For second-generation members of a diaspora community, ethnic and cultural affiliation are less straightforward than for the first generation. We compare information on identity construction in London's Indian Diaspora with the participants' linguistic integration into the host community. Our study is novel and exploratory in that it combines quantitative, variationist methodology with a qualitative approach. We employ two standard sociolinguistic instruments to model subjects' ethnic identity: a questionnaire and sociolinguistic interviews with a focus on discursive identity construction. In a second step we investigate possible connections between morphosyntactic variation and ethnic identity in language use data from three different communicative contexts. The results show that, while interview data on ethnic identity are amenable to quantification, clear correlations between the resulting identity scores and vernacular morphosyntactic features are difficult to find. In particular, patterns of style-shifting between the different communicative contexts are not as expected.gories are decided on or the social theory underlying these categories.' It is the aim of this article to take a step in the direction of filling this research gap. In order to do so, we employ two standard sociolinguistic instruments, namely: (a) a questionnaire eliciting information on the subjects' background (including language proficiency and occupation) and their degree of integration in both ethnic and host/mainstream 1 communities; and (b) sociolinguistic interviews with a focus on discursive identity construction (Wodak, de Cillia, Reisigl, & Liebhart, 1999) and 166