Working Group

Modeling the diversification of human languages

PI(s): Michael Gavin (Victoria University of Wellington)
Start Date: 10-Oct-2010
End Date: 10-Oct-2012
Keywords: cultural evolution, biogeography, phylogenetics, mathematical modeling

The nearly 7000 languages humans collectively speak are unevenly distributed across the globe with far more languages at lower latitudes. Although geographic patterns of language diversity are well documented, little is known about how these patterns formed. Several different academic disciplines (e.g., linguistics, anthropology, geography, biogeography) have produced theories describing the mechanisms determining geographic patterns in language diversity. Some focus on historical, economic, political, and social drivers, whereas others suggest that ecological and biogeographical variables, alone or in combination with social factors, are key drivers of language diversification. However, few studies have empirically examined predictors of spatial patterns in language diversity, and results to date are equivocal with each study pointing to the importance of a different set of predictors. Although these previous studies provide an important starting point, they have three major limitations: they do not incorporate key ideas from linguistics and anthropology, the correlation approaches used do not explicitly include diversification mechanisms, and they do not account for temporal shifts in key predictors over the millennia during which language diversity developed. We will overcome these limitations and advance the understanding of language diversification by: (1) synthesizing the literature from linguistics, anthropology, and biogeography to produce a review of the major mechanisms hypothesized for diversification; and (2) testing a subset of these mechanisms drawing on techniques (i.e. simulation modeling and Bayesian phylogenetic methods) that have begun to be used in evolutionary biology and biogeography, but are novel to the study of spatial patterns in human language diversity.