In the middle of 2002, a scientific scandal came to light in one of the most respected laboratories in the field of electrical and optical properties of organic materials and devices, Bell Labs. One of its highly acclaimed young scientists was accused of misrepresenting results. For Jan Hendrik Schön this spells the end of a career. It must be a personal tragedy for him and I think he deserves some sympathy and understanding from the community. Like cases before this, one cannot be sure of how much (or little) of this activity has contaminated his publications. A wide withdrawal of many of his scientific papers followed the scandal, across a range of the most highly respected journals. Many questions have been raised by this affair: Why did he do it? Why didn’t the peer review of his papers detect something earlier? How much awareness was there on behalf of his coauthors, who must surely share some of the blame? Is there something about this particular field of science that warrants further scrutiny?
In this day and age, understanding why someone might be driven to publishing large numbers of papers is quite simple: all assessment exercises, whether for individuals, laboratories, or universities depend on a high paper count in well-respected, high-impact journals. This increases the temptation to publish unreliable work, massage data, conceal shortcomings, and engage in the ‘multiple publication’ scenario. The latter is especially pernicious. Authors publish virtually the same results in several journals with only minor additions to the ‘story’. There are hundreds of examples in the literature of this ‘several for the price of one’ activity, which only serves to keep the reputation of the authors and their labs high in the league tables.
Most disturbing, however, is the publication of data that is not based on truth. Most scientists have performed experiments in which the predicted outcome does not happen. It is frustrating, challenging, tedious, and often takes years to find the reasons or answers. This does not fit well into our age of instant gratification and, with the modern tools of word processing, computing, and illustration, the means are available to ‘assist’ the data or ‘spin’ the story. We must be constantly on our guard to avoid letting this happen to ourselves, our colleagues, and the students we teach.
These recent events also point to the need for a much tougher and more rigorous line in the refereeing procedure. Something has gone seriously wrong if journals, such as Science, Physical Review, and Applied Physics Letters, among others, were taken in by Schön.
Perhaps it is not fair to point the finger at the editors of journals, maybe we all share some blame. The increased pressure on our time means that we tend to take the chore of refereeing less seriously, something to be done at home in the evening after a 12-hour day. If the paper reads well, comes from a famous lab, and the figures and equations are clear, then “it must be OK”. I have been guilty of this behavior and I am sure I am not alone. The sheer number of papers submitted for publication is one of the reasons why refereeing is getting less rigorous. If authors did not try to maximize their number of publications, but waited until they had the full story before writing one ‘complete’ paper, the pressure would be relieved.
The responsibility of coauthors is one of the key questions raised by the Schön affair. I have always adhered to the view that all authors are responsible for the ‘message’ in a scientific paper. Each author might contribute in different ways, one gathering electrical data, one making microstructure determinations, and someone else doing the theoretical analysis, but all must carry responsibility for the final paper. It is of paramount importance for each to check their own contributions and question those of their colleagues’. I think this seldom happens now, such is the relief “to add another paper to the list”. If authors behave responsibly, as I suggest, then this ‘internal peer review’ will both capture unreliable data or analysis and reduce the output of scientific papers.
The final point concerns the scientific field in which Schön worked. It has a strong interdisciplinary element, combining organic chemical synthesis with physical, optical, and electrical assessment more akin to electronics. Crosschecking colleagues’ work is more difficult when several disciplines are involved. This is going to be an issue in many of the physical/chemical/life science interactions that are becoming some of the most exciting aspects of modern materials science.
We have many lessons to learn from this sad affair and I hope we will all participate in applying them immediately.
[1] Peter J. Dobson is academic director of Oxford University Begbroke Science Park.