“In Search of the Drowned in the Words of the Saved: Testimonial Fragments of the Holocaust” by Gabor Toth, PhD (University of Oxford) 2018-2019 Center Postdoctoral Research Fellow, April 2, 2019 (lecture summary)

Gabor Toth, 2018-2019 Center Postdoctoral Research Fellow, gave a public lecture at the USC Shoah Foundation Center for Advanced Genocide Research focusing on his project to find, represent, and reflect on victims’ experiences during the Holocaust. 

He opened his talk by playing a clip of Cecilie Klein-Pollack’s USHMM (United States Holocaust Memorial Museum) testimony where she shares a story about witnessing the brutal treatment and death of a young ballet dancer during the Holocaust. Toth explained that while we understand a great deal about the history of the Holocaust (what, when, and how perpetrators did what they did), we cannot understand how the little ballerina experienced her persecution. There is a gap, Toth argued, between our historical understanding of what happened during the Holocaust and the experiences and perspectives of millions of voiceless victims.

To attempt to define, recover and represent victims’ experiences, Toth has been working on a digital monograph entitled “Let Them Speak”, which will include a series of essays about victims’ perspectives, as well as an anthology of 2700 transcripts from three different collections of survivor testimonies: the Fortunoff Video Archive for Holocaust Testimonies (specifically a small set of those gathered by the Holocaust Survivors Film Project), the United States Holocaust Memorial Museum, and the USC Shoah Foundation. This edition is centered on victims’ perspectives. The digital tool was developed in conjunction with the Yale University Digital Humanities Lab. 

To demonstrate the gap between our historical understanding of the Holocaust and victims’ experiences, Toth discussed an example of a major turning point that survivors describe -- seeing their parents crying for the first time. This experience is related again and again in testimonies. While victims seeing their parents crying does not figure prominently in our historical understanding of the Holocaust, Toth argued, it is a profoundly relevant emotional and mental experience for victims. What other significant emotional, mental, and physical experiences are relevant for victims that we are not aware of? Or even if we are aware of them, have we really grasped their significance for victims? 

Toth defines victims’ experiences as intangible but existing emotional and mental dimensions of persecution. There have been oral history projects to document, preserve, and share victims’ perspectives. Toth asserted that it is from these projects that we can identify experiences that could have been those of any victims – collective experiences shared by survivors and murdered victims. From these, we can try to grasp the experiences of victims. The goal of the “Let Them Speak” digital monograph and transcript tool is to give a tangible and real shape to victims’ perspectives and make them searchable and accessible to researchers, educators, and students.

When setting out to find recurring experiences in the 2700 testimonies that are part of “Let Them Speak”, Toth faced potentially thousands of similar episodes. So he focused on emotional, mental, and physical experiences in eight experience domains. (Future scholarship could identify even more testimonial fragments.) Based on the recurrence of experiences in survivor testimonies, Toth focused on what victims were likely to have experienced. He detailed the computational challenges he faced and described the methods he used to find recurring experiences in testimony.

After addressing the problem of how to computationally discover recurring experiences, Toth faced the challenge of representation – how to represent recurring experiences in testimony as pieces of collective experience without simplifying and homogenizing them. Toth wanted the victims to speak for themselves. He chose to approach this challenge by representing these experiences as incomplete and decontextualized “testimonial fragments” that unlock the subjective realities of survivors and victims. In the “Let Them Speak” tool, one can click on a fragment to go to the full transcript of the interview and see the fragment in its context. Toth made the decision to always represent fragments in groups.  One testimonial fragment alone cannot represent the whole experience, he argued. The goal is for fragments to complement each other and convey the heterogeneous dimensions of collective experience. The digital sphere, Toth argued, allows oscillation between the collective and the individual.

Throughout his lecture, Toth illustrated his points with clips from testimony and demonstration of the "Let Them Speak" tool. Returning to the example of survivors talking about seeing their parents cry, he demonstrated a search of “mother crying” and got 34 results from that search. Then he illustrated the steps he would follow to refine and expand the search and the steps to find and match patterns in the text of the 2,700 transcripts. With each step he demonstrated, the search results grew. He ended up with about 1,000 search results of survivors talking about their parents crying. In a third of the interviews in “Let Them Speak”, survivors describe this pivotal experience. 
 
Toth concluded the lecture by arguing that there is a moral imperative to humanize the voiceless victims by revealing and presenting the emotional, mental, and physical experiences that they were likely to have gone through.

In the lengthy Q&A period following the lecture, Toth discussed how he chose the eight experience domains he focuses on, which include other recurring experiences that don’t feature in the history of the Holocaust (such as running or nakedness); his editing process; how he defines and labels turning points; how the fragments can help stand for the experiences of all victims, not just those who were murdered but those who never gave interviews and never talked about their experiences; the effects of institutional practices and what his tool can help discover about the form of interviews; and the ethics of text and data mining. 

Summary by Martha Stroud