In 2001, 40 members of the editorial board of Machine Learning—then published by Kluwer Academic Publishers, now part of Springer—tendered their resignation. The reason behind the mass resignation was Kluwer’s policy of restricting online access to Machine Learning articles to only subscribers. “Our resignation… reflects our belief that journals should principally serve the needs of the intellectual community, in particular by providing the immediate and universal access to journal articles that modern technology supports, and doing so at a cost that excludes no one,” the members wrote in an open letter. Instead, they decided to found the Journal of Machine Learning Research (JMLR), a peer-reviewed, open access publication whose articles are freely available at the journal’s Web site.
“Open access is a model that makes sense,” says Lawrence K. Saul, JMLR’s editor-in-chief and a professor of computer science and engineering at University of California, San Diego. “You want your research to be read by as many people as possible. You don’t want it to be gated by artificial barriers.”
The sentiment is common among today’s computer scientists. In a paper world, goes the logic, the need of the scientific community to see its research circulate widely was aligned with the business model of commercial publishers, which had a financial incentive to ensure broad distribution. In the Internet age, of course, distribution is as easy as connecting to the Internet. Many researchers have thus grown reluctant to entrust their papers to journals whose online archives are restricted to those who can pay for them.
A variety of issues are bound up in the discussion, from peer review and proofreading to research and the scholarly record. At the core, for many scientists, is copyright, and the conviction that it does not make sense to transfer ownership of their work to organizations that profit from distributing it, especially when the work was given freely, and possibly even based on taxpayer-funded research. Also, the annual profit margins at publishers like Elsevier are more than 35%. The cost of the site licenses that grant access to scholarly journals continues to increase, accounting for an estimated 50%65% of library budgets. Computer science journals are inexpensive compared to those of other scientific disciplines, but their cost is rising nonetheless. According to a Library Journal study, the average cost of an annual subscription in computer science was $1,593 in 2011, up from $1,472 in 2009.
And for many computer scientists, the days when publishers were thought to perform a valuable service are largely gone. “Computer scientists can format their own papers, and there is no need for so-called professional typesetting—we have the tools to do it ourselves,” says Yann LeCun, Silver Professor of Computer Science and Neural Science at New York University. Though production delays have fallen, the allure of free, instantaneous online publication is undeniable for fast-evolving scientific disciplines.
The number of open access journals has exploded during the past decade. At press time, the Directory of Open Access Journals (DOAJ) listed more than 7,300 titles, up from 2,750 in 2008 and 560 in 2004. Of them, 317 are computer science titles. The open access model has obviously proven itself for smaller journals like JMLR, which continues to thrive more than a decade after it was founded. Nonetheless, questions remain about the future of open access. Is it sustainable in the long run? Will a standard business model emerge? Is there room for both open and paid access publications?
Matters of Money
Nearly all academic journals rely heavily on volunteer labor, from editorial board members to authors and reviewers. Yet there are a variety of additional expenses, from managing the peer review process and maintaining digital infrastructure to archiving old articles, creating reference links, proofreading, and providing metadata to services like Google Scholar. Commercial and learned society publishers typically incorporate these costs into the price of their subscriptions. At many open access journals, these costs are borne by volunteers, and offset by direct or indirect institutional subsidies. JMLR, for instance, is hosted by the Massachusetts Institute of Technology, and the journal’s founder and first editor-in-chief, Leslie Pack Kaelbling, personally paid for its domain registration and operating costs. The journal now covers such expenses through two grants from Google and Microsoft. Some open access journals receive support from their national governments. Others charge authors publication fees to offset their administrative costs. These fees vary from journal to journal, but typically range from $1,500 to $5,000 per accepted article. The practice is more common in some fields than in others, though it is not, on the whole, as widespread as it is often thought to be. In 2009, Stuart Shieber, the James O. Welch, Jr. and Virginia B. Welch Professor of Computer Science at Harvard University, found that only 23%30% of open access journals listed in the DOAJ charge publication fees. On the other hand, as the University of Quebec at Montreal psychology professor Stevan Harnad points out, the majority of top open access journals—those with an impact factor greater than one, like PLoS Biology—do charge publication fees.
How open access will evolve in the future is unclear. In the meantime, many traditional publishers have adopted a similar scheme. Under this hybrid model, authors can pay a fee to release their articles from the constraints of site licensing terms. Since subscription prices usually are not affected, the practice remains controversial (it is known pejoratively as double dipping), although up to 40% of authors participate in some cases. “It can be a very successful commercial model, particularly with high-volume operations,” says Bernard Rous, director of publications at ACM. “It also introduces a significant bias toward heavily funded research areas and rich institutions, where there tends to be a lot of grant money. Without institutional underwriting, you’re left with an unfair burden on individuals that have to pay.” For that reason, ACM decided against the idea during a recent review of its policies.
The Directory of Open Access Journals contains more than 7,300 titles, up from 2,750 in 2008 and 560 in 2004.
What ACM does offer authors is the ability to self-archive their work on personal Web sites and institutional repositories. Previously, scholars could post the author-prepared version of an article, after peer review. Now, thanks to a new service called Author-Izer, they can create a link that grants free access to the definitive version in the ACM Digital Library. Harnad calls self-archiving “green open access” (as opposed to “gold open access,” such as publishing in an open access journal). According to statistics he has compiled, more than 90% of scholarly journals enable authors to self-archive preprint versions, and more than 60% of them enable authors to self-archive the refereed final draft upon acceptance for publication. “Almost all publishers are already green,” explains Harnad.
Wouldn’t it be easier to let authors retain copyright and publish their work directly on their Web sites? Not necessarily, says Rous. “When some publishers shift from copyright to license, they put a lot of conditions in an exclusive publishing license that can leave the author very little room to exercise the copyright they retain,” he explains. “On the other hand, non-exclusive publishing licences may undermine the subscription model since authors may publish the same work in other titles and grant permission for inclusion in any and all aggregations.” The fear may seem unrealistic to senior scientists, although it is perhaps not completely unfounded in the current publish-or-perish climate, especially among younger researchers who are looking to inflate their publication records. “Except for the mind-set that if you retain copyright, it’s yours, authors are actually in a much better position to reuse the materials and do selective reposting and distribution under ACM’s copyright policy than under many of the licenses examined in our recent review of policy,” says Rous.
With Author-Izer, ACM offers authors the ability to self-archive their work on personal Web sites and institutional repositories.
While self-archiving has long been popular among computer scientists, it does not satisfy all scholars. Many would like to reform publishing and peer review in one fell swoop, starting from scratch and building a new system from the ground up. Their proposals often try to harness the power of the Internet to accelerate the review process. LeCun, for example, envisions a centralized online repository such as ArXiv, a site used mainly by physicists to distribute pre-peer-review versions of their papers, with an additional layer for collective review. “You still need a system to count points for tenure and promotions,” LeCun says. “But if we, as a community, cannot define our own standards, then who can?”
Others are more focused on reforming distribution alone. “I strongly support open access, but I believe there’s room for reader-pays journals as well,” says Eric Van de Velde, a technology consultant and former Library IT Director at the California Institute of Technology. “What’s wrong with the current system is the inefficiency that’s involved. People don’t realize how many middlemen there are between the reader and the publisher. A library doesn’t buy directly from a publisher. It buys from an aggregator, so of course the aggregator gets a cut. Then the library negotiates consortium deals and bundle deals. It gets ridiculously complicated.” Van de Velde would like to see publishers create a national, iTunes-like system for scholarly articles, freed from the market distortions of site licensing subscriptions. “You could pay for each article; that might be one model. Another model could be a monthly subscription that gives you access to download X number of articles. Once you put this national distribution infrastructure in place, you can experiment with all kinds of models at an individual basis.”
Harnad, on the other hand, feels that broader adoption of green open access is sufficient to address scholars’ goals. “Speed is a red herring. Having a central repository is a red herring. Self-archive the minute your article comes back from peer review, and CiteSeer will harvest it immediately—you can’t get it any faster than that. The only thing standing between us and 100% open access for every one of the 2.5 million scholarly articles that are published every year is a few keystrokes by the author.”
Of course, not all of those authors may be willing to take the additional steps to make their articles available via open access. One thing, however, is clear: Open access is growing fast in both recognition and popularity, making it a force to be reckoned with in the future of academic publishing.
Further Reading
Bosch, S., Henderson, K., and Klusendorf, H.
Periodicals price survey 2011: Under pressure, times are changing. Library Journal: April 14, 2011.
Delman, S.
ACM offers a new approach to self-archiving, Communications of the ACM 54, 11, Nov. 2011.
Harnad, S.
The green road to open access: a leveraged transition, The Culture of Periodicals from the Perspective of the Electronic Age. Gacs, A. (Ed.), L’Harmattan, Paris, France, 2007.
Rous, B.
Electronic publishing models and the public good, Nature, Feb. 13, 2012.
Join the Discussion (0)
Become a Member or Sign In to Post a Comment