Phonetic distance and dialect clustering on the Qinghai-Tibet plateau

Abstract

The goal of this paper is to use string edit distance to describe the synchronic relationship between the Tibetan speech varieties located on the Northeastern edge of the Tibetan Plateau. String edit distance provides a statistical way to compare a large number of linguistic features, in essence producing a statistical bundle of isoglosses. In this way, it can be used as a tool in dialect mapping and synchronic clustering. In this paper, the aggregate distance matrix produced by string edit distance reveals that the great degree of phonetic continuity on the grasslands of the northeastern edge of the plateau is matched by an equal degree of phonetic discontinuity in the mountains forming the eastern border of the plateau. While the dialects located on the grasslands can be grouped together into one cluster, the dialects in the mountains can be grouped together into six clusters.