Frank Norman is a well-known librarian in the biomedical research field, working more than 35 years in this sphere. As head librarian at the Francis Crick Institute (a.k.a. the Crick), one of the world’s top biomedical research organizations, he serves more than 100 research groups across a range of disciplines.
With a keen interest and expertise in open science and scholarly communication, including preprints, Norman writes articles on this and related topics, serves as an advisory board member of the bioRxiv preprint server, and promotes preprints to the Crick’s scientists.
We reached out to Norman to better understand why he and his team are encouraging their researchers to post preprints of their work - and to get his insights on preprints and their role in the evolution of scholarly publishing.
From a librarian’s perspective, why should authors care about preprints?
Researchers will see faster dissemination of their research if they post preprints and can gain early feedback on their results. Early career researchers can benefit, too, as they can include their preprints on their CVs while waiting for peer review of submitted papers. And, of course, authors are also readers, so they’ll be interested in preprints as they contain the newest research findings. Preprints can help protect against scooping, too. If you put your manuscript on a preprint server, you get a DOI and a date stamp. So if another lab in the same field publishes a result similar to yours before you do, if you have posted a preprint, you can still publish your paper in one of the increasing number of journals that have an anti-scooping policy.
Why would librarians at research institutions and universities care about preprinting?
Preprints can help to create a more healthy scholarly communications system, so I think we, as librarians, need to take an interest in that. More researchers are posting preprints but there is confusion about some aspects of preprints. I often get asked, “What license should I choose?”, for example. It’s important for librarians to know about preprints so that we can answer these questions. ASAPbio is a great source of information on all things connected to preprints. Biomedical research funders mandate preprints for certain circumstances, such as in public health emergencies. Some of the research groups at the Crick turned their focus to COVID research last year. Any results from that research must be preprinted. Librarians also need to be clear about funders’ open access rules. In biomedicine, usually posting a preprint is not sufficient to satisfy funders. The accepted manuscript or version of record must also be made available openly. This point is not always obvious to researchers.
How do you educate and/or encourage researchers at The Crick to post preprints?
We run little pop-ups in the Crick’s collaboration areas, present posters at internal conferences, answer questions when people ask them, put information on the intranet, and we mention preprints in inductions for new research staff. We are trying to slip in mentions of preprints in general features on the intranet. There is a series of ‘science dates’ on the intranet. These are transcripts of conversations between two researchers at the Crick who don’t know each other. I arranged one of these to be between a preprint enthusiast and someone who had just started using them. What we need to do is get preprints on researchers’ radar. We also try to give preprints a higher profile by occasionally recommending to the communications department to highlight a preprint in their weekly round-up of Crick research on the intranet. We got high-level buy-in, too, when Crick management agreed that preprints were acceptable in job applications and lab review processes.
Preprints have existed for decades across some disciplines, like physics and statistics. Why do you think preprints are still new to some disciplines while they’re so rooted with others?
Different disciplines have different publishing habits. In high-energy physics, even before the internet, they were already accustomed to sending paper preprints, so arXiv was a natural development. A typical experiment would involve huge numbers of collaborators, and it’s impossible for an individual lab to recreate a high-energy physics experiment. In molecular biology, it was more straightforward to repeat an experiment you’d seen on a preprint somewhere so there was reluctance to release manuscripts early. Intense competition for funding in biomedical sciences research added to this reluctance to go down the preprinting route. On the medical end, there’s also concern about releasing unverified information that might mislead patients; but peer review isn’t actually a guarantee against flawed or incorrect research getting published, and it’s better to be cautious about all published research, whether peer-reviewed or not.
Some journal publishers, particularly those tied with academic societies, perceive fully open-access preprint servers as threats to their subscription revenue models. Would you not renew a journal subscription if much of that journal’s content were preprinted?
As a librarian, my key mission is to do what the researchers need. At present I believe there’s still value in having the version of record, particularly in terms of ease of finding an article. Researchers I serve still find value in subscription journals and the version of record. If the preprint becomes the primary version that people use and journal brands become less dominant, then there would be a case for cancelling subscriptions. However, we’re moving away from straight subscriptions to more and more read-and-publish deals. In 2022, I expect most of the Crick’s journals to be part of read-and-publish deals. Hence, even if all articles are issued as preprints, we will continue to transact these read-and-publish deals.
Are there any trends that you see lending steam to preprints? What are they, and why do you think so?
COVID has given a huge boost to preprints, thanks to the requirement to share data related to public health emergencies. I wouldn’t be surprised if research related to other global emergencies, like climate change, becomes subject to a similar requirement. Open research platforms like F1000Research and Wellcome Open Research combine preprint and final published versions on one platform, making preprints more clearly part of the process of publication. I do like this model of publishing. The number of articles published on these platforms is still quite small though. I also think we’re seeing more experimentation in peer-review. There are more varieties of peer-review, and that’s making people less fearful about preprints. In a world where everything is preprinted, how do researchers discriminate between what's worth reading and what’s not? Often readers use the journal brand to help them decide. If artificial intelligence or machine learning can come up with an effective way of guiding readers to papers of interest, this will make consumption of preprints more productive.
What trends in the information sciences or scholarly publishing may create headwinds for preprints in the future?
I think the worries about misinformation and conspiracy theories on medical information and our inability to control pseudoscience is problematic, because that fuels fears about preprints. I’m not a big fan of publishers trying to control preprints, setting up their own restricted access servers for example. I think open, unrestricted preprint services are far better.
What challenges can preprints create for librarians?
We maintain a record of all published research outputs from the Crick, and preprints make that a little bit harder. Knowing when a preprint becomes a published version is still a bit challenging. I think we just need to master the CrossRef API. We also need to decide whether to include preprints everywhere where we currently list published papers, like web pages for our research labs. Some authors confuse preprints and open access. They post a preprint and don't worry about making their article open access. Many of our funders require open access for the final peer-reviewed version. Also, the proliferation of preprint servers can be puzzling. Crick researchers have preprints on arXiv, bioRxiv, chemRxiv, medRxiv and Research Square. We’re used to dealing with multiple journals, so we should be able to cope with this; but we need to be alert to new preprint developments.
Do you think preprints will eventually have a permanent role in scholarly publishing across most or all scholarly disciplines? Why or why not?
In terms of most sciences, I see a continuing role for preprints. Most researchers see the benefits of preprints. Funders too can see how they speed up science. I expect the preprint habit will spread to all other science disciplines. People are much more willing than in the past to post preprints. Now we can see that maybe, for some research outputs we may not need heavy-duty peer-review processes. In some cases, we might not need any. For example, publishing null results as preprints may be an acceptable way to get those studies into the research arena. So I think we’re a little bit less tied to a rigid view of peer review now.
Five to ten years from now, what do you think preprints’ role will be in scholarly publishing?
I think there will be a variety of models. There won’t be a single route to go from manuscript to published research. There will be different routes through different servers. There will be some cases where the servers are the journals. We may see some overlay journals becoming more established, more journals like eLife which now only accept submissions that are preprints, some preprint review services will become more formalized, and some things that we haven’t even imagined yet! I think learned societies have an important role in publishing and I expect some new ideas and approaches from that quarter. Some society publishers are part of Review Commons. That’s a good development, streamlining their peer review processes. I hope learned societies will seize their moment. There’s still a lot of trust in learned societies, and they are our scientific community.
Help expand the conversation on preprints. Please share your thoughts on this interview through social media.