Person reading in library

Publishing the thoughts of peer reviewers could help science, but is not popular with everybody.Credit: Getty

When Nature asks experts to review manuscripts for possible publication, we promise that the reports they send back will be kept confidential. But should we? This week we publish a Comment article that comes with a provocative challenge: more journal editors should commit to publishing peer-review reports. Doing so, the authors argue, benefits science. It puts published work in useful context and helps junior scientists to understand how review works.

Nature and the Nature research journals have long welcomed suggestions to make peer review work better for the communities we serve. In 2016, Nature Communications started to publish referee reports — with names removed — as long as the authors of the papers agreed.

The reaction has been instructive. For one, it demonstrated that authors in specific fields of the life sciences are more likely to welcome such openness. Take-up from those in other disciplines, including many in the physical sciences, has been much slower. In fact, Nature Communications lost several reliable reviewers in chemistry when the referees were told their unsigned reviews would be made public if the author opted for it. They resented not having a say in the process, and felt that their reports would have little value outside the small intended audience.

As such, Nature and the Nature research journals have no plans to make publishing referee reports compulsory for all. But we are actively exploring ways to offer it as a wider service in future, when readers, referees and authors say that they want the option.

The desire for transparent peer review is likely to vary across different communities as they consider the questions involved. Will reviewers shift their focus from a small audience devoted to improving a single manuscript, to persuading a broader audience of their own views of the topic? Will authors be as able to take criticism in their stride, knowing that it will be made public? Will the scientific community get confused by reading criticisms of an earlier draft that no longer apply? Will sections of peer-review reports be presented out of context by campaigners or opponents?

For a publisher, there are other issues to address. An important concern (as when publishing any critical and opinionated material) is the risk of libel in a reviewer’s comments or, more commonly, the inclusion in author responses of copyrighted or sensitive third-party material that helps in the assessment of manuscripts but which cannot be made public.

Some clinical-sciences journals now routinely identify reviewers to avoid charges of conflict of interest. By contrast, many journals in the social and political sciences keep authors and reviewers secret during the review process — to encourage frank reports that are not overly awed by prestige or dismissive of under-represented groups. Such double-blind review has been an option on all Nature research journals since 2015.

Nature editors find review reports invaluable. We know that some readers would find them useful as well. We hope that the Comment piece helps to stimulate wider debate. We welcome insights and feedback on this issue from across the scientific spectrum as we continue to align our own practices with the needs of different disciplines.