Abstract The paper illustrates the results of a correlation study focusing on linguistic variation in an Italian region, Tuscany. By exploiting a multilevel representation scheme of dialectal data, the study analyses attested patterns of phonetic and morpho-lexical variation with the aim of testing the degree of correlation between a) phonetic and morpho-lexical variation, and b) linguistic variation and geographic distance. The correlation analysis was performed by combining two complementary approaches proposed in dialectometric literature, namely by computing both global and place-specific correlation measures and by inspecting their spatial distribution. Achieved results demonstrate that phonetic and morpho-lexical variations in Tuscany seem to follow a different pattern than encountered in previous studies.

INTRODUCTION

It is a well-known fact that different types of features contribute to the linguistic distance between any two locations, which can differ for instance with respect to the word used to denote the same object or the phonetic realisation of a particular word. Yet, the correlation between different feature types in defining patterns of dialectal variation represents an area of research still unexplored. In traditional dialectology, there is no obvious way to approach this matter beyond fairly superficial and impressionistic observations. The situation changes if the same research question is addressed in the framework of dialectometric studies, where it is possible to measure dialectal distances with respect to distinct linguistic levels and to compute whether and to what extent observed distances correlate.