Theories of culture have long relied on spatial metaphors to describe how meaning systems are internally organized. Constrained by limitations of data and visualization, prior spatial models of meaning have commonly taken the form of two-dimensional “maps” that can be easily rendered on paper. However, recent advances in computational linguistics show that the vast array of semantic associations that characterize a cultural system can only be effectively represented by expansive models with hundreds of dimensions. In this talk, I outline how we may harness such high-dimensional models to study cultural systems structurally and holistically. Focusing on collective understandings of class and politics, I put forth methods to identify periods when cultural systems undergo structural shifts.
Austin Kozlowski is a PhD candidate in the Department of Sociology at the University of Chicago. His research lies at the intersection of culture and politics and explores how political ideas link together to form belief systems and ideologies. His work draws on a range of methods including computational text analysis, survey research, and qualitative in-depth interview analysis. His prior work has examined how word embedding models can be used to measure shared meaning, how changing political coalitions drove the polarization over science, and how ideology is linked to overconfidence among elite economists. In his dissertation, he argues that political divisions have recently aligned with longstanding cultural divides in the United States, amplifying contemporary polarization.