The open-access journal eLife has unveiled plans to introduce a new publishing model. Starting next July, the journal will adopt a “publish, then review” policy, and will make all of its peer-review reports publicly available.
Under the policy, which the journal announced1 on 1 December, eLife will only review and publish papers that have already been posted on a preprint server, such as bioRxiv, medRxiv or arXiv. Submitted papers that aren’t already on preprint servers will be posted on bioRxiv or medRxiv.
This model “amplifies the value of preprints”, says Peter Suber, director of the Harvard Office for Scholarly Communication in Cambridge, Massachusetts, and the Harvard Open Access Project. “eLife is showing that the benefits of a preprint can be conjoined with the benefits of peer review.”
One of the driving forces behind the policy change was an internal analysis, which revealed that approximately 70% of the papers reviewed by eLife during May, June and July 2020 had already been posted as preprints. “We’ve always known that the eLife community was very supportive of preprints, but we didn’t realize quite how much it had taken hold within our community until we did a survey of our papers,” says Michael Eisen, a biologist at the University of California, Berkeley, and editor-in-chief of eLife. Eisen suspects that the reason the remaining 30% of authors aren’t currently posting their manuscripts as preprints before submitting them for publication is “just a matter of inertia, rather than any kind of real opposition”.
The “publish, then review” model has already been adopted by others. The London-based publisher F1000 Research, for example, provides platforms in which academics post their manuscripts before receiving public reviews; only those that pass review are indexed in databases such as PubMed. But eLife’s policy combines this idea with the conventional journal system: the journal will publish only the manuscripts that pass its review process. The benefit of eLife adopting such a model is that academics don’t need to abandon the conventional journal system, Eisen says. “I think it is essential, in order to bring as large a fraction of the community along with us as possible, that we not force people out of [the existing] model while they come into the new one.”
The new policy won’t come into effect right away. Over the next six months, authors will be able to opt out of posting their submitted manuscripts as preprints, but will be asked to explain why.
eLife also plans to start posting all of its peer-review reports on preprint servers, whether or not a paper is accepted for publication (although authors whose manuscripts are rejected will be allowed to delay the posting of their reviews until their articles are accepted elsewhere). The journal is also developing a platform, Sciety, for sharing public peer reviews.
The authors of reviews will remain anonymous, in response to feedback on the idea from early-career scientists and other academics who felt they’d be unable to provide honest critiques in identifiable public reviews for fear of retaliation.
Although a system in which reviewers are completely anonymous is not ideal, eLife and other publications are able to provide “authenticated anonymity”, Eisen says. “This is the critical thing that organizations like a journal can do — being careful about who we pick as peer reviewers to make sure that they actually have the relevant expertise, that they’re not conflicted and that only reviews that are fair and constructive get posted.” Currently, eLife, which is partly subsidized by private research funders, charges US$2,500 to publish a paper, and that fee will stay in place for the time being, Eisen adds.
“eLife’s new model is an important development in making preprint sharing, open peer review and availability of reviews with preprints the default, and thus an important contribution to culture change in scholarly publishing,” says Bianca Kramer, a librarian and scholarly-communication researcher at Utrecht University in the Netherlands. “At the same time, the model in its current implementation still relies on the selectivity of eLife as a journal.”
Eisen also thinks that a publishing model is flawed if it treats the journal in which a paper is published as the primary measure of quality. He says that as eLife adopts its new reviewing policy, it also plans to develop other evaluation metrics. “Getting people away from valuing a publishing decision as a once-in-a-lifetime event that dictates a paper’s value is what we ultimately hope to do,” Eisen says.
Eisen, M. B. et al. eLife 9, e64910 (2020).