Is there really any place for peer review in modern scientific publishing? It seems to me that this process, grounded in 18th century gentlemanly science, has little or no place in a world with such wonders as the internet. At best, the submission of research to a group of scientific peers (also sometimes know as ’scientific rivals’) is rather quaint. At worst, if we are to believe the complaints levelled this week in the field of stem cell research, it is corrupt.

Dissatisfaction has surfaced among stem cell researchers with the publication of an open letter to the editors of academic journals. The complaints centre around the time and quality of peer reviewing, which they say delays unneccessarily the publication of important findings, while promoting papers of lower scientific value. Some also voice fears that work may be suppressed in order to allow others to claim breakthroughs.

Peer review has been questioned before, but is staunchly defended by those who see the value of the status quo. Certainly, for most researchers most of the time, the system works adequately - but that is about as much endorsement as I can give it. Peer review is not there for the benefit of science, but rather for the benefit of publishers. It is not a form of scientific quality control, and is barely even a certificate of validity. Its main purpose is simply to stop publishers getting egg on their face.

But even in this limited role, it has a questionable track record. This week also saw the final and full retraction of Dr Andrew Wakefield’s notorious 1998 paper in the prestigious UK medical journal The Lancet, in which he linked the MMR vaccine to autism. That paper had been fully peer reviewed. Fleischmann and Pons’s work on cold fusion was published in the peer reviewed Journal of Electroanalytical Chemistry, and yet the very phrase ‘cold fusion’ has now become a synonym for bogus science.

Is there an alternative? Well, yes - in fact there are a few. In astronomy, peer review has largely been abandoned in favour of paper ‘preprints’ which circulate widely and freely thanks to the internet. Alternatively, peer review could be conducted in a more open and honest way - reviewers could lose their anonymity, and discussions could take place in the open (again thanks to the internet). Exchanges may need to be refereed, but that shouldn’t be too onerous. Or the whole idea could be abandoned in favour of professional editorial quality control.

Some of the greatest scientific discoveries of the last century or so found their way into the public domain without the benefit of peer review. Crick and Watson’s structure of DNA was first published as a letter, rather than a paper - and, given the intense rivalry between teams seeking the structure of the molecule, that may just have precluded any temptation for bias by reviewers. Albert Einstein’s greatest works, published during his ‘annus mirabilis’ in 1905, were reviewed only by the editors of the journal. It seems to me that peer review may cause more problems than it solves, and that the time is right for a major overhaul of the system.
This comment was originally published in the Laboratorytalk Newsletter