Proceedings of the Annual Meeting of the Cognitive Science Society

We present an exploratory computational analysis of cognitive patterns in U.S. presidential campaign manifestos through metaphorical concept mappings. We examine documents from 1960 to 2024, using MetaPro to extract metaphors from both the Republican and Democratic parties' rhetoric. Our goal is to propose a novel way of political analysis by generating insights into the cognitive foundations of election rhetoric and revealing distinct patterns. Our research focused on three key ideas: (1) the evolution of the conceptual landscape of campaign rhetoric over six decades, (2) the different metaphorical strategies of incumbent and challenger campaigns, and (3) the divergence of Democratic and Republican parties in their metaphorical repertoires. The analysis reveals a moderate decline in cross-cycle mapping turnover over the observed timeline, as well as structural differences between incumbents and challengers, and the parties themselves, suggesting that political polarization is observable even at the level of metaphors.