When a molecular biologist or a biochemist has made a discoveryoften after many months or even years of tedious experimentsthey tell the rest of the world by publishing their results in a scientific journal. So far, these journals have controlled who can read them and who cannotbut maybe not for much longer.
E-mail, Internet discussion groups, electronic databases and pre- or e-print servers have already transformed the way scientists openly exchange their results. And in the life sciences, researchers are now demanding that their work be included in at least one free central electronic archive of published literature, challenging the traditional ownership of publishers. The demand has sparked widespread discussions among scientists, publishers, scientific societies and librarians about the future of scientific publishing. The outcome may be nothing short of a revolution in the scientific publishing world.
It all started last fall, when an advocacy group called the Public Library of Science distributed an electronic open letter urging scientific publishers to hand over all research articles from their journals to public online archives for free within six months of publication. To add weight to their demands, the authors threatened a boycott starting in September 2001, pledging to "publish in, edit or review for, and personally subscribe to, only those scholarly and scientific journals" that agreed. As of April 21, some 15,817 life scientists from 138 countries had signed the letter, among them several Nobel laureates.
The authors of the letter feel they have every right to make these demands. After all, it is the scientists who supply the journals with their productsthe manuscriptsfor free. Scientists also help journals by reviewing and judging the quality of each others work, a process called "peer review," without pay. Publishers, in exchange, edit the articles, organize the review process and provide news items and other content. Finally, they produce, market and distribute a printed or electronic journal.
In the eyes of Michael Eisen, one of the initiators of the Public Library of Science initiative, the work that publishers do, however, does not justify that they then own the copyrights to the articles. "We think of the publishers as being like a midwife," he says. "They are paid for their role, and at the end of the day, they give the baby back to the parents."
Publishers argue that unless they own the copyright, they cannot protect articles from misuse. And scientific publishing is big business: like other scientific societies, the American Association for the Advancement of Science (AAAS), for example, finances most of its activities with income from its publication, Science magazine. "I think scientists all over would be shocked to realize what a phenomenally lucrative business scientific publishing can be," Nicholas Cozzarelli, editor-in-chief of the Proceedings of the National Academy of Sciences of the USA (PNAS), says. "There are huge sums of money to be had in this field."
Journals Don't Play the Game
What urged the authors of the open letter into action was the slow progress of PubMed Central, a free electronic full-text archive of research articles started by the National Center for Biotechnology Information (NCBI) at the NIH in early 2000. By storing articles in a common format on a single site, PubMed Central wants to facilitate sophisticated literature searchesfor instance, those restricted to certain parts of a paper, such as the figure legends. Ultimately it also wants to link the literature to other online databases.
PubMed Central asks journals to contribute their articles voluntarily as soon as possible after publicationat most after a yeargiving the journals time to offer exclusive access to make a profit (studies have shown that the demand for research papers decreases sharply after only a few months). But so far, only seven journals, including PNAS and a collection of e-journals, are participating, and a few additional journals have signed up. Even though some journals make their back issues freely available at their own Web sites, they are reluctant to give them away elsewhere. "Journals have just not wanted to play the game," Eisen says.
In physics, free electronic archives are old hat. Scientists have been submitting their own research papersboth before and after publicationto the Los Alamos e-print archive since 1991, without the participation of publishers, which simply had to accept the practice. Yet the American Physical Society, for example, still sells subscriptions to three journals that publish 14,000 research articles a year.
Perhaps not surprisingly, though, many publishers, threatened with either financial losses or a boycott, have been overtly hostile to the open letter. A number of scientific societies depend on the income from their journals to support their activities. But some scientists liken this system to a tax on their papers and think societies should subsidize their activities in other ways.
Also, some journals worry that outside archives hosting their articles will introduce errors into the files, lowering the reliability of the information. What if a g (microgram) suddenly becomes a mg (milligram)? PubMed Central actually detected errors in some of the papers they were given, thereby increasing the overall quality. "The more eyes to look at it and fingers trying to work with it, the more things you can find," says David Lipman, director of the NCBI.
On another level, some publishers resent a central, NIH-run archive like PubMed Central because they fear that technical failures would affect all users at once, and because the government might impose restrictions in the future, for example, by ruling not to publish certain kinds of research. On the other hand, PubMed, another NIH-managed database that grants free access to references and abstracts from 4,300 biomedical journals and links back to their Web sites, has been extremely successful and popular among both scientists and publishers.
Moreover, publishers point out that a commercial electronic archive, managed by HighWirePress and including nearly 250 journals from many scientific disciplines, already exists and that government money is wasted. Unlike access to PubMed Central, however, most of the HighWire Press journals are not free.
As a group, commercial publishers appear unsure about the recent developments and do not seem to have formulated their policies yet. Elsevier Science, Nature Publishing Group (a sister company to Scientific American, which is not a peer-reviewed journal), Cell Press and Academic Press declined interview requests, and Springer Verlag, as well as Allen Press, did not return phone calls. In a written statement, Annette Thomas, managing director of the Nature Publishing Group, commented that "many complex issues have been raised, and we are currently soliciting feedback from scientists, librarians, and other interested parties."
Charging Authors, Not Readers
One of the main questions to come from the current controversy is, Who will pay for publishing original research articles in the future if subscriptions decline? Only a small fraction of the publication costs of a print journalsome estimate as little as 10 percentcovers the editorial and peer review process. Many journals produce a costly print edition and add news, review articles and other valuable information, for which they have to pay. To offset their costs, journals derive income largely from subscriptions, as well as from advertisements, both in print and online, and reprints.
But since subscriber numbers may decrease if the access to journal information becomes free elsewhere, various publishers are thinking about changing their business model: instead of billing readers, they plan to bill authors, a practice that is already common in the form of page charges. Overall, these submission charges would amount to only a small fraction of a scientist's total research costs, they say, and could easily be included in research budgets. Libraries, freed from subscription charges, could also chip in on behalf of authors at their institutions.
Publishers would make exceptions for researchers from poor countries to ensure that no one is excluded for economic reasons. "We feel it is probably a better system to put the charges on the authors than the other way round," says Peter Newmark, editorial director at BioMed Central, a commercial publisher from the U.K. But steep submission charges could steer budget-conscious scientists away from these publications.
Many libraries seem to be in favor of open access archives like PubMed Central. "I think these are important efforts, and the library community is very supportive of them," says Joseph Branin, director of the Ohio State University libraries. In recent years, rapidly rising subscription rates for scientific journals have forced libraries to cancel many titles. Most of them now negotiate for electronic access to large sets of journals in consortia, giving them greater bargaining power.
If journal articles became freely available after a while, some libraries might stop subscribing to them. But for many scientists, instant access to the literature is crucial to keep up with current developments, so libraries will probably keep subscribing to the most important titles. "Because its available freely over the Internet after the first year of publication does not necessarily mean we are going to cancel our subscriptions to those," Branin remarks. Smaller, specialist journals, however, might be in danger of going out of business.
Libraries hope that subscription rates for the first few monthsbefore free access takes holdwill come down. But the opposite might be true: if many libraries opted out, publishers might try to recover their costs from the remaining ones. "And for those institutions, my own surely included, this free information could be very expensive indeed," writes Ann Okerson, a librarian at Yale University, in a contribution to a Nature Web debate. Scientists and libraries in developing countries, which often cannot afford subscriptions, would probably benefit most from free electronic archives.
A Possible Compromise on the Horizon
Come September, will the scientists who signed the open letter really go through with a boycott? Journals depend on their authors, but equally, researchers in the life sciencesespecially young investigatorsneed to publish in "brand name" journals, such as Cell, Nature and Science, to advance their careers. "I cant afford to boycott these journals because my career is not established yet," says an assistant professor from a New York medical school, who asked to remain unnamed. Nobel Prize winners, on the other hand, may find it easier to divert their papers to less established publications.
One of the practical problems of a boycott would be providing enough alternative journals for scientists to publish in. Some are thinking about starting their own journals. In mathematics, for example, some editorial boards in Europe have already left their commercial publishers and created new titles at their own institutions. "They are finding that while it does cost money, the costs are actually quite minimal," notes Mary Case of the Association of Research Libraries. BioMed Central also offers to provide the logistics for scientists who want to start their own journals.
That said, a possible compromise has recently appeared on the horizon: only two weeks ago, PubMed Central announced it would allow participating publishers to link back to their own Web sites, rather than insist that they display full-text articles on the NIH server. PubMed Central would still obtain a full-text copy for search purposes, but they would hide it from public view. Many publishers are currently considering this solution. "I think lots of publishers will grant free access after a period of time on the basis proposed in this compromise," says Donald Kennedy, editor-in-chief of Science. He also thinks that "under those circumstances, the threat of a boycott will vanish."
But for Eisen and many others, such an arrangement doesn't go far enough. Eisen still wants to see free access to alternative archives as well: "I remain absolutely convinced that the real future of publishing, five years out, is one in which nobody controls the literature."
Whatever the outcome, the scientific publishing world is in turmoil. Both Nature and Science have started e-debates on their Web sites, and contributions from many sides are pouring in. "It [the open letter] was not an unreasonable proposal," Kennedy comments. "It has gotten a good conversation started." In the end, it will probably be the authors who decide the issue. As Case puts it, "It is the scientists who are going to have to figure out how they want their work to be available."