If they cannot be easily located, referenced again later, and connected with other relevant scholarship, expansive digital publications easily become silos that fail to connect with existing scholarship and are not likely to have much impact in their fields. Discoverability in digital publishing is both paramount and complex, as digital scholarly publications are often aimed at multiple audiences and rely on both traditional and digital, scholarly and commercial, and human and machine apparatuses in order to reach those users. At present, the discoverability of expansive digital publications is characterized by gaps and loose ends.
Gaps in traditional scholarly communication. While traditional scholarship relies on online databases and reviews in journals and professional associations to help with diffusion and visibility, exapnsive digital publications lack comparable infrastructure. Some major journals, such as The Journal of American History, regularly review digital publications, but many do not. The scholarly practice of citation is another key component of improving discoverability: more citations lead to greater visibility for digital publications. Scholars who hesitate to cite digital publications further reduce the likelihood that these works will be discovered through traditional means. Yet scholars are understandably reluctant to cite digital publications that do not possess a stable URL (and in some cases, may prefer to cite a print analog rather than risk link rot). It also may be unclear what, exactly, should be cited — the work itself? a subsection? — or how to point others directly to the idea or information that is referenced. Such challenges are compounded for expansive digital publications, whose components may change and migrate over time as the work matures.
Gaps in libraries’ cataloging workflows. In addition to being underrepresented in scholarly literature, expansive digital publications rarely show up in library catalogs and registries. Interviewees for our study lamented that libraries are not well equipped to manage digital objects. “Unlike with a purchased physical book,” explained one interviewee, “there is no ‘trigger’ for getting scholarship on the open web into a cataloging workflow.” To carry that metaphor further, one might argue that there is no “barrel” either, to direct cataloging once the trigger is pulled. The fluid and open-ended quality of expansive digital publications makes them difficult to classify within existing genera and thus into appropriate cataloging channels.
Workshop participants also agreed that citation of digital publications remains a very thorny issue: scholars want a persistent identifier/locator (such as a DOI) in order to cite digital publications, but the creation of a unique identifier can also become a trap. “Should we assign DOIs or ARKs or ONIX metadata to everything to make them all citable and more discoverable?” posed one participant. “Therein madness lies. Some things may not deserve DOIs or ARKs because they may be too atomic and not worth it.”
Limitations of web-based and commercial discovery tools. Digital publications are unlikely to be found in the commercial indexes that libraries subscribe to. Discoverability services are often provided by content aggregators, who fund this service through sales commissions; open digital content breaks that model. Many workshop participants pointed out that while library catalogs and publication indexes still matter, they no longer serve as the starting points of discovery for many users (including scholars).1 We need to develop cataloging strategies that can make expansive digital publications discoverable on the open Web. For instance, describing content with open schemata, such as those expressed by Schema.org vocabularies, can improve machine discoverability beyond the library catalog using tools like Knowledge Graph.
Yet while our interviewees and workshop participants expressed strong agreement about the importance of integrating digital outputs into popular discovery tools, they also raised serious concerns about the difficulties this work entails. Users may treat platforms like Amazon and the Apple Store as discovery tools, but these platforms don't consider themselves as such. In order to be sufficiently “discoverable” through these channels (e.g. Portico, Amazon, Gobi), a digital publication must conform to criteria used for searching and browsing within that platform (e.g., price; deliverable format) and may also need to meet further requirements specific to the platform. Workshop participants also lamented that Google Scholar, in the words of one interviewee, “is much more trouble than it’s worth. They dictate to publishers how to design article pages; they won’t index you unless you follow very specific rules. Google Scholar has become a set of cuffs around our hands.” They also noted that Google Scholar is not stable or reliable, as it is subject to discontinuation as a service, just as with many other “free” web services.
Project creators’ approach to discoverability. Poor discoverability may also stem from shortsighted design. As indicated in some of our interviews, scholars planning expansive digital publications rarely consider how the audience will discover the work. One possible explanation for this oversight lies with the lack of incentives for considering discoverability. In some cases, review processes (for example, as part of grant funding) force consideration of audience at an early stage. In the absence of these external mechanisms, however, scholars have few prompts to plan early and carefully how to reach their target audience. To be fair, audience and design considerations are not typically the creator’s responsibility: in traditional publishing, scholars rely on existing infrastructure (e.g., publisher markets; library catalogs; journal reviews; citations) to connect their work with audiences. Even if they recognize that discoverability of expansive digital publications is an issue they must address, scholars have little experience or training to guide them.
Our study points to several ways libraries can play a role in improving the discoverability of expansive digital publications while strengthening their credibility and worth.
Introduce "design thinking" in the planning process. By harnessing design thinking as a method that considers users’ needs, librarians can help creators integrate discoverability into the design of their publications from the outset.2 As part of this process, research services can help creators develop a clearer sense of their potential audiences, and consultants can help with planning effective use of metadata in order to maximize discoverability through existing platforms. Relatedly, this process could help to identify the unit(s) of discovery for the project and ensure that they contain stable and citable URLs. Here, again, training can be another key intervention, by helping creators employ design strategies that improve discoverability.
Leverage intermediary relationships. Libraries are substantial customers of the information intermediaries who contribute to the discoverability of conventional formats, so they could use their financial influence to encourage such services to make less familiar formats discoverable. For many libraries, the most significant obstacles to this approach would be internal: rethinking and better connecting the work of acquiring and providing access to content with the role of facilitating scholarly communication.
Make the research process discoverable. While the expansive digital publishing project exists primarily online, artifacts of its development may take more traditional forms (such as articles or conference presentations). Libraries can help authors document and demonstrate their research process, exposing the many steps and various components that often comprise an expansive digital project. In addition to increasing the likelihood of other scholars discovering these publications, these components can offer a behind-the-scenes look at how an expansive digital publication develops and can open new avenues for scholarly conversations.3
Toward a taxonomy for expansive digital publications. Libraries should foster a holistic dialogue amongst scholars, indexers, and cultural institutions (e.g., NISO) to define an overarching taxonomy for digital publications that can effectively describe the landscape of digital objects. This taxonomy can become the basis for developing functional descriptive metadata requirements that will lead to improved discoverability for individual digital publications.
Bridging the gap between publication creators and academic publishers. Libraries do not always find themselves in the best position to help creators increase the visibility of digital publications to a larger public. Indeed, academic libraries often focus on serving their home institutions rather than performing outreach beyond the immediate community. However, they typically have comparative advantages in connecting project creators to academic publishers, who possess the expertise to create marketing strategies intended to reach new audiences, and can serve as valuable partners.
Working with tool providers. Encouraging cooperation with programmers and tool designers was another theme that emerged during our workshop. Tool providers serve as essential collaborators in the search for better forms of citation in digital publications. By incorporating feedback from digital publication creators, libraries, and funders, programmers can develop better discoverability features for existing platforms as well as encourage better practices for improving discoverability and diffusion.