Generative Agent-Based Model of Language Adoption

This BSc thesis project built a generative agent-based system using Large Language Models (GPT-4o-mini) to simulate language adoption dynamics in populations. The system modeled how languages evolve and spread through social interactions, implementing evolutionary algorithms for linguistic adaptation and cultural transmission. Over 2,000 simulations were conducted and analyzed using Bayesian statistical models in R, revealing complex patterns of language evolution from simple interaction rules. The research explored how individual agent behaviors and communication strategies lead to emergent linguistic phenomena, providing insights into the mechanisms of language change and cultural diffusion in human societies.