Mapping Moral Cultures as Meaning Networks: A Computational Framework for Comparative Analysis

• In Event: Sociology of Culture Section Roundtables
  In Refereed Roundtable (90 minute): 9. Computational Sociology & Semantic Mapping

Abstract

Moral culture, the shared mental representations of what is right and wrong or good and bad within groups, is a central theme in cultural sociology. However, research relying on aggregated survey scores has not fully captured how values are relationally organized and contextually interpreted. This study proposes a computational framework that reconceptualizes moral culture as a value-centered meaning network by integrating open-ended text data with Schwartz’s value model. Using survey data from the 2016 U.S. presidential election (N = 2,038; 1,016 liberals and 1,022 conservatives), I implement a workflow, applying three different techniques, Structural Topic Modeling (Roberts et al. 2019), network estimation and comparison via the Network Comparison Test (van Borkulo et al. 2023), and Moderated Network Analysis (MNA; Haslbeck et al. 2021), to extract value-related topics and compare their relational structures. The results indicate that although the two groups exhibit comparable overall network connectivity, they differ substantially in how values structure specific topics. In the conservative network, Self-Transcendence is positively associated with election-related discourse and significantly moderates ideologically sensitive topic pairs, including Tradition–Gender and Political Change–Family/Religion, suggesting that other-oriented values reinforce in-group moral cohesion. In the liberal network, Self-Transcendence is negatively associated with education and immigration discourses and does not exert a significant moderating effect, suggesting that these domains are organized by alternative value logics. By contrast, Conservation operates as a cross-ideological moderator, conditioning the strength of several topic–to–topic connections (e.g., Political Change–Family/Religion and Tradition–Gender). By enabling systematic comparison of both the content and structure of shared meaning systems, this framework advances moral culture research beyond the distributional analysis of individual value scores and toward a relational understanding of moral meanings.

Author

• Donghyun "Henry" Kim, University of Iowa