Abstract
The aim of this investigation is to explore the suitability of speech posted on Twitter to examine regional language. Accordingly, existing differences among data extracted from Twitter was examined and compared to data extracted from a dialect survey administered to old-aged people and university students in Japan. The result of the comparative analyses revealed that (a) the dialectal speech used by old-aged people hardly reflected the speech posted on Twitter, and (b) the data accumulated from the survey administered to college students mostly matched Twitter speech. Hence, we can utilize Twitter conversations as data in undertaking dialect studies. In particular, these Twitter conversations can be utilized as a data resource in the study of regional variations in word forms. Though Twitter speech cannot be utilized in elucidating regional variations pervading traditional dialects, Twitter speech may serve as a means of finding distributions of new dialects and forms of expressions used mainly by youth groups as well as regional variations pervading expressions that are hardly noticed as dialects, for instance, unnoticeable dialects.
