Sir

In an extremely important, timely and welcome development for science, the E-Biomed proposal has evolved into PubMed Central, a free online public archive of the peer-reviewed and non-peer-reviewed literature in biology. It will be launched by the US National Institutes of Health (NIH) next January (Nature 401, 6; 1999). There is only one fundamental question that needs to be answered about the revised proposal: will authors be able to self-archive their refereed articles in PubMed Central?

The revised proposal is not clear about this question: it could be that only publishers will be able to archive refereed papers. This would be regrettable, because publishers are not likely to want to give away papers free, whereas authors are.

If authors are allowed to self-archive their refereed articles, PubMed Central will not only quickly make the biological literature into the optimal free resource for biological science, but it will provide a model for adoption by all other learned disciplines.

The director of the NIH, Harold Varmus, says that PubMed Central will be a web-based repository for barrier-free access to primary reports in the life sciences. Assuming that “barrier-free” means free for one and all in perpetuity, this will be an invaluable contribution to the advancement of biological and medical research.

Varmus also says that the screening of non-peer-reviewed reports will be the responsibility of groups that have no direct relationship to the NIH. This is as it should be. Peer review should continue to be implemented by scientific publishers and societies, and reports should be provided to PubMed Central from participating publishers and societies that have mediated the review process. But what about peer-reviewed reports from non-participating publishers and societies? Will the authors of such work be able to archive it in PubMed Central too? Or will the work available for free for all in PubMed only be that published by ‘participating’ publishers and societies?

The non-peer-reviewed reports will also enter PubMed Central through independent organizations, which will be responsible for screening this material. Will authors be able to self-archive their non-peer-reviewed reports? Some screening is a prudent idea but it must not be so restrictive as to prevent the self-archiving of preprints that are being submitted to peer-reviewed journals.

Will the availability of the peer-reviewed literature online free for all be conditional only on the active collaboration of publishers (who currently derive their revenue from selling it) or also on the active collaboration of authors (who give it away)?

Varmus has asked publishers, societies, editorial boards and other organizations interested in depositing content in PubMed Central to contact him at PubMedCentral@nih.gov. But what about authors interested in depositing their peer-reviewed and non-peer-reviewed reports? Is, say, university affiliation sufficient (which would be a good first step), or is it still to be only publishers who determine whether or not their authors' freely given reports can be given away for free? A great deal rests on the answer to this question.

It is to be hoped that, as PubMedCentral accrues more and more of the literature and makes it available to everyone for free, the bioscience community will become as addicted to this online archive as the physics community has become to the Los Alamos archives. In that case, the freeing of the rest of the literature will not lag far behind.

An online debate on this topic is at http://amsci-forum.amsci.org/archives/september98-forum.html