Unstable Concepts, Measurable Processes: A Macro-Level Approach to Racialization

• In Event: Racial and Ethnic Minorities Section Roundtables Session
  In Refereed Roundtable (90 minute): Table Eight: Racialized Constructions of Social Identities and Categories

Abstract

The paper addresses a core problem in the sociology of race: once U.S.-centered concepts of “race” and “ethnicity” are exported to other national contexts, they become unstable and undermine analytic clarity. In settings where race, ethnicity, and nation are fused or organized differently than in the United States, the same empirical case can be plausibly classified as “race,” “ethnicity,” or something else, depending on which conceptual tradition is imported. The paper argues that this instability is best confronted not by choosing between competing definitions, but by shifting analytic attention to racialization as a process and treating that process as an object of macro-level, longitudinal measurement.
Building on Wimmer’s boundary-centered approach and Wacquant’s account of racialization, the paper conceptualizes racialization as multi-dimensional (naturalization, eternalization, hierarchization, homogenization), multi-level (micro, meso, macro), and non-linear, entailing likely deracialization and reracialization processes, as fields are contested. The key claim is that parts of this process—especially hierarchization and homogenization—are measurable in cultural and institutional discourse (news media, policy debates, and other elite-oriented arenas). Using diachronic word embeddings and related natural language processing tools on large text corpora, the approach infers changes in (1) the concentration, diversity, and temporal stability of descriptors attached to groups (homogenization) and (2) the evaluative valence and relative ranking implied by associated terms (hierarchization). This capacity to track when stereotypes loosen, harden, or reconfigure brings deracialization and reracialization into view.
This framework makes racialization, deracialization, and reracialization empirically tractable at the macro level, without presupposing a fixed answer to what “race” or “ethnicity” a group is. Conceptually, it offers an alternative to exporting U.S. race concepts; methodologically, it demonstrates how macro-level racialization can be observed and quantified over time; substantively, it enables comparative analyses of racialization in contexts where census categories are absent or organized around different classificatory logics.

Author

• Elisa Kim, Stanford University