Sir

As part of the continuing debate over the future of scientific publications, I wish to make two comments and to advance a suggestion.

First, what is the difference between a published paper and a paper that has been uploaded by its authors onto some electronic archive? The difference is that one has been refereed, the other has not. So, refereeing is clearly at the core of the scientific publishing business.

Second, what are the calls on a scientist's time? These, in order, are: (1) to bring more money into the laboratory; (2) to write and publish more papers and conference communications; (3) to write reports for the funding agencies; (4) to get some science done; (5) to teach and/or carry out administration; (6) to sit on committees that distribute the money; and (7) to referee papers.

It is a real temptation today for a referee to clear a paper off his or her desk by writing a casual and ill-considered report. If refereeing is to be done better than of late (and recent scandals involving famous laboratories are revealing), it must move up the list of priorities.

The best solution would be to include refereeing as a factor in assessment for recruitment and promotion, and perhaps even for the allocation of funds. This requires a measure to be devised, which should probably be based on information provided by the quality scientific journals.

If scientific publishers wish to avoid embarrassing fiascos and to ensure that the publication process remains reliable, they would do well to address this issue. Otherwise, the core of their business is at risk.