Quantification and Statistical Analysis of Structural Similarities in Dialectological Area-Class Maps

Abstract

Some of the main aims of dialectology have always been the division of a given geographic space into areas where different dialects are spoken, the detection of dialect boundaries, and the investigation of the strength of these boundaries. For this purpose, a sub-discipline of dialectology, „dialectometry‟, has introduced the method of counting or measuring the differences between lects spoken at different locations. (For a concise overview of dialectometric methods, see, for example, HEERINGA 2004, 9–24.) The higher the number or degree of differences between two locations are, the higher is the chance that they are placed in two disjoint dialect areas. This implies, however, that each pair of locations sometimes exhibits agreement, sometimes disagreement, even within one dialect area. Otherwise, there would only be completely disjoint, mutually unintelligible lects. Instead, we find that while some of the variants fit neatly into the dialect areas, others show somewhat divergent or even completely different geographical distributions. Consequently, the distributions of single variants must be considered more than mere deviations from one underlying pattern as represented by the dialect areas; in fact, they differ so greatly from one another that they need to be studied, too.