Abstract

In this presentation I will discuss a series of dialect studies where I have applied new computational approaches to the analysis of regional linguistic variation. These studies are computational in two senses. First, as opposed to traditional dialect studies, which are based on data collected through linguistic interviews, in these studies I analyse corpora of natural language and hit counts harvested from Google. Second, as opposed to traditional dialect studies where regional linguistic variation is analysed by hand, in these studies I analyse regional linguistic variation using statistical techniques borrowed from geographic information systems, data mining, and quantitative corpus linguistics. I will argue that by adopting computational methods it is possible for regional linguistic data to be collected and analysed with greater efficiency and precision than is possible using a traditional approach.