| dc.description.abstract | The uneven global distribution of human languages remains a significant question in linguistics. Prior research suggests that more politically complex societies tend to reduce language diversity by spreading their languages over larger areas through cultural group selection. To explore this, this study refines an existing agent-based model to simulate the emergence and spread of languages, incorporating political complexity and cultural group selection mechanisms. Languages are distinguished in the model using divisive clustering with an optimal Levenshtein distance normalised (LDN) threshold determined by silhouette scores. The model simulates societal interactions over 7,000 years, comparing outcomes with real-world data from West Africa. The results indicated an optimal LDN threshold of 0.624 for distinguishing languages, though a threshold of 0.500 was used to ensure sufficient languages emerged. Further, the results suggested that political complexity might reduce language diversity, though this could not be validated when compared to the real-world data. Despite the studies limitations, it provides insights into the use of agent-based models for simulation language emergence and evolution, along with providing insights into how political complexity and cultural group selection mechanisms shape language diversity. | |
