If universities are so concerned about the subscription prices and article fees charged by publishers, why don’t they disseminate their research for free via their own repositories?
One answer to this question is that researchers want to target their papers at others in their field, but some disciplines have their own subject-wide repositories. The most celebrated is arXiv.org, which covers an ever-expanding range of fields within physics and mathematics.
Many researchers in those areas - such as former Royal Society president Lord Rees of Ludlow and Sir Timothy Gowers, the Fields medal-winning mathematician based at the University of Cambridge - already do most of their reading on arXiv. Yet even mathematicians and physicists still feel the need to publish their papers in standard journals, too.
The reason is that the established pecking order of journals confers academically vital prestige on the papers they publish. Even Sir Tim admits to “taking note” of where a paper was published when he is sitting on a hiring committee.
But might it be possible at least to reduce the cost of journal publishing by taking it out of the hands of commercial publishers and using arXiv itself as the publishing platform? This is precisely what the Episciences Project aims to achieve by establishing a series of “epijournals” in maths.
The project, expected to launch soon, has been developed by the Fourier Institute’s mathematics laboratory, part of France’s Joseph Fourier University (Grenoble 1), with support from French open-access facilitators the Centre for Direct Scientific Communication (CCSD). According to Sir Tim, who broke the news of the project in a January blog post, “the idea is that the parts of the publication process that academics do voluntarily - editing and refereeing - are just as they are for traditional journals, and we do without the parts that cost money, such as copy-editing and typesetting”.
The project’s leader, Grenoble 1 professor of mathematics Jean-Pierre Demailly, said that publishers “do not offer substantial added value considering the high price they charge” and represent a “barrier to the development of modern tools for indexing and searching data”.
Spotlight on pre-prints
For him, the only useful function performed by modern journals is to organise peer review, and this is what the epijournals - each of which will have its own editor and editorial board - will do. The handful of epijournals (Professor Demailly estimated between five and 10) that will launch at the same time as the platform itself will initially scour arXiv for high-quality “pre-prints”: the draft versions of papers that researchers often post for comment prior to journal submission.
Final versions of the papers, incorporating referees’ suggestions, will then be reposted on arXiv and linked to by the epijournal.
Professor Demailly hoped that once the journals became accepted, authors would proactively send manuscripts to them (although they will be allowed to submit them to standard journals as well).
He also hoped that the presence of four Fields medallists (including Sir Tim) on the committee overseeing the Episciences Project would confer instant prestige on the journals, to be reinforced by the “demanding” selection standards they would adopt.
“Of course, if Episciences is to become big, which I sincerely hope it will, it will be necessary to accommodate a larger part of scientific production,” he said.
For this reason, the journal platform his team are developing will allow editorial board members - and possibly others - to post reviews of papers highlighting the “added value” of each.
But significant expansion would come at a “substantial” cost, he admitted. He hoped the Episciences system would minimise administration costs by permitting more efficient communication between authors, editors and referees, but international backing of the kind enjoyed by arXiv was likely to be necessary.
Sir Tim said that any radical departure from the traditional journal format - including use of the term “epijournal” - would initially be avoided so as not to jeopardise the epijournals’ acceptance. He called on leading mathematicians to declare publicly when on hiring committees that they would take epijournals as seriously as traditional publications - “and there are many who are ready to do that”.
Although the platform had been designed with the needs of mathematicians in mind, he said, it could be applied to any subject that “doesn’t require significant processing of author manuscripts”, such as preparing illustrations.
Indeed, some subjects have already attempted their own small-scale epijournals - often known as “overlay journals” - that bring together papers published elsewhere. However, Frederick Friend, honorary director of scholarly communication at University College London, said the most prominent UK examples, such as the Overlay Journal Infrastructure for Meteorological Sciences, had not survived recent changes at their funder, Jisc.
“We have learned…that success depends upon a partnership between a substantial source of articles of value to researchers and a substantial organisation able to set up and maintain the overlay journal platform. The Episciences project has an ideal source of articles in arXiv and an ideal organisation in CCSD,” he added.
Cameron Neylon, director of advocacy at PLOS, said the atmosphere of rebellion over open access, engendered by last year’s boycott of the publisher Elsevier inspired by Sir Tim, only enhanced Episciences’ prospects. He described the project as “potentially highly disruptive because it decouples the publication…from the peer-review platform and allows them to evolve separately”.
“I certainly think it poses a real threat to more traditional mathematics journals,” he added.
However, Michael Mabe, chief executive of the International Association of Scientific, Technical and Medical Publishers, said it would be a “gargantuan” task for an epijournal to sift the entirety of arXiv for pre- prints worthy of being “plucked out for greatness”, and he was not clear what would motivate editors - particularly if they were not even offered an office or expenses. If, by contrast, an epijournal merely courted submissions, it would be “just another journal”, he added.
Mr Mabe also worried about the potential for confusion if multiple epijournals each created their own versions of the same papers.
He said that mathematicians often favoured “non-journal solutions” because peer review in the discipline - which often involves redoing the work - was very different from that in other fields. However, he added, the practical and financial difficulties facing Episciences would be “challenging” and he wished the project luck.
Even Sir Tim cautioned against “over-hyping” the project.
“There are a lot of people who hope it will lead to mathematicians taking publishing into their own hands. But there are many difficulties and uncertainties, so it would be foolish to make any predictions,” he said.
“What I can say more confidently is that if we are to get to a position where we are no longer reliant on commercial publishers, then we need experiments like this. So the fact that a very serious experiment is about to start is extremely good news.”