In 1946, Professor David Boder from the Illinois Institute of Technology traveled to Europe to interview Holocaust survivors in displaced persons’ camps in France, Germany, Switzerland, and Italy. Using a wire recorder, he collected over 100 interviews in 9 different languages, asking survivors about events which were still relatively unknown in the United States. The recordings were later transcribed and translated, providing a unique collection of early testimonies. Starting in the 90’s, thousands of recorded
testimonies were also collected by the Shoah Foundation Visual History Archive at the University of Southern California, which today hosts nearly fifty-five thousand video interviews in dozens of languages. The content of these interviews was indexed at 1-minute intervals, providing an insight into the most recurring topics of memory and narration.
These testimonies are the narration of lived traumatic experiences collected at different points of the life of survivors: Boder interviews are very early, while Shoah Archive interviews are decades apart. The languages, narratives, and memories displayed in these interviews change over time. This is why our project focuses on analyzing language use and narrative structures in these testimonies, to identify their unique features as well as
their commonalities. In particular, our project seeks to understand the dynamics of language use and code switching in the interviews, as well as the presence and change of recurring topics represented by index terms.
By applying several Digital Humanities approaches, we pursue two main goals:
● Providing interactive visualizations that give the observer a clearer perception of the overall corpus narrative;
● Enabling a quantitative comparison of genocide narrative features across different genocide corpora with testimonies from survivors of different origins, ages, and cultural backgrounds.
Our visualization and analysis processes use Python programming for data cleaning and analysis, and Tableau for data visualization. The digital humanities research approach we used in this work is applicable to other similar corpora, and it seems productive for at least two reasons: it allows reproducible analyses grounded in data that can complement the
qualitative analysis of a delicate subject matter such as genocide testimony, and it allows us to analyze some of the subtleties and variations of these testimonies in a way that humanizes survivors and does justice to their experience rather than reducing them to numbers.

Authors: Anna Bonazzi , Lizhou Fan, Todd Presner (Supervisor)