Sir

The issues of peer review and authorship are of central importance to scientists, and are frequently debated in journals. But many such discussions fail to refer to the increasing number of critical investigations into the workings of the publishing process.

The 15 July issue of the Journal of The American Medical Association, for example, contains 33 articles based on presentations at the 3rd Congress of Biomedical Peer Review in Prague in 1997, and constitutes a rich source of analyses of current and proposed publication procedures in biomedical research. Such studies suggest that two easily implemented systems could improve current practice.

Listing authors' contributions, and listing those who could guarantee the integrity of an article, offer a big improvement over the current ways of dealing with authorship. Similarly, the predominant system of editorial review, where the reviewers are unknown to the authors, is considered unfair. The two justifiable systems are a fully closed one (where the reviewers, authors and ideally the editors are unaware of each others' identities) and a fully open one (where all the parties know each others' identities). Research shows that the fully closed system suffers from poor success in masking identities, suggesting that the fully open system is the more favourable alternative.

It is encouraging that both explicit listing of contributions and open peer review are starting to be adopted (for example in The Lancet and Cell Calcium, respectively). But faster and more general implementation of these systems would be facilitated by a leading interdisciplinary journal such as Nature showing the way.