Abstract: As machine learning becomes increasingly tasked with consequential real-world decisions, ever more concerns are raised about the kinds of social biases that it reinforces and perpetuates. Machine learning algorithms are not neutral observers of the world, but see what they are trained to see, which means seeing with the human biases that data encodes. While industry and academic experts have responded by considering how to make these algorithms more fair, cultural historians have begun using them to read for patterns of gender and racial bias in the archive. This talk provides an example of what such readings might yield by using word embeddings to explore the semantics of racial discourse in a large corpus of Japanese periodicals and fiction written during the rise and fall of Japanese empire (1890-1960). Explorations of bias at larger scales, I argue, can offer insights into the character of racial discourse as interpretable pattern, whether by algorithm or human.
Bio: Hoyt Long is Associate Professor of Japanese Literature at the University of Chicago. He publishes widely in the fields of Japanese literary studies, media history, and cultural analytics. He co-founded the Chicago Text Lab with Richard Jean So and currently co-directs the Textual Optics Lab, which focuses on the creation of large-scale, multi-lingual text collections and the development of tools to explore them. His recent publications include “Race, Writing, and Computation: Racial Difference and the US Novel, 1880-2000” (Journal of Cultural Analytics, 2019) and “Self-Repetition and East Asian Literary Modernity, 1900-1930” (Journal of Cultural Analytics, 2018).