Using correspondence analysis to model immigrant multilingualism over time

The notion of immigrant and linguistic superdiversity privileges the multidimensional nature of multilingualism and requires analytic models that attend to the bidirectional interaction of many social and psychological variables. In this chapter I cast multilingualism as an interactive field of language proficiency and language practices, and I propose the use of multiple correspondence analysis (Greenacre 2007) and geometric data analysis (LeRoux & Rouanet 2004) as means for visualizing and analyzing this field. Further, I examine the literature on longitudinal multiple correspondence analysis to suggest ways of modeling the development of language proficiency and the expansion of domains of language use over time. Data are taken from a study of Puerto Rican bilinguals in the city of Chicago in the United States (Schrauf 2009).