Do different linguistic features pattern together?

What is it about?

Since the inception of the variationist enterprise in the 1960s, studies of language variation and change have generally focused on one variable at a time. This focus is understandable, given previous computational limitations on quantitative analysis, as well as the need to make analysis more tractable by isolating a single variable rather than trying to analyse all features of speech at the same time. However, since speakers (and listeners) are presumably capable of attending to variation at multiple levels of the linguistic system simultaneously, there is a tension between the goal of variationist analysis – to understand linguistic behaviour – and the practical realities of conducting research on language variation and change. This paper looks at cases studies from Toronto English and Bequia English, using Principal Component Analysis to explore the extent to which different variables pattern together in subgroups of speakers. The paper looks at these questions for phonology and syntax.

Featured Image