Imagine you’re holding a book– an illustrated novel published in Philadelphia in 1770, filled with marginalia. Where did you find this information? Or, how did you find this book? Most likely, by using library catalog or database. But these digital databases are not always ideally designed. Often built to mimic an older, analog model, the card catalog, a quick search through their records often misses valuable information–what a book contains, where is it from, what its social history is, etc– either because that information is hidden in labyrinthine layers of interface or missing altogether. That’s where my work comes in.
This summer I’m working with the Early Novels Database (END), a database based on the Singer Mendenhall collection of 17th, 18th, and 19th century novels at UPenn’s Van Pelt-Dietrich Library. Among other things, the database allows you to search ways that early novelists relied on “paratext”– prefaces, editor’s asides, footnotes, tables of contents, etc – to define the nascent genre. As webmaster and social media coordinator, I’ve spent the summer translating this information into a new digital form and building a social media presence to share the social life of the book that is captured in END’s records.
My main project has been building the framework for the new END website, which will be live by August 16th. Like 18th century novelists, my job is to keep readers interested while they navigate through an unfamiliar medium – for early authors, the novel, in my case, the digital database. Building a new interface has meant ways to display and present this information digitally has had me thinking quite a bit about access and what it means to have “open data.” Digital databases are often difficult to navigate, contain minimal or highly specialized documentation on how to use them, and, frequently, lie behind pay walls accessible only to scholars and individuals with the economic and institutional privilege to bypass them. Even “open data” often translates to data available in idiosyncratically encoded file, often with multiple levels of encoding–another tremendous barrier to anyone without some background in computer science or programming. Rethinking database design in order to cater towards wider access has meant juggling aesthetics, usability, complexity, and users’ varied levels of expertise – from librarians to early novel scholars and general readers.
Driven by the desire provide multiple avenues into our database for these different audiences, I’ve been experimenting with “novel” ways of presenting bibliographic data. In addition to a traditional, library style-keyword search, I’ve developed a web interface that integrates multiple social media–Twitter @EarlyNovelsDB, Storify, Flickr, and SoundCloud– for accessing our data, transforming bibliographic description into tweets, images, narrative, and audio clips. The interface will include both downloadable files of the dataset, as well as a simple environment for users to visualize and analyze the data. Because of barriers to access for many digital tools/interfaces, I have tried to create a sandbox way that allows users to freely interact with the data and presents a large amount of data in a navigable and friendly medium.
While coding the website and curating our social media presence, I also cataloged novels, which made quickly realize the extent to which eighteenth century novelists were also deeply concerned with information management. Samuel Richardson’s The History of Sir Charles Grandison and Memoirs of the Year 2500, for example, both feature myriad footnotes and indexes. In these and other 18th century novels, authors include paratextual directions on how to read (read: navigate) the texts that invite the reader to conceptualize the novel not as a single entity, but as a composite whole, filled with individual, pieces of data.
Working to put previously inaccessible data at users’ fingertips has made one thing incredibly clear: there is quite a lot of labor required to produce easily accessible data. When I catalogued Memoirs of the Year 2500, I spent most of the morning individually photographing and cataloguing the book’s 300 footnotes. The exact location of these footnotes will now be available in the novel’s record and linked to their photo facsimiles on our website and on our Flickr. While the footnotes in this example are an extreme example, the work of cataloging the novel is not. Examining every page of these texts is a time consuming task and a single book can take hours; yet this kind of work is necessary for capturing the hidden aspects of the book and copy-specific information that reflects the uniqueness of each book as a material object. Just as 18th century novelistic practices informed my web design, social media has also illuminated aspects of early novel cataloging. Inspired by Twitter widgets in websites, I created a “500 feed” – a rotating feed of general cataloger notes fields from our records in the effort to expose otherwise-overlooked information and represent the quirky, idiosyncratic, and subjective aspects of books and the cataloging process. Reconceptualizing END as an “open database” has also been, for me, an attempt to re-capture and re-present the human labor behind it.
In homage to 18th century literary conventions, the latest installment of my internship could be titled “Volume II: A Venture into the Visual Realm.” I have begun designing two data-visualization applets to help users better read and study the techniques 18th century authors used to annotate, structure, and market their novels, as well as a video guide to the END project and website. Stay tuned for more updates!