Give credit where credit is due

Tim Birkhead proposes a simple way to protect the innocent from accusations of academic fraud

The great British physiologist Sir Edward Albert Schäfer, who died in 1935, was described in his obituary as having "a superb intolerance of intellectual dishonesty". Wonderful! And so should we all. As the comment makes clear, scientific misconduct has been around for a long time, but I can't help feeling that it is becoming more frequent, driven perhaps by our increasingly competitive, league-table culture.

On being exposed, miscreants often drag down all those who have been associated with them and, in particular, those who have co-authored papers with them. Some philosophers of science, and indeed many pseudo-philosophers chatting over coffee or beer, propose that anyone with their name on a paper should take an equal share of responsibility for it.

Morally they are probably correct; pragmatically they may be missing the point.

We collaborate with people because often they have a particular skill that we don't. Someone may be very good at designing an experiment but less so at developing a theoretical model that makes the most of the results. We collaborate and share authorship because the overall result is bigger than its individual parts. We may also motivate and reward hard-working assistants with co-authorship.

Laboratory directors add their name to a paper even though they may have done nothing specific to a particular piece of work, yet without their lab and all that goes with it there may not have been a study. To be sure, directors ought to be responsible for the information their laboratories generate, but it may be unreasonable or unrealistic to ask them to be familiar with and fully "accountable" for the work. In the same way, if I include particularly diligent assistants as co-authors, I don't expect them to be in any way accountable for the analysis, writing and interpretation of the data they have helped to collect. Some may balk at the idea of rewarding assistants with co-authorship, but until the recently negotiated pay scales, there were precious few other rewards for research assistants and technicians.

The problems with co-authorship emerge when things go wrong; when someone is accused of scientific misconduct. If no one is accountable, the chances are that all authors will sink with the ship.

But there is a simple solution, proposed as an option by the journal Nature , and I can't believe it hasn't been more widely adopted. The suggestion is that in the Acknowledgements section, the authors make a clear, mutually agreed statement about who did what. In order of authorship: w designed the study and wrote the paper; x collected the data; y conducted the statistical analyses; and  z ran the lab, got the grant and observed from a safe distance. I can't see many disadvantages, except perhaps the embarrassment of a laboratory director whose contribution is rather diffuse, or a case where someone prestigious has been made a guest author just to enhance the paper's status.

I suppose it is still open to fraud. The senior author could inflate the significance of his or her contribution: w designed the study, got the grant, collected the data, conducted the statistical analyses and wrote the paper; x , y and z all made encouraging noises from a safe distance. But providing the statement is mutually agreed, apportioning the blame if something is later discovered amiss limits and focuses the damage. I think we should adopt statements of responsibility as a matter of routine.

Tim Birkhead is professor of behavioural ecology at Sheffield University.