Annual Meeting Preview: Digital Labor History and Historical Sources as Data

This session takes place on Saturday, April 6, at theOAH Annual Meeting in Philadelphia and is endorsed by the Labor and Working-Class History Association (LAWCHA).
#AM2955

Chair and Panelist: Vilja Hulden, University of Colorado Boulder
Panelists:
• Tobias Higbie, University of California, Los Angeles
• James Gregory, University of Washington

Digital Labor History and Historical Sources as Data

In this session we ask historians to envision how their sources can become “data” without excising messiness that matters. The session involves members of the digital history committee of the Labor and Working Class History Association (LAWCHA), who are currently drafting a multi-year initiative on sources and skills we are calling the Digital Labor History Incubator. We welcome participants from all subdisciplines of history into a conversation—what sources in your specialty could become “data?”

The panelists in this session have all done work—mapping, social networks, text mining—that can broadly be construed as falling under the rubric of “digital humanities.” We proposed the session (and the Incubator) because it seems to us that humanities scholars sometimes feel that digital humanities emphasizes the computational tool over the humanities content and sees humanities scholars as in need of (re)training. By contrast, we want to invite our colleagues to draw on their expertise in sources to consider the question of humanities “data.” What types of new analyses might those sources allow if we were to reimagine them as repositories of data? How do we think about our sources as data?

What do we mean by “sources as data?” Basically, any approach that grapples with the source through some method other than traditional close reading or examination. For example, Toby Higbie’s project takes a single source—a biographical compendium—and transforms it into tabular data that shows, for example, the birthplaces and organizational affiliations of a large number of early-twentieth-century labor activists. One can then create networks of people who belong to the same organizations or social clubs, map the birthplaces of activists, and so on. Suddenly, a book that does not make for particularly exciting reading becomes a rich source of new insights about patterns and connections in the American labor movement.

We believe that historical data cannot be left for data scientists to mine; we as historians need to develop approaches to sources as data that bring the historian’s awareness of complexity, nuance, and uncertainty to bear on how those sources can and cannot be analyzed at scale. Thinking of sources and archival collections in such terms is becoming increasingly common, primarily among library and archives professionals. Our hope is that our session will contribute to a greater interest among historians in engaging in that conversation. On these themes, we recommend that panel attendees read Miriam Posner’s short blog post on “Humanities Data: A Necessary Contradiction.”

We see the sources-as-data approach as 1) promoting greater awareness of hidden possibilities in the sources we use, 2) encouraging public-facing history and undergraduate involvement in the production of historical knowledge, and 3) forging an avenue to greater collaboration among historians as well as between historians and scholars in other fields (whether archivists, computer scientists, statisticians, or other humanities and social science scholars). All of these, we contend, offer potential both for achieving new research insights and for engaging scholars from other disciplines as well as students and the public at large in historical analysis.