The Future of Libraries is…Open

Academia has always been about generating knowledge, with libraries being the repositories of much of that knowledge. We need to both record and communicate the knowledge we create through our scholarship, and historically much of this has been through dissemination via monographs and journals. Although the diversity of outputs has always been greater than this (and is growing), these remain the primary vehicles for scholarly communication in many disciplines.

But for how much longer? The monograph (or variants of long-form communication) is as old as libraries themselves, whilst the journal article is at least 400 years old (the French will always remind us that the first scholarly journal was the Journal des sçavans, first published in January 1665, whilst the second-oldest journal – the Philosophical Transactions of the Royal Society, first published in March 1665 – is the oldest in continuous print).

Other innovations, such as peer review, are considerably more recent, but the idea that the primary mode of scholarly communication in many disciplines has largely remained unchanged for this long is striking. Even the growing number of online-only journals still publish articles that are effectively formatted as print articles, albeit rendered as PDFs. Limits on article length (word counts, display items, and so on) are largely a vestige of a time when paper was expensive and journals had to be posted to subscribers.

And yet, we are producing an ever-more diverse array of outputs that capture different elements and stages of the research process. This varies across disciplines, but we are seeing an accelerating trend towards sharing of intermediate research outputs – study plans, analytical approaches, datasets and increasingly available on a range of institutional and third-party platforms. There will always be limits to what can be made truly open, but an increasing amount of our scholarly output is at least FAIR.

Despite this, we continue to distil this rich scholarly process – and the outputs that capture it – into short- or long-form summaries. And we continue to entrust publishers with the Version of Record (I won’t open the can of worms that is for-profit scholarly publishing, transformative deals, gold open access, and the rest). Surely, with the technology available to us and the platforms that already exist, we could conceive of a new model of scholarly communication – one with institutional libraries at its heart.

The idea of shifting from a Version of Record to a Record of Versions – a dynamic and updating web of knowledge that links intermediate research outputs as they evolve – is not a new one. But it is only recently that attempts have been made to develop the digital architecture to support this, such as Octopus, funded by Research England. But for this to evolve into a new model of scholarly communication will require more than just digital architecture.

Who owns this record is key.

Institutional repositories could be integrated with research workflows – certainly in some disciplines – allowing an almost real time record to emerge. Institutions themselves would be guarantors of the output of their academics, providing an incentive for them to ensure the robustness and integrity of the workflows that generate the deposits. We would move from a model of peer review focused only on the end product to one of assurance and provenance focused on the whole process.

Of course, we would still need journals. These could provide key functions that a repository alone cannot – searchability, exegesis, critical review, and so on. In fact, exactly the functions that professional editors with academic qualifications are well placed to deliver. And functions that organisations would pay for. But the scholarship itself – and a complete and evolving record of it – would remain within the institutions that generated it.

And the benefits from this could be enormous. Ensuring the robustness and integrity of research workflows would drive quality, accelerate innovation and protect against fraud (in turn protecting institutions against the financial consequences of fraud). Making intermediate research outputs more widely available would enable their reuse by civic partners such as local schools. And the transparency on which such a system would be built would itself drive economics benefits.

There are moral reasons to reform our model of scholarly communication to be fit for purpose in the 21^st century, and better capture the dynamic and evolving nature of knowledge. But there are also instrumental ones – hard-nosed benefits, financial and reputational, that can be realised by building a system that relies on transparency and institutional accountability. It won’t be easy, but we can’t continue to rely on a model built (literally) on communicating via dead trees…