Examining Political Rhetoric with Epistemic Stance Detection

TL;DR

This paper introduces a RoBERTa-based model for automatically detecting epistemic stance in political discourse, enabling large-scale analysis of belief attribution trends across U.S. political ideologies.

Contribution

It presents a simple yet effective stance prediction model and demonstrates its novel application to political science research on ideological belief patterns.

Findings

Model outperforms complex state-of-the-art approaches.

Reveals ideological differences in cited belief holders.

Enables large-scale automated analysis of political texts.

Abstract

Participants in political discourse employ rhetorical strategies -- such as hedging, attributions, or denials -- to display varying degrees of belief commitments to claims proposed by themselves or others. Traditionally, political scientists have studied these epistemic phenomena through labor-intensive manual content analysis. We propose to help automate such work through epistemic stance prediction, drawn from research in computational semantics, to distinguish at the clausal level what is asserted, denied, or only ambivalently suggested by the author or other mentioned entities (belief holders). We first develop a simple RoBERTa-based model for multi-source stance predictions that outperforms more complex state-of-the-art modeling. Then we demonstrate its novel application to political science by conducting a large-scale analysis of the Mass Market Manifestos corpus of U.S. political opinion books, where we characterize trends in cited belief holders -- respected allies and opposed bogeymen -- across U.S. political ideologies.