Session Abstract

The Digital Humanities - and the attendant subfield of Digital History - have flourished commensurately with the increasing availability and ubiquity of digital data. Both large and small scale digitization projects--often combined with increasingly sophisticated OCR and crowdsourcing components--have begun to utterly transform the historical record, imbuing it with novel interoperable and extensible possibilities. Not only have well known traditional historical sources been converted into a digital form, but previously isolated and inaccessible repositories have gained a new visibility in the scholarly discourse as well. Combining computer science, social science, literary theory, and visual design, historians across the globe have already begun to explore and interrogate these new digital sources with myriad new techniques. Emerging research methods like ‘distant reading’ or ‘macroanalysis’ (the computational reading of thousands or more sources), facilitated by topic modeling, corpus linguistics, and network analysis are now constantly pushing the boundaries and scale of historical research, as well as taking it into new disciplinary domains, enabling historians to ask new questions and fundamentally new kinds of questions.

Novel research tools and techniques have brought not only intriguing possibilities for historical inquiry, but also daunting methodological challenges. Yet historiographical and methodological discussions have struggled to keep pace with the new ways in which historians have begun to interrogate the digital historical record. This gap has been widened by virtue of the interdisciplinary nature of working at new scales and scopes than historians are accustomed to. To what extent must historians explicitly confront their methodological relationship to other disciplines? How have new statistical and graphical techniques reframed traditional questions about causality and correlation? To what extent does a digital historical record compress the distances between social, cultural, and intellectual history? What is the relationship between analytical error and interpretive error? Can design and improved visual literacy effectively bridge the gap between statistical inference and historical interpretation?

This panel, “Text Analysis, Visualization, and Historical Interpretation,” brings together a set of digital historians (four speakers and a commentator) who will draw from their diverse disciplinary and methodological backgrounds to address some of these questions. Micki Kaufman will explain how she has fruitfully used text analysis visualizations to facilitate historical inquiry ranging from emotional history to more structural and comparative approaches. Ian Milligan, using examples from his ongoing work with large Web archives, will explain the potentials and pitfalls of this unique source base while presenting a workflow that enables researchers to explore such collections through both distant and close reading. Michelle Moravec will explore how scholars can interrogate the existing historiography of social history through visualizations of large data sets, in her case a large corpus of feminist periodicals. Fred Gibbs will address the methodological centrality of the oft-neglected role of design and aesthetics in creating and communicating historical arguments based on computational processes. Finally, the session’s chair and commentator, Robert Nelson, will tie together the speakers’ contributions and offer general reflections on the opportunities and challenges posed by history’s digital turn.