I have spent many hours these past few months listening sympathetically to complaints by quite a few young scientists about their treatment by editors and referees of well-known journals. Since their future careers - jobs, grants, recognition - turn on the issue of publishing in the right journals, they have all the right to be worried and even angry.

I can tell several stories which outdo anything I have heard recently, and I will recount one, which seems to reveal the most serious defect with the present system. Some years ago, we published a paper in a genetics journal on a class of genetic suppressors which we argued were due to enhanced expression of an alternative gene product produced by duplications and triplications. When the genes were finally cloned we were able to get physical evidence for this hypothesis, and a paper was duly submitted to a journal. It was rejected. One referee had no complaints, but the other said we should do genetic experiments to prove our point. The editor's letter urged us to pay attention only to this referee's comments and said the manuscript was seriously defective and could not be published without the genetic experiments. The following telephone conversation then took place:

S.B. [After introducing the matter]: Did you read the paper yourself?

Editor: No, I cannot be expected to read everything that crosses my desk.

S.B.: Are you aware that the referee you selected either can't read English or, more likely, is a total moron? The experiments he asks for were done and published a few years ago. They are clearly referred to in the paper, and the physical evidence supports them.

Editor: [Silence.]

S.B.: Who is the referee?

Editor: I can't tell you that.

S.B.: You should now accept responsibility for your bad choice and since his comment is both groundless and worthless, I assume you will now accept the paper.

Editor: No, we cannot go back on our original decision; there is no appeal.

I could cite several other instances where authors have been compelled to pay for the mistakes of editors who seem to value decisiveness more than truth and justice. It is incidents such as this that have led me to question whether the anonymity of referees needs to be guarded so closely. The standard argument for anonymity, of course, is that referees can speak their minds (if they have any) without fear of professional retribution. But it also allows their motives to remain opaque. For the innocent among you, here are two examples from S.B.'s glossary of referees' comments and their true meanings:

Referee: The treatment of the literature was cursory.

Meaning: The author has failed to quote my papers.

Referee: I am concerned about the interpretation of the experiment; the author should repeat these twenty times with different conditions of pH and temperature and wearing yellow socks.

Meaning: If I can slow him down I can get my own paper on the subject into print before him.

Removing the anonymity of referees may help, but there are more radical solutions, too. One was invented by Leslie Orgel and myself. Editors would be provided with printing inks with a range of different lifetimes from a few months to decades; they would publish everything received but would decide whether to use 2-month, 2-year or 20-year ink for each paper. At the appointed time the paper would vanish from the literature.

Another scheme suggested itself to me when I received a copy of the first issue of volume 1 of a new journal, with a title such as The Journal of Invertebrate Psychiatry. On the inside cover it contained the remarkable statement, "Back numbers of this journal may be obtained . . .", which led me directly to the concept of negative volume numbers. Again, editors would publish everything they received and would only have to decide when it would have been appropriate for each paper to have been published, assign it, for example, to volume -33, 1963.

In case anybody has not yet noticed, soon none of these or other ingenious schemes will he necessary because the whole system will have been 'done in' by electronic publishing. Papers are now being given publication dates when they go on the net, which can be months before the hard copy appears. The electronic preprint with open discussion (not refereeing) will soon become commonplace; in fact, labs could go into the publication business by themselves. We will need something to substitute for the present ratings given to papers appearing in 'superior, peer-reviewed publications' (and commercial publishers will find ways of making people pay for this). Perhaps we should have a readership index; it should not be beyond the wit of man to devise a way of recording whenever a paper is read, hard-copied or cited. Perhaps papers that are not frequently consulted should be progressively consigned to slower and more remote storage facilities, and ultimately perhaps only exist as printed copies in bound volumes in one library in Antarctica.

Sydney Brenner is currently president and director of science at the Molecular Sciences Institute in Berkeley, California, and on the Scientific Advisory Board of the Institute of Molecular and Cell Biology, National University of Singapore, where he also maintains a lab. His "False Starts" columns are seen regularly in Current Biology.

Andrzej Krauze is an illustrator, poster maker, cartoonist, and painter who illustrates regularly for HMS Beagle, The Guardian, The Sunday Telegraph, Bookseller, and New Statesman.

Tell us what you think.

Endlinks

Peer Review Process of Journals Indexed in Index Medicus - a look at whether these journals have a peer review process and an evaluation of this process.

Biomedical Journals and the Internet - a detailed introduction. From Journal Club on the Web.

Refereeing - links to many related resources.

Electronic Management of the Peer Review Process - presentation of a management system. From the Fifth International World Wide Web Conference.

When Fuzzy Thinking Is a Good Thing

A Moveable Feast:

How Good is Good Enough?

The Global View of Evolution

What is Lamarck's Signature?:

Neoplasia