An astrophysicist has launched a low-cost community peer-review platform that circumvents traditional scientific publishing — and by making its software open-source, he is encouraging scientists in other fields to do the same.

The Open Journal of Astrophysics works in tandem with manuscripts posted on the pre-print server arXiv. Researchers submit their papers from arXiv directly to the journal, which evaluates them by conventional peer review. Accepted versions of the papers are then re-posted to arXiv and assigned a DOI, and the journal publishes links to them.

By piggybacking on or 'overlaying' the arXiv repository, the journal should operate at a fraction of the cost of traditional publishers and will be free for both readers and authors, says journal founder and editor-in-chief, Peter Coles, an astrophysicist at the University of Sussex in Brighton, UK. He announced on 22 December that the journal was open for submissions. It will go live later this month, once its first papers have undergone review.

Development of the software that powers the journal's peer-review system was led by Arfon Smith, chief scientist at the popular code repository GitHub. Because the software is open-source and available at GitHub, Coles hopes that researchers in other fields will adopt the same platform to create their own open journals. “Just cross out 'astrophysics' and write 'condensed matter' or anything else, and you’ve got your open journal,” he says.

Similar overlay journals already exist in computer science and mathematics; Tim Gowers, a mathematician at the University of Cambridge, in the UK, launched one high-profile example, Discrete Analysis, in September. But the Open Journal of Astrophysics is thought to be the first of its type in physics. Gowers says he is excited that the platform behind it is open-source "since it potentially reduces the costs for others even further".

The overlay model

Coles believes that traditional journals and their associated costs are no longer needed in fields such as astrophysics and cosmology, because most researchers already both submit their work to arXiv and read papers on it.

“The only objection to just putting things on arXiv is that it’s not peer reviewed, so why not have a community-based effort that provides a peer-review service for the arXiv?" he says — pointing out that academics already carry out peer review for scientific publishers, usually at no cost.

Coles himself covered the costs of developing the software platform for the journal, amounting to a few thousand pounds, he says. (Discrete Analysis licenses different software and is helped by a grant from the University of Cambridge, UK.)

GitHub is covering the costs of hosting the platform, so the only remaining expense is editors’ and reviewers’ time, which they give up voluntarily, says Coles. If the experiment proves successful and the volume of papers balloons, the journal may eventually have to charge authors a handling fee of a few tens of pounds, he adds. (The journal also relies on the continued existence of arXiv, whose running costs amount to less than $10 per paper).

The journal does not have the resources to offer services provided by conventional journals, such as heavy editing of papers. Instead, poorly written articles will be rejected and the authors referred to a list of professional copy-editing services, Coles says.

Gaining traction

Gowers welcomes the new journal; the arXiv overlay model is much more likely to succeed, he says, if many examples of it can be seen to be working. The journal has amassed an editorial board with high-profile physicists including Pedro Ferreira, a theorist at the University of Oxford, UK, and Andrew Jaffe, a cosmologist at Imperial College London.

But astrophysicists will not necessarily jump to publish in Coles' journal. Ewine van Dishoeck, an astrophysicist at the Leiden Observatory in the Netherlands, says she, for one, is unlikely to submit her work there. "We have a small number of well established and high quality journals in astronomy that everyone respects," she says.

Papers in astrophysics are effectively open already, van Dishoeck points out, because anyone can view pre-print manuscripts immediately on the arXiv, while journals in the field make final accepted versions open after a delay — typically 12 months after publication. An issue for researchers can be slow peer-review of papers, she adds, but the Open Journal of Astrophysics has yet to prove it can be faster.

Whatever their costs, the main problems facing all new journals hoping to achieve traction among researchers are ensuring speed and editorial fairness, adds Andrew King, a cosmologist at the University of Leicester, UK. "Reliability — and particularly fairness — are very hard to guarantee," he says, pointing out that the backing of long-lived organizations with a stake in the future of a field, such as learned societies, is often crucial to a journal's success.