Originally posted on the Nature news blog

The European Research Council (ERC) announced today that it has joined the list of more than 170 institutions to financially support arXiv, the major online repository for pre-print papers operated by Cornell University Library in Ithaca, New York.

The cash involved is peanuts: $1,500 for the remainder of 2013 and $3,000 a year thereafter. (The ERC is already giving €90,000 this year to the life-sciences archive website Europe PubMed Central). But the ERC’s support is significant because it is a taste of what may be to come, says arXiv program director Oya Yildirim Rieger.

At meetings today and tomorrow in New York, she and other arXiv advisory board members will discuss how to allow other research funders, and potentially publishers, to support the website, which now funds its annual costs (around $800,000) mainly through libraries and philanthropic organizations. They will also discuss whether publishers might deposit articles directly into arXiv, which could change the website’s author-centric character.

ArXiv currently holds 875,000 e-papers, mainly in physics, computer science, mathematics and statistics — although it has expanded to cover other disciplines, including biology — and gets around 8,000 new pre-print submissions each month.

As staff and server costs escalated, in 2010 Cornell University Library said it would seek outside financial support from its users, mainly academic institutions. The appeal went well: this year, arXiv is expecting some $320,000 from those users (now including the ERC, but it is not the first research-funding organization to join the supporters’ list); and $350,000 from the Simons Foundation, a private foundation based in New York, which agreed a five-year funding deal in August 2012. Cornell University Library chips in with $75,000. But, says Rieger, arXiv would now like to expand its budget from its current “minimal, skeletal” level to one allowing innovation and new features, including modernized server architecture.

Rieger says that arXiv will have to be sensitive about conflict-of-interest issues if it is to get support from publishers. In August, arXiv overseers made a conference call with eight publishers, including Springer, Elsevier, Wiley and the Institute of Physics, to discuss wider support, and a pilot effort in which publishers might directly deposit the published, peer-reviewed version of an article onto arXiv after a suitable delay, perhaps 12 months after publication. (At the moment, authors do sometimes upload final versions of research articles, but on an ad hoc basis, and often not the publisher’s ‘version of record’, but a peer-reviewed final manuscript.)

That idea is connected to publishers’ efforts to make published research papers free to access under what they are calling the ‘CHORUS’ initiative, which was compelled upon them by the White House’s announcement in February that publicly funded US papers be made free to access.

Exactly how readers will view their free papers online isn’t yet clear: papers may be stored both on publisher’s sites and on repositories such as arXiv, PubMed Central, or smaller national or institutional stores; or the repositories may only link to articles stored on publishers’ sites. What also isn’t clear is who gets to choose the reuse rights for open articles posted online — and what those reuse rights are going to be.

But for arXiv at least, the nascent plans have the potential to change the website’s character from an author-centric ‘preprint’ server, as it began life more than 20 years ago, to a repository also supported and used by research funders and publishers to store final versions of articles.

That change is already happening to some extent with the Episciences project, a mathematician-led effort to create free open-access journals which would store their peer-reviewed articles on arXiv.

If peer-reviewed articles were stored on arXiv, journal publishers would be paid for their services in one of two ways. Articles would be made available on arXiv only after a delay — say 12 months — in which case the conventional subscription model serves, requiring libraries and other users to pay the publisher to get past the paywall on their website for immediate access. Alternatively, publishers would be paid up front so that articles would be made open immediately on publication.

In particle physics, this is currently happening under a worldwide attempt to make all the discipline’s articles free: the SCOAP3 initiative, in which thousands of institutions are paying into a €10-million annual pot that will go in pre-agreed contracts to 12 journals, amounting to €1,200 per paper. Oddly, many of these paid-for papers would also have been uploaded by authors to arXiv in the first place.

The ERC sees today’s announcement as further vindication of its support for open access. All papers resulting from ERC grants from 2014 will have to be open within six months of publication, and the organization “strongly encourages the use of discipline-specific repositories” such as arXiv and Europe PubMed Central, an archive of full-text articles in the life sciences. ERC hasn’t said how it will enforce its open-access policy, however — a detail that will be crucial to its success.

This article is reproduced with permission from the Nature news blog and the magazine Nature. The article was first published on September 16, 2013.