his family suffered very deeply during this inquiry. Among other things, he felt forced to resign his post of president of Rockefeller University. But, as Kevles points out, while he may have overreacted during the congressional hearings, and hence delayed publication of the correction of some of the mistakes in the paper, his refusal to abandon Imanishi-Kari over a ten-year period of immense personal pressure was an act of great courage. Indeed, the reader is left wondering at the sanity of a society that could submit one of its most brilliant and creative scientists to such pressures for so many years.

This disturbing book highlights some extremely important questions to which, perhaps not surprisingly, Kevles does not offer any easy answers. Are scientists capable of regulating their activities? Many of them hold that science is, in effect, a self-regulatory activity. If the questions that are being addressed are sufficiently important, mistakes or even fraudulent practice will come out in the wash, simply because others will always attempt to validate the findings. While this is largely true, unfortunately it is not an argument that seems to be cutting much ice with governments, funding bodies and society, given the bad press science is receiving from so many different directions.

There are, in fact, several ways the scientific community can help to keep its house in order. Many of the problems of the Baltimore case appear to have started with poor communication within and between laboratories. Young (and not-so-young) scientists need constant support, must not be put under undue pressure and, above all, must never become isolated from the work of their colleagues. It is vital that heads of laboratories allow free, open and critical discussion about the research they are pursuing. While it is inevitable that tensions will develop from time to time among a group of driving, creative people, secrecy in a laboratory, regardless of whether it stems from personal insecurity, commercial interests or misplaced ambitions, creates an environment in which the objectives of good science can be lost in an atmosphere of unproductive intrigue and in which scientists may lose those day-to-day interactions with their co-workers that are such a major disincentive to self-delusion, or worse.

At the same time, universities and other research establishments need to set in place transparent and effective mechanisms for the speedy and informed investigation of any possibility of scientific misconduct, with clearly defined ways to protect the careers of whistleblowers and with input from outside the institution from the beginning of the inquiry. If, as now seems inevitable, it is thought necessary to establish national or international bodies to overlook the activities of science, those who seem so ready to tackle this difficult task should first absorb the lessons so clearly set out in Kevles's book, particularly the problems posed by an over-hurried attempt of this kind by the National Institutes of Health in the early days of the Baltimore inquiry. And, in particular, they must avoid generating the kind of over-heated atmosphere of suspicion and mistrust that was so unhelpful throughout this saga. As highlighted in Kevles's account, ill-conceived witch-hunts bring out the worst in accusers and accused, and objectivity is lost in a mist of emotion and prejudice.

Of the many lessons to be learnt from this book, perhaps the most important is that, because of the complexities of modern-day science, those who sit in judgement on scientists must be properly informed about the field under investigation. The fact that somebody is a distinguished politician or lawyer, a forensic science expert or even a scientist of high repute and impeccable character does not necessarily mean they are qualified to judge the merits of the work of others.

In one of his previous works, In the Name of Eugenics, Kevles unravelled the origins of another gathering source of tension between science and society. This new book is equally important, reflecting the increasing mood of mistrust of science on both sides of the Atlantic which, if not dealt with, could do great harm at a time when the potential of the benefits of research has never been greater. It is therefore essential reading for those grappling with this problem and a fascinating and beautifully written tale for anyone who simply wants to find out more about what makes science and scientists tick.

Sir David Weatherall is regius professor of medicine, University of Oxford.