Measuring Division: The Language of Dissent in the Supreme Court and American Polarization

Abstract: Political polarization in the United States has been widely examined through legislative behavior, electoral politics, and mass political discourse. Far less attention, however, has been paid to how polarization is discursively constructed within elite judicial institutions. This paper examines how ideological conflict is enacted and negotiated through the language of dissenting opinions in the U.S. Supreme Court. Rather than treating polarization solely as a function of voting behavior or ideological scores, the study adopts a corpus-assisted discourse analytic approach to examine how disagreement is patterned linguistically across time.

Dissenting opinions are treated as a distinctive institutional discourse genre. While they permit explicit disagreement, they remain constrained by norms of judicial legitimacy, professionalism, and institutional authority. This tension makes dissents a particularly revealing site for examining how ideological conflict is rhetorically managed within the Court. The analysis focuses on shifts in evaluative stance, rhetorical framing, and linguistic markers of disagreement across judicial eras, asking whether these patterns align with broader trends in American political polarization.

The study draws on a diachronic corpus of Supreme Court dissenting opinions spanning the nineteenth century to the present, constructed using digitized judicial texts from the Caselaw Access Project. Texts are extracted, cleaned, and annotated with metadata including year of decision, justice identity, Court era, and ideological orientation. Using corpus-driven keyword analysis, measures of evaluative language, lexical diversity, modality, and semantic similarity modeling, the study identifies patterns of rhetorical divergence. Quantitative findings are interpreted alongside institutional and historical context to understand how ideological conflict is expressed within judicial discourse.