Abstract
This study reports on the methodological challenges of determining and quantifying ethnic identity in questionnaire and interview data from second-generation members of the San Francisco Chinatown community, and linking these identity scores to the use of durational characteristics that we argue are part of the ethnolinguistic repertoires of our participants. We analyse durational features from free speech in interactions with an in-group and an out-group interlocutor for an exemplary sample of four speakers. By thus combining qualitative data on ethnic identity orientation with quantitative sociophonetic results, we show that rhythmic variability in free speech is both possible and traceable, and suggest that speech rhythm can be used as a flexible feature to index ethnic identity.