Moving toward digital libraries
How one man’s vision for a digital library shaped how a historic service can be rethought for the current age.
a conversation with Klaus Kempf
Klaus Kempf has been a protagonist in the digital transformation of libraries over the past 20 years. A traditional librarian both in terms of education and spirit, as well as a lover of books and culture, Kempf has promoted change from within, without being a digital hypebeast for digital’s sake.
A.B.: Klaus, you have been the director of Munich’s new Bayerische Staatsbibliothek (BSB) digital library since 2011. It is one of Europe’s most important libraries. Can you tell us what a digital library is, exactly?
K.K.: We talk about “digital libraries” as an evolution in the history of libraries. As the world of data and communications began its transformation toward digital in the 60s, we first had “automated libraries”, in which computers were used as work tools to automate specific tasks. Then we moved on to “hybrid libraries”, as we talk about them today, where both traditional, analog media and modern, digital media live under the same roof dynamically. In some of these hybrid libraries, such as university libraries and STM (Scientific, Technical, and Medical) ones, digital content is already largely dominant, which means those are the ones closer to the idea of a fully digital library. This is a broad overview, but of course, it largely depends on how each library interprets the idea of “digital” individually.
A.B.: What are the essential elements needed for a digital library to work well?
K.K.: Whether a library is focused on the production of digital material or the handling of digital services that offer access to third-party resources (and/or open access publications coming from a headhouse, like a university), there are three key elements that need to be considered: first is having expert collaborators in-house; second is strictly keeping the end user’s point of view in mind when ideating, designing, and developing new services, while considering how they are going to use (and reuse) the resources; last but not least is a magic word, cooperation.
A.B.: I’d like to hear more about that magic word, if you don’t mind. Who needs to cooperate with whom, and why?
K.K.: Librarians and libraries can rely on a traditional culture of cooperation. Well before the digital era, cooperation was already a sine qua non condition for many services. Let’s just think about interlibrary loan (ILL), born in Italy after the Risorgimento (the mid-1800s) and then “copied” all around the library world. With the arrival of digital services and the explosion of the internet, cooperation and collaboration — frequently referred to as networking — have become standard procedures. With the aforementioned hybrid library, cooperation had already moved beyond mere collaboration between libraries: data centers (data maintenance and infrastructure management), I.T. companies (software development and management), and other institutions (book conservation and restoration) have all become indispensable partners. But these partnerships go even further, looking at the world of business, like in the case of Google (because of Google Books). And so other than cooperation, competition also grows, which is new. Competition between libraries and big players like Google and Amazon (which started out as an online library!), but also online, non-profit aggregators like the Internet Archive and HathiTrust. The latter, however, are also ready to collaborate or even enter institutional partnerships with libraries. Cooperation and competition coexist, and alternate.
Andrea Bolioli: The second thing you mentioned is the user’s point of view. Disciplines like UX and user-centered service design have become critical in many fields. Have libraries not had users as their main viewpoint from the get-go? What has changed in the move to digital?
Klaus Kempf: In the analog era — i.e. the print world, also known as Gutenberg Galaxy — libraries were mainly information archives. Yes, there was cooperation and services like the aforementioned ILL, but each library was focused on and organized according to its own needs, not the overall network’s; let alone people’s. In the digital era this has changed: services are now tailored to the user’s needs, due to competition, if anything. This is also new: users have largely cut loose from their libraries. Because of companies like Google and Amazon, the offering of libraries has had to change and adapt too.
Completely outsourcing would mean wasting the opportunity to learn on the job all the things a digital library actually needs. I think there is a better strategy.
A.B.: It is important for digital libraries to deal with the use and reuse of content. Why digitize and archive everything if something is not used, then?
K.K.: The digitization of a digital library, and more specifically its management — archiving material, keeping services up and running, etc. — is costly; much more than a traditional library, in fact. However, the end service is usually of much higher quality. And so marketing these services and measuring their success is absolutely essential; it would be art pour l’art if it weren’t. Keeping track of the use and reuse of the available material and data is a key indicator of the information’s overall quality and people’s satisfaction with it; and, in turn, a good way to understand how good the service as a whole is.
A.B.: Let’s circle back to the first thing you mentioned, the importance of having expert collaborators in-house and so not delegating too much. What are the necessary skills, and what should be outsourced instead?
K.K.: When it comes to the digitization process, all libraries face a shortage of human resources in both specific skills and overall numbers. It thus becomes easy to give in to the idea of delegating, especially when stakeholders offer to chip in with some extra cash and good options on the market come to the surface. However, completely outsourcing would mean wasting the opportunity to learn on the job all the things a digital library actually needs. I think there is a better strategy, which is what we have done at BSB.
A.B.: Tell us a little bit about that. You have been the head of BSB and have led the largest library digitization project in Europe. What are some of the biggest challenges you have faced?
K.K.: In a nutshell: restructuring and reorganizing the entire library; designing and developing new work processes; retraining staff and collaborators. We didn’t have trained personnel at first, so we had to follow a double strategy: on the one hand, we (successfully) tried to buy projects using third party fundings (mainly from the Deutsche Forschungsgemeinschaft, DFG); on the other, we kickstarted an extensive, profound reorganization of the entire production part of the library, i.e. the purchase and cataloging departments (including subject indexing and categorization), in order to rationalize existing workflows and free people to work on other things. The introduction of elements of project management has allowed us to become more flexible, even less hierarchical to an extent, and to give birth to the “independent” collaborator, who is none other than someone with sufficient skills and education to act independently on projects (according to their role inside the organization).
A.B.: You have collaborated with Google on this project. What have the advantages and limitations of this partnership been?
K.K.: It has been a very well-defined collaboration on the Google Books project, one that has seen them work with other libraries around the world both before and after us. We have mainly worked on printed material, like books and magazines, with no copyright, from the 1600s onward. Google allows to enter a different dimension of digitization, a “mass industrial digitization” of sorts: in BSB’s case, for instance, we were talking up to 5,000 volumes a week. Google scans the material and creates a digital copy, whose quality gets better over time — moving from black and white to color, for instance. But the real advantage is that Google offers texts that are both full and fully searchable. And that absolutely sets the gold standard for digitization projects like this.