There is a long history of data-driven methods applied to confront the painful legacy of lynchings in the United States. It begins with Ida B. Wells who first utilized a "computation of lynching statistics" to draw attention to the issue in the 1890s (Wells 61). Wells drew data from annual lynching records published by the Chicago Tribune. In the decades that followed, Monroe Work and the Tuskegee Institute also compiled data drawn from newspapers. So, too, did the NAACP. These data became a critical tool for an international anti-lynching movement that grew throughout the period. The data were used to dispel myths that sought to excuse lynchings. They were used to prove how lynchings had become a national issue. Through intensive lobbying campaigns that wielded national data, the anti-lynching movement was able to combat racial violence across the country. In this presentation, I will be tracing a history of the anti-lynching movement through an analysis of its data-driven activism. This history contextualizes–and is contextualized by–a pipeline I have built for identifying reports of lynchings in historical newspapers. This pipeline includes text-mining the Chronicling America archive, fine-tuning BERT to classify reports of racial violence, and enriching the extracted data for downstream tasks. In its current iteration, this project has identified over 50,000 lynching reports published between 1865 and 1921. Since newspapers were the primary source for lynching data, these reports uncover the information ecosystem that allowed the anti-lynching movement to compile and utilize its data. Because white mobs most often targeted Black communities, I have prioritized identifying lynching cases where victims were Black. Lynching was a threat to all marginalized groups, however, including immigrant groups, Native Americans, and people with disabilities. I will therefore conclude by outlining future plans to adapt this pipeline to identify reports based on other demographics, locations, and periods.
Matthew Kollmer (Wed,) studied this question.