Go to ACL ARR 2026 January homepageQuantifying Cultural Evolution: A Computational Analysis of Intellectual History in Chinese and Latin Traditions
Keywords: Cultural Analytics, Intellectual History, Cultural Evolution, Cross-Cultural Comparison, Knowledge Graphs, Large Language Models, Low-Resource Languages
Abstract: Understanding how intellectual traditions evolve across cultures remains a fundamental challenge in cultural analytics. We present a computational framework for quantifying cultural evolution by extracting idea graphs from classical texts using LLMs, detecting thought communities via graph algorithms, and discovering cross-temporal idea lineages through semantic similarity. Applying this framework to 243 Chinese and 636 Latin texts spanning over 2,500 years, we construct graphs containing approximately 360,000 entities and 390,000 relations. Our analysis reveals fundamentally different evolutionary dynamics: Chinese tradition exhibits asynchronous Politics-Religion dominance with cumulative inheritance, while Latin tradition shows synchronized dynamics with substitutional inheritance. We identify two critical transition periods where diversity trajectories diverge in opposite directions, and through structural analysis of thought communities and lineages, reveal why these patterns emerge and how they shaped divergent modern trajectories. Our quantitative findings align with historiographical narratives about Chinese continuity versus European transformation, providing measurable evidence for cultural analytics and cross-civilizational comparison.
Paper Type: Long
Research Area: Computational Social Science, Cultural Analytics, and NLP for Social Good
Research Area Keywords: cultural analytics, computational social science, network analysis, knowledge graphs, digital humanities
Contribution Types: Model analysis & interpretability, Publicly available software and/or pre-trained models, Data resources, Data analysis
Languages Studied: Chinese, Latin, English
Submission Number: 10688
Loading