Towards faithfully visualizing global linguistic diversity

Abstract

The most popular strategy for visualizing worldwide linguistic diversity is to utilize point symbology by plotting linguistic features as colored dots or shapes on a Mercator map projection. This approach creates illusions due to the choice of cartographic projection and also from statistical biases inherent in samples of language data and their encoding in typological databases. Here we describe these challenges and offer an approach towards faithfully visualizing linguistic diversity. Instead of Mercator, we propose an Eckert IV projection to serve as a map base layer. Instead of languages-as-points, we use Voronoi/Thiessen tessellations to model linguistic areas, including polygons for languages for which there is missing data in the sample under investigation. Lastly we discuss future work in the intersection of cartography and comparative linguistics, which must be addressed to further advance visualizations of worldwide linguistic diversity.

Resources

Details