Viewpoint

Is a Government Archive the Best Option?

Science  23 Mar 2001:
Vol. 291, Issue 5512, pp. 2318-2319
DOI: 10.1126/science.291.5512.2318b

The Editors

Rich Roberts and his colleagues have constructed a thoughtful argument for an online archive of published science. A seamless way of getting access to the scientific literature is an objective many scientists have sought, and the version outlined in the Roberts piece is being pursued with vigor and understandable passion by its advocates. We admire the goal, and suspect that evolutionary forces may be moving us toward it. We have decided to make our own back research reports and articles freely available after 12 months—at our own Web site—later this year.

The specific proposal of Roberts et al. goes further. It urges our readers to sign a petition that “advocates the free and unrestricted distribution of scientific literature 6 months after publication.” Actually, the petition does quite a bit more than that. It urges an economic boycott: signers agree not to submit papers to, review for, or subscribe to journals that do not submit to the petition's proposals. To begin a conversation among scholars with a threat of economic boycott is unfortunate.

However, we would rather focus on the qualities that Roberts et al. believe are essential to the archive they advocate. It should include all scientific papers and the content should be in a common format that allows for advanced search capabilities. Content should be free and “open distribution” should be allowed. PubMed Central (PMC) is given as the model of an archive that will meet these criteria. We believe other alternatives exist that can meet most of these goals faster and more effectively without putting nonprofit scholarly publishing at risk.

There already are multiple-journal sites—for example, the nonprofit HighWire Press (HWP), which archives over 230 journals, including biological, physical and interdisciplinary papers. More than 200,000 articles are freely available at this site. By comparison, there are only about a dozen journals at PMC, limited currently to biology.

Advocates of PMC argue that sites in which each journal is archived separately are insufficiently integrated. But searching across multi-journal, full-text repositories is already possible at sites such as HWP. In addition, 60% of this content is in a common format already. Why not begin with the already populated venue and add the integration, rather than the other way around? Why not use taxpayer dollars to promote innovative search technologies that do not require taking control of services provided by the private sector?

The proposition of Roberts et al. raises problems for Science, and for other journals. First, it will reroute an economically important source of online traffic for journals that offer content and other products on their sites. Second, unlimited redistribution of content could lead to misuse of content and loss of quality control. Third, it may expose users to risks historically associated with monopoly suppliers. For example, recently PubMed—on which PMC will depend—unexpectedly failed to process new content for over a month, inconveniencing authors and publishers.

We also wonder whether enough attention has been given to some of the economic issues. Experience shows that demand for scientific papers drops to about 1/10th within 4 to 5 months, but then continues at a low level for years. We plan to track our experience with free back issues carefully, but in the meanwhile, we take little comfort from the assurance that “costs of participation in open archives will be minimal.” Subscription and advertising revenue will be at some risk and transferring primary access to someone else's site may expose us to further losses. The value we add—through peer review, perspective and context-setting analysis of research, and good news coverage—requires revenue support from advertising. Moreover, Science supports other activities of AAAS—including science and public policy, kindergarten through 12th- grade education, a career-mentoring Web site for young scientists, and innovative “knowledge environments.” These benefit scientists from all fields. Posting our back content on a site that primarily serves biomedical scientists would confer a benefit on one group by taking benefits away from another—creating, in effect, a transfer payment from the sciences in general to biology in particular. That bothers us.

We worry, too, about another group of journals that will be entering a riskier environment. Our association is an umbrella organization, including many specialized scientific societies as affiliates. Their more focused journals must remain viable to ensure continued publishing options in highly specialized fields and for younger scientists. In most cases, academic library subscriptions provide the economic “floor” that guarantees financial sustainability. If papers from specialized journals were to become available on the PMC site, budget-conscious library directors would be tempted to cancel subscriptions. Some of the signers of the petition are scientists who belong to those very societies. Have they considered that their initiative will put PMC in competition with their own journals? When tax-exempt organizations go into competition with commercial entities they must pay unrelated-business income tax. When tax-supported organizations compete with commercial entities and nonprofits, the public has usually raised strong objections.

There are also questions about whether the proposed location for PMC—the National Library of Medicine, part of the National Institutes of Health—is the right one. NIH already sponsors, through its extramural programs, much of the biomedical research PMC will archive. It regulates the conduct of that research, controls much of the training of the next generation of researchers, and archives primary data. It now proposes that the results of the research it funds be given over by publishers and authors to a server subject to its exclusive control. The Congress or the President can eliminate support for certain kinds of science and have done so in the past. Would PMC then be able to archive papers on those subjects? Concentrating this kind of womb-to-tomb control in a single federal agency has risks, and we should ask whether we are entirely comfortable with a state-run, centrally managed economy in biomedicine.

Proponents of this plan include scientists of high reputation: Nobel laureates, leaders of institutions, and others whom we all admire. Nonetheless, we think its potential consequences require careful analysis and policy debate. We at Science are determined to participate in a constructive spirit.

Subjects

Navigate This Article