Drivers of geographical patterns of North American language diversity

Abstract

Although many hypotheses have been proposed to explain why humans speak so many languages and why languages are unevenly distributed across the globe, the factors that shape geographical patterns of cultural and linguistic diversity remain poorly understood. Prior research has tended to focus on identifying universal predictors of language diversity, without accounting for how local factors and multiple predictors interact. Here, we use a unique combination of path analysis, mechanistic simulation modelling, and geographically weighted regression to investigate the broadly described, but poorly understood, spatial pattern of language diversity in North America. We show that the ecological drivers of language diversity are not universal or entirely direct. The strongest associations imply a role for previously developed hypothesized drivers such as population density, resource diversity, and carrying capacity with group size limits. The predictive power of this web of factors varies over space from regions where our model predicts approximately 86% of the variation in diversity, to areas where less than 40% is explained.

Keywords: geographically weighted regression; language diversity; path analysis.

Figure 1.

Observed language diversity. Language ranges are shown in the gridded map. Blank spaces on the map indicate regions in which no information about language distribution is available and thus were not compiled in the grid map. (Online version in colour.)

Figure 2.

Global path model quantifying direct and indirect effects of environmental and sociocultural factors on North American language richness. The numbers marking each arrow represent the standardized β coefficients (i.e. path coefficients) for language diversity. Model fits (R²) are shown for variables directly affected by other factors. (Online version in colour.)

Figure 3.

GWPath applied to North American linguistic diversity. (a) In the GWPath model, the standardized β coefficients of variables, as well as the R² for the direct relationships are represented by the average value over the continent, followed by its standard deviation. (b) Model fit varies over the geographical domain of North America. (c) Variables with the highest total coefficient (sum of direct and indirect effects) also vary across the continent. (Online version in colour.)

Figure 4.

Direct effect of predictors mapped over the North American domain. The standardized β coefficient is mapped for (a) river density, (b) ecoregion richness, (c) topographic complexity, (d) climate change velocity, (e) precipitation constancy, (f) population density, and (g) carrying capacity with group size limits. (Online version in colour.)