Séminaire de Recherche en Linguistique

Establishing patterns of linguistic variation traditionally requires extensive fieldwork, whether this is for a sociolinguistic study of a single community or a dialect atlas covering a whole region. Social media potentially provide large amounts of data from many more individuals than we could hope to interview using traditional means. In this talk, I will introduce the Tweetolectology project, which aims to establish the geographic and social distribution of linguistic variants using the vast amount of data available in Twitter messages. We will look at some examples of primarily morphosyntactic linguistic variation from Welsh, British and Irish English, and Haitian Creole, seeing how tweets demonstrate both well-established patterns of variation and some newly emerging ones, potentially demonstrating linguistic change in progress and diffusion of new variants.