New Journals Piggyback on arXiv

By Emily Conover

A non-traditional style of scientific publishing is gaining ground, with new journals popping up in recent months. The journals piggyback on the arXiv or other scientific repositories and apply peer review. A link to the accepted paper on the journal’s website sends readers to the paper on the repository.

Proponents hope to provide inexpensive open access publication and streamline the peer review process. To save money, such “overlay” journals typically do away with some of the services traditional publishers provide, for example typesetting and copyediting.

Not everyone is convinced. Questions remain about the scalability of overlay journals, and whether they will catch on — or whether scientists will demand the stamp of approval (and accompanying prestige) that the established, traditional journals provide.

The idea is by no means new — proposals for journals interfacing with online archives appeared as far back as the 1990s, and a few such journals are established in mathematics and computer science. But now, say proponents, it’s an idea whose time has come.

The newest such journal is the Open Journal of Astrophysics, which began accepting submissions on December 22. Editor in Chief Peter Coles of the University of Sussex says the idea came to him several years ago in a meeting about the cost of open access journals. “They were talking about charging thousands of pounds for making articles open access,” Coles says, and he thought, “I never consult journals now; I get all my papers from the arXiv.” By adding a front end onto arXiv to provide peer review, Coles says, “We can dispense with the whole paraphernalia with traditional journals.”

Authors first submit their papers to arXiv, and then input the appropriate arXiv ID on the journal’s website to indicate that they would like their paper reviewed. The journal follows a standard peer review process, with anonymous referees whose comments remain private.

When an article is accepted, a link appears on the journal’s website and the article is issued a digital object identifier (DOI). The entire process is free for authors and readers. As APS News went to press, Coles hoped to publish the first batch of half-dozen papers at the end of January.

To manage the submission and peer review software, the founders of the Open Journal of Astrophysics have created a web interface that allows a markup of the article. Reviewers comment on parts of the text and authors can respond, with all discussion occurring in the online interface. Authors make revisions by posting a new version to arXiv. “It’s traditional peer review but with a modern interface,” says Adam Becker, managing editor of the journal.

Another overlay journal, Discrete Analysis, led by mathematician Timothy Gowers of the University of Cambridge, has been accepting submissions since September, with the launch planned for the end of January.

Repeating a refrain common among supporters of open access publishing, Gowers points out that academics write and review articles for free, but then must pay to read them. So the new journal, Gowers says, is “a natural thing to want to do.”

Discrete Analysis uses software called Scholastica for managing submission and peer review. Scholastica charges $10 per submitted paper, but Gowers has secured funds from Cambridge to cover the cost.

Gowers is known for having an axe to grind with commercial publishers, particularly Elsevier; in 2012 he sparked a boycott of the company, protesting the high cost of subscriptions to their journals, large profit margins, and practice of bundling journals into packages so that libraries are forced to subscribe to journals that they otherwise wouldn’t.

An overlay journal platform called Episciences currently hosts a handful of journals in computer science and mathematics. The two-year-old project is an effort of France’s Center for Direct Scientific Communication (CCSD), and it interfaces with arXiv and other repositories, like HAL, a scholarly archive created by CCSD.

Episciences’ portfolio already includes some longstanding journals that have moved to the platform, as well as new journals. Additionally, “we are discussing with quite a few journals [that are] interested in joining the platform,” says Laurent Romary of the French Institute for Research in Computer Science and Automation (INRIA), a leader in the effort.

Romary says one of the important things such journals do is decouple publication and peer review. He’s an advocate of “post-publication peer review,” in which the article is available while it’s being reviewed, allowing researchers to get their results out as quickly as possible.

Another issue where Romary sees potential improvement over traditional journals is licensing. “The only thing that counts for a scholar is to be cited, so the license should be as unconstrained as possible so that on the one hand anyone can take the material and cite it, and second... you facilitate activities like text and data mining,” Romary says.

Proponents of overlay journals hope that others will follow their lead. “But whether they actually will, I don’t know,” Gowers says. “By being one of the trailblazers I hope we can make it more likely that others will do the same.”

“Some people are not convinced,” Coles admits. One issue with these journals is they don’t have the prestige associated with established journals. But, Coles says, “It’s a lot to pay for a status symbol. What we should be doing is paying for the dissemination of scientific knowledge, not the epaulets.”

And it remains to be seen whether overlay journals will be able to move beyond small-scale operations. “I very strongly support these things; I think they’re fantastic. But I have concerns about scalability and long-term sustainability,” says Paul Ginsparg of Cornell University, who founded arXiv.

“The hidden expenses in these things are never really taken into account when they operate on a small scale,” Ginsparg says. “There’s either personal or institutional subsidies going into the time, labor, internet connection, and all of the rest.”

“I can easily see how it would work for small journals in very specific topic areas,” says APS Editorial Director and Interim Editor in Chief Dan Kulp, who oversees the peer-reviewed journals published by APS. “Once you start to get thousands or tens of thousands of manuscripts a year, you begin to need additional overhead to track, coordinate, and confirm peer review. ... These costs are not negligible.”

As a result, says Ginsparg, “I don’t think they’ll ever be big enough to change the entire landscape, and I don’t think it makes sense to move the landscape over to this.”

But Ginsparg supports the effort. “We’re still poking around trying to figure out what is the right long-term solution for all of this and if nothing else that’s one of the reason why these experiments currently remain so important.”

“We have to keep trying these different things,” Ginsparg says, “because we’d like a more functional and a more financially stable model 20 years from now than we currently have.”