APRIL 13, 2017 –
4:00PM TO 5:30PM
Lecture by Christopher M. Church, PhD, assistant professor, Department of History and co-director, Nevada Center for Data and Design in the Digital Humanities (NDAD), University of Nevada, Reno
Democracies have flexed their imperial muscle the world over since the onset of the nineteenth-century, when the French, and Europe more broadly, focused on empire-building as a way to achieve national glory and international security, shaping international relations into the twentieth and twenty-first centuries. Liberal ideology and the communication revolution have simultaneously enlarged the empires of Western democracies while serving as their most vocal critique. Consequently, it is incumbent upon scholars to investigate the historical relationship between empire and modern democracies, particularly between public policy, the press, and the populace in order to fully understand contemporary developments in these relationships.
To this end, this presentation will explain how creating interactive cartographic visualizations by text mining historical periodicals can enable scholars to analyze how cultural imagination informed political conquest. By performing “distant reading” on the popular French weekly, the Journal des Voyages, we can unearth the imperial narrative targeted not only at the reading public, but most interestingly the one endorsed for use in French schools. While visual text analysis holds great promise for gaining insights into how French newsprint portrayed colonized peoples and locales throughout the new imperial period, effective implementation requires interdisciplinary collaboration focused equally on data visualization, text analysis, machine learning, and humanities research questions. Therefore, this presentation addresses both the opportunities and challenges involved in performing a “distant reading” of historical periodicals in French, including maximizing insights from cleaned OCR data, aptly performing natural language processing on non-modern, non-English languages, and creating easy-to-use data visualizations that grapple with their source material’s inherent biases.