As technology allows us to reach more people than we realize, how can public historians ensure we’re connected with our publics? Ironically, technology allows us to maximize access to archival sources while also preventing us from engaging with our publics. Digital archival projects must be portable, discoverable, and accessible in order to meet the public where they are. At the Digital Public Library of America (DPLA), I apply our mission of maximizing public access to reliable information and to our shared culture to a specific format: books. In highly popular – and sometimes high-profile – archival book digitization efforts, the public’s needs and interests can get lost in the scale of the projects; quantity alone does not demonstrate usefulness to the public. When creating creating large-scale digital humanities projects serving the public at a macro level, engaging expert intermediaries who can help us at the micro level can lead to a better way of serving our public.
First, some background on DPLA and ebooks: Launched in 2013, DPLA’s platform connects users to more than 20 million sources from America’s libraries, archives and museums. At DPLA we’re currently developing Open Bookshelf, a substantial, free, and open collection of widely-read and widely-held ebooks, available through the public domain or open access licenses. This collection is designed for library use, with a goal of improving discoverability through metadata and curation. Our vision is that these books will be curated by librarians for a library audience; rather than add every public domain or open access book possible, the collection will be thoughtful, deliberate, and useful. These titles are added to Open Bookshelf by members of the Curation Corps, a group of librarians with a broad range of experience in collection development and with serving diverse communities of all ages. A distributed network of librarians nationwide will then select the titles that meet their patrons’ needs (through the new DPLA Exchange), and patrons can easily discover these sources alongside trade publications.
One of the benefits/issues with a digital archival books project is the potential size of our audience. We can optimistically say our projects serve anyone who can find us with Internet and device (itself another problem), but our intended audience must be more deliberate. With an end-user base of all public library patrons in the country, we needed to break down our audience for Open Bookshelf. Borrowing from software developers, we’ve created user stories to envision who might be using a particular content type, on which device, and with what technical features. For example: “As a community college teacher, I want to access open textbooks so I can assign them to students at no cost on all devices and with note-taking.” Marketing Open Bookshelf only to librarians allows us to trust the expertise of the librarian while also making the project scalable for serving a national audience.
What will it look like when we successfully serve a national audience? One obvious criterion for success seems simple enough: use. Analytics could offer such clear-cut metrics; for example a web page for an individual book title was visited X times, or the book was downloaded Y times. Yet this doesn’t tell us the full story about our potential reader. Was the user deeply absorbed in a learning more about a book, were they confused by how to access the file – or did they walk away from the computer for 4 minutes, 34 seconds? These metrics are a helpful start, but do not give us a complete picture of our public.
Loftier – though harder to measure – criteria for success could be that the project made the books easier to access, or that they inspired a love of reading. Because libraries have a rich tradition and strong commitment to protecting user privacy, it can be difficult to learn more about patrons beyond usage metrics. For Open Bookshelf, we will only know which books a librarian added to their collection, but not whether they’re being read repeatedly by those patrons, and we certainly won’t know the book’s impact on the readers. This is another area where a relationship with the expert intermediaries is useful: we will rely on regular interactions with librarians to learn more about how patrons are using the books. The method is less tidy than collecting data on readers, is subjective and relies on self-reporting, but it helps to complete the story the analytics don’t tell.
Similarly, data about a book or collection has been hugely helpful in learning more about what our publics enjoy reading. To maintain our goal of a deliberate Open Bookshelf collection, I’ve looked at data on the most circulated books from public libraries in the US, the most accessed books on sites like HathiTrust, and the most assigned books thanks to data from the Open Syllabus project. Data quickly offers an overview of user behavior, but it can also reinforce information. I found the usual suspects on those lists. But what about authors and books from the past that haven’t benefited from decades of pop culture relevance? A librarian recently introduced me to Phillis Wheatley’s Poems on Various Subjects, Religious and Moral (1773), a collection that I now feel must be included. Without that personal interaction with an expert, I may have missed this book of poems written by a slave woman in the eighteenth-century – something that the public would likely find interesting. What else are we missing by only following the data? How do we get more information without talking to every librarian in the country?
An important barrier between our digital archival projects and our public is the digital divide. Yes, making information available online is critical in making it accessible – but only those with the means to do so. Solving the digital divide may seem like something outside the scope of a digital humanities project – there is only so much grant funding, after all. But there are things we can think about to meet people where they are and help make our information truly accessible. Does the project require an application that consumes large amounts of mobile phone storage, or data usage? Is that app only compatible with the latest version of iOS? Can the content be translated to other languages? Is the website mobile-friendly? Can the content be accessed in a web browser? Is the content accessible on a screen reader? In building our projects (and in seeking funding for them), we must consider where our public is, and not ask them to come meet us on our terms.
One way DPLA has fostered buy-in from the public on our projects is by having our users help build the DPLA with us. Because we are a small organization building a large project for a nationwide (and then some) audience, we rely on a community of users to help us spread the word and make the DPLA even better. We have strong relationships with our content and service hub contributors, and trust them as the best people to facilitate meaningful relationships with libraries, archives, and museums within their region. Our Primary Source Sets are compiled by the Education Advisory Committee, a group comprised of the project’s target audience: teachers. We have a strong developer community who have used our open API to make fascinating (and often hilarious) apps. These are in addition to our Curation Corps, who select content for Open Bookshelf and for our K-12 library app for low-income readers, Open eBooks. This method can still raise barriers: some organizations may be unable to participate because of lack of staffing, which we try to help where we can; and while we offer honoraria for experts helping us, working with us still requires free time outside work.
There are many fantastic examples of digital humanities projects engaging their publics to help them build something together. Crowdsourcing projects allow the public to contribute raw content or technical skills to improve a project. As we continue learning from the computer sciences we’re also building user communities around the open-source software and open data powering our projects. While the digital component remains exciting, our humanities training is equally and critically important. Mass digitizing humanities texts and making them available online does not make it a humanities project – it needs that layer of expertise, of curation, of annotation, of interpretation. It also needs to meet a need of the public’s (even if that need is not explicitly articulated), and meet them where they are. Generating engagement with the public – whether through a distributed network, crowdsourcing, usability testing, surveys, etc. – and then translating the final product for broadest possible use by the intended audience will help ensure a digital archival project is both useful and used.