A Science Publishing Revolution

Scientists and publishers generally agree that the Internet is sparking a science publishing revolution.1 They have yet to agree, however, on how to cultivate that revolution without alienating one another. The latest effort to push the online publishing envelope has a sizable group of scientists threatening to boycott journals whose content is not freely available in a public database six months after publication. This call for a "public library of science" (PLOS) has already caused quite a stir among both scientists and publishers. The initiative was started late last year by an advocacy group of scientists including Michael Ashburner of the University of Cambridge, Patrick O. Brown of Stanford University, Michael B. Eisen of Lawrence Berkeley National Laboratory and the University of California, Berkeley, Richard J. Roberts of New England Biolabs, Matthew Scott of Stanford, and Harold Varmus of the Memorial Sloan-Kettering Cancer Center. Frustrated with the rate at which journals were agreeing to put journal content into PubMed Central (PMC) (a National Library of Medicine (NLM)-sponsored public research repository that went online in January 2000), the researchers circulated an open letter, at first informally then on the Web (www.publiclibraryofscience.org). The letter contends that the "full contents of the published record" should belong to the public and not to journal publishers. Publishers are asked to make their content not only available for free on their own Web site, but on that of any public provider. The PLOS wants the establishment of several PMC-like repositories. The letter also pledges that as of September 2001, scientists who sign the petition will only "publish in, edit or review for, and personally subscribe to" those scholarly journals who have agreed to their terms. According to Varmus, whereas PMC is a freely accessible archive of literature intended to serve the general public, the PLOS is a kind of political movement based on the idea that the scientific literature ought to be more directly controlled by the people who generate it and use it. Says Eisen, "What we're really trying to do is recover some voice in the scientific community--in the decision making process--with respect to what the future of science publishing is going to look like." Scientists in developing countries where high journal subscription costs are particularly intolerable could benefit most. Libraries worldwide dealing with high subscription rates of obscure journals could also benefit--that is, if they're willing to wait six months to see journal content. Approximately 13,000 scientists from 124 countries had signed on to the letter by early April; about half of the signers are from developing countries. Some journals, most prominently Science, the journal of the American Association for the Advancement of Science, fear that releasing content for free after six months will unfairly sacrifice revenues. As with many science societies, AAAS relies on its journal as its primary source of income. In a recent editorial written in response to an accompanying editorial by PLOS advocates, the editors at Science announced plans to make their content freely available on their own site 12 months after publication,2,3 measures that fall well short of PLOS demands. Donald KennedyScience editor Donald Kennedy claims that there is continuing demand for back research content after six months, although at a low level compared to the first couple of months. He adds that content retrieval from the Web site enables exposure to other material and other products. "We can't estimate such losses accurately, but we think they are highly likely," Kennedy told The Scientist via E-mail. "We are reluctant to accept bland assurances from those who (a) aren't familiar with the economics of publishing and (b) are at no risk if they're wrong." Science rival Nature has yet to take a definitive stance with regard to the PLOS or making content available. Nature Publishing Group managing director Annette Thomas says that the issues surrounding the PLOS are more complex than many of the signers realize, and that Nature Publishing is still in "active discussions" with regard to this and other initiatives. Thomas expresses particular concern about small scientific societies in Europe that depend heavily journal-related income. She adds that Nature authors have expressed concern that revenue loss could mean that the authors themselves would have to absorb peer review costs. "That's not something that we're actually considering," says Thomas. "But it has come up in other discussions." According to Eisen, Nature Publishing has actually been more receptive to the PLOS than Science, although they've made no commitments. Journals supporting the PLOS and PMC argue that Web readership drops significantly after just a couple of months, making income losses negligible. "That's the basis of [the PLOS] initiative and why it's laudable," says Elizabeth Marincola, executive director of the American Society for Cell Biology. "They've done the calculation that [publishers] really have nothing to lose and I believe that's true." Even prior to the PLOS initiative, ASCB's journal, Molecular Biology of the Cell, had made its content freely available via PMC two months after publication. Marincola says that the journal's institutional subscriptions, the primary income of society journals, have not waned. "Publishing is an extremely lucrative business," points out Nicholas Cozzarelli, editor of the Proceedings of the National Academy of Sciences and PMC advisory committee member. "This idea that people have that it's some delicate flower that needs to be protected from the vicissitudes of the marketplace is total nonsense." As of about a year ago, PNAS started making its content freely available via its Web site and PMC four weeks after publication. It's the only high-impact journal to join PMC (and thereby conform to the demands of the PLOS as well). Cozzarelli suggests that by increasing readership, the move has actually resulted in a net economic gain rather than a loss. "We in scientific publishing should be allowed to make a reasonable return on what we do," says Cozzarelli. "But I don't think we should be in the business of running cash cows for people." 'GenBanks' for Scientific Literature The PLOS initiative has encouraged some journals to rethink their policies. Largely as a result of the initiative, the Journal of Cell Biology recently started making its content available for free on its own site after six months. JCB editor Ira Mellman says that the journal wholeheartedly endorses the idea of making scientists' published work as freely available to the community as possible--a prospect he calls not only noble, but long overdue. However, Mellman, like Kennedy, is vehemently opposed to the idea of releasing all content to any interested provider, a model the PLOS has borrowed from sequence databases such as GenBank. In the GenBank model, scientists send DNA sequence to one of three databases in a form all three centers agree on. The data is available to anyone--a company, for example, might use the data to set up its own system tailored to its needs. This is essentially what the PLOS would like to see happen with research papers: multiple centers like PMC with completely searchable, linkable research data in one underlying format. The PLOS does not approve of making content freely available on a given journal's Web site only. As Mellman writes in a letter submitted to Science in response to the two March editorials, he fears that passing along content to multiple providers will inevitably infuse multiple errors into research papers--Greek letters interchanged, images skewed or mislinked. He also suggests that having all journal content in one or more public repositories is a completely unnecessary, outmoded approach akin to "AltaVista claiming that they can only index a Web site if the complete content of that site is sent to them and hosted on their server." NLM director David Lipman, the executive secretary of PMC, says both concerns are completely unwarranted. He emphasizes that the process of placing journal content on PMC, for example, doesn't cause errors, but rather frequently catches them. Multiple centers, says Lipman, would actually ensure better quality control. And he points out that the "AltaVista"-type searching that Mellman refers to would only allow for "messy, noisy" retrieval. PMC and the like, for example, are able to much more precisely recognize accession numbers of entries to sequence databases, or gene names and symbols. Looking for Compromise Lipman and Varmus note that PMC is as much about archiving research data as providing access to that research. "What we're doing," says Lipman referring to PMC, "is saying to publishers 'whatever you do ultimately, let's make sure we get an archive started now.'" The desire to establish an all-inclusive archive--and to allay the concerns of potential PMC contributors who fear a loss of control, income, and readership--prompted PMC to offer a compromise at its late March advisory committee meeting. PMC will still require that its members submit their content for incorporation into the PMC archive. But, if the publisher so chooses, PMC will point back to the publisher's Web site exclusively for display of search results--that is, as long the publisher freely displays content on its own Web site within one year of publication. The journal's data will be in the archive but visible only through the journal Web site "window," and not PMC's. The move is not in line with the demands of the PLOS. Cozzarelli calls the compromise a "way station" that could help journals gain enough confidence in PMC to eventually join without such stipulations. He also hopes that the proposal will be seen as a good faith effort that will focus the threat of a PLOS boycott on only those "scofflaws" that refuse public release altogether. The principle protagonists involved in the PLOS, PMC, and science publishing in general will address an array of e-publishing issues at a closed-door, invitation-only meeting starting today at the Cold Spring Harbor Laboratory's Banbury Center. Nature recently started hosting an online debate on these issues at www.nature.com/nature/debates/e-access/index.html. It remains to be seen whether the aggressive PLOS strategy will pay off. Many principal investigators who sympathize with the PLOS initiative have not signed on because they feel boycotting prestigious journals will unfairly hurt the careers of the more junior people in the lab. "It puts principal investigators in a very difficult position," says Marincola, who supports the principles of the PLOS but worries it might be too aggressive. "I think the junior level people are the ones that have the most to gain by having the system altered," counters Eisen, who characterizes himself as "junior-level" faculty. "They're the ones who have the longest future ahead of them in science." Mellman disapproves of a PLOS strategy that attempts to "in one fell swoop change entirely a paradigm that has served the community extremely well and ignore the fact that ... all of us are already thinking about these terms." Mellman would have preferred an initiative that first focused on the profitable monopolies of big commercial publishers like Nature Publishing and Elsevier Science rather than one that lumps together such a diverse group of scientists, journals, and publishers. "At least do something to make [commercial publishers] serve the community a bit better," he comments. Eisen envisions new models of science publishing that remove journals' "gatekeeper role" by somehow directly subsidizing them, perhaps through grants. "It would be much, much better for science," he says, "if nobody was sort of sitting there, protecting small snippets of the scientific literature just in order to make a profit." The revolution in science publishing continues. Eugene Russo can be contacted at erusso@the-scientist.com References 1. D. Amber, "Scientists, Publishers, Societies--and Turf," The Scientist, 14[17]:1, Sep. 4, 2000. 2. D. Kennedy et al., "Is a Government Archive the Best Option?" Science, 291:2319, March 23, 2001. 3. R.J. Roberts et al, "Building A 'GenBank' of the Published Literature," Science, 291:2318, March 23 2001.

A Science Publishing Revolution

Interested in reading more?

Become a Member of