Cartographic Visualization in Support of Dialectology

Abstract

Using data from the Syntactic Atlas of Swiss German Dialects (SADS), several methods for the cartographic visualization of linguistic data are proposed, demonstrated, and evaluated against the requirements of linguistic research. After reviewing the challenges of lin- guistic data visualization, point symbol maps are introduced as a baseline. We then present alter- native visualization methods that present linguistic data in new ways. The first uses Voronoi polygons about the data points to color in the dominant variant per location. The second uses kernel density estimation (KDE) to interpolate intensity values of all variants and thus infer and display the dominant variant per location (incl. at missing value locations). The KDE-based method helps to better see trends in the data and also automatically infer isoglosses. The third new technique uses 3-D visualization to support the exploration of spatial trends, as well as co- occurring variants. As a fourth alternative, measures and methods from geostatistics are used for the visualization of specific, global and local, variations and patterns.