The once and future library: will there be, and what might a research library look like in 2030?

Editor's Notes

• #3: It is notoriously difficult to predict the future. For this presentation I will take a relatively simple approach: look at trends that we already understand and fast-forward them to 2030. While this misses out on developments that may just barely be visible on the horizon, it should enable a discussion around the known unknowns – and of course how I have chose to interpret them.
• #4: Globally, the percentage of open access content increases every year. Now we have another push from funders around Plan S, there are better workflows and a younger generation of researchers who seem to embrace “open” even more. Therefore my assumption is that by 2030 open access will be the default scholarly publishing model, even though there are likely to be some challenges still around monographs. Also, I assume there will still be commercial databases and publications that will be funded by libraries on a subscription model. Even so the change will be substantial enough to, largely, bring the role of research libraries as providers of access to scholarly content to an end (certainly in the sciences, somewhat less in the arts and humanities and in particular in historical disciplines. This does not change the overall message though.
• #5: Currently, research libraries are on a transition from a licensing model for content to one where they facilitate open access through processing of article processing charges, advising authors and providing a green open access infrastructure. Our role in discovery is already shrinking, but with a range of powerful free discovery tools from Google to Unpaywall there may not be much of a role for research libraries in content discovery (beyond special collections etc.). A few larger research libraries run OA services, such as BASE, but there won’t be many such aggregators across the sector. There is another potential challenge though: we already see funders running their own publishing platforms. I can easily envisage a model where funders would either make this the default or where payment processes are so advanced that there is little work for librarians here. With another decade of user experience enhancements, training needs may diminish too. APC payment processes are highly inefficient, but it isn’t hard to envisage a model where they could be largely automated, along with compliance checking. So potentially, research libraries could become victims of their own success. This is a good thing for open access, but likely to change our role.
• #6: There will still be a demand for contemporary material that is commercially produced and not available as open access, for example text books and various databases. Libraries will continue to license this type of material.
• #7: Another example where we could become victims of our own success is persistent identifiers. If we assume current efforts to fully embed ORCID, DOIs and other identifiers for funders, organisations, instruments, collections etc. into workflows, the active role of libraries will be much reduced. Not only will we not be needed to enhance metadata as much as we need to now – this will all done by machines/software at the source, with researchers – all the PID information would be embedded in objects and the various PID registries. PIDs need to be generated at the point where data is generated, and that isn’t usually in or related to a library. This is likely more data than each library would want to hold, so our role would be to keep local copies of a subset of the data. This won’t be of much practical use beyond preservation/backup, so local systems would at best be an interface to the full record living in the cloud. Our main role would therefore likely be to handle the subscriptions to these services on behalf of universities.
• #8: Looking at open science much broadly, I would argue the challenge is too big for any one organisation to address. We will see this market mature and consolidate, and much of the current playful exploration in which libraries engage will come to an end. There will be a mix of discipline-based services and a few service providers who can offer platforms at capacity that we cannot match. So again our role would be to pay licensing fees for commercial services. There is probably a role supporting the long tail though, and training.
• #9: Currently there is a lot of talk about supporting researchers on mobile and different devices. Arguably, the device will increasingly become less relevant, it will be the software on the device, or rather the software in the cloud the device talks to. Conceptually, we will have to shift from a user interface design to interfacing with software. Whatever the software our users use can’t interface with, even if it is just Google for discovery, will become invisible. We already now see younger researchers increasingly expecting that everything is remotely available and of course digital, and that trend can only become stronger. This will eventually mean the end of the catalogue as a user tool and involve us pushing (or at least opening up) all our data to the cloud, for others to build tools around it.
• #10: It seems very likely that at some point all our data will be in the cloud, processed and, via services, fed to us from companies who use machine learning and AI technology. In some areas that has already happened, but we are moving that way for library services too. This will likely ring the death kneel to cataloguing as a task done in libraries locally and/or by humans, perhaps with the exception of new heritage acquisitions and some print material. There will be too much content for any one library, and systems will rely on users and software doing most of the work (also consider the impact of PIDs). Libraries that still have hardware will phase that out, but the same goes for locally administered software. Procurement of services will become an even more important task, with some local level of promotion/training perhaps.
• #11: Arguably, we already struggle with running software in research libraries – just consider the out of date state of many repositories – and with increasing market concentration in tech our ability to produce anything meaningful will be reduced. If we were more agile we could still make very focused contributions though. Arguably, working with vendors and shaping their offering may be our best bet.
• #12: Academic publishers are actively preparing for the future that I have just outlined. We are currently dependent on them for content. In the future there is a chance we may be dependent on them for software. This may result in a more competitive market, but how much value we can still as systems and approaches mature and consolidate add remains to be seen.
• #13: Libraries have, and I think will continue to have a role in facilitating dialogue and raising issues related to information management. This is likely to continue, in particular when we pick up on the ethical issues of the new information age. However, this will require us to have the rights skills.
• #14: Again, we could be a victim of our own success here. Where we have fully digitised our heritage collections and made them available under open licences they can go wherever our users need them. We will still have a preservation copy somewhere – more likely at a service provider though – but our role in access may be limited. Only where our user experience is bad, where we fail our users, i.e. where content has rights restrictions or is print only will users still have to come to us directly. One area where the original will remain relevant is for exhibitions though. Arguably, those may become more popular as the original analogue copy will feel even rarer in a world full of digital.
• #15: Whatever may happen in the digital realm, we can confidently assume humans will still have physical bodies in 2030. So we will still need places to work. In fact, experience from many libraries suggest that digital content available online does not lead to reduction in visitor numbers – quite the opposite in some cases. Even where our collection items aren’t consulted, university libraries in particular provide work and social space to users (notably students). There are no signs of this trend being reversed, although we will have to continue developing our spaces alongside evolving user needs. Libraries have hundreds of years of experience in this though. Another activity that is set to become more important is engagement with local libraries and communities. Higher education libraries are well placed to become interfaces to local communities.
• #16: While we can hope the usability of future systems will be better, digital literacy skills will still be in demand – and may even be more essential than now. That will require us to keep training our staff and build more expertise than we currently have – and with current salary structures we often can’t afford it.