This paper highlights prosody as a fundamental feature of bilingual conversation. My data show that syntactic structure does not impose constraints on codeswitching, as one prevalent line of inquiry regarding codeswitching claims, but rather certain discourse structures make codeswitching at any given point more or less cognitively and interactionally profitable according to conversationalists' ability to produce and comprehend information.
My corpus consists of one hour of conversational data from four competent bilinguals of Mexican heritage living in Southern California yielding a total of 782 analyzable units. Using the transcription methods developed by Du Bois, Schuetze-Coburn, Paolino, and Cumming, (1993) wherein each line of transcription consists of one Intonation Unit (Chafe, 1979, 1987, 1993, 1994), the prevalent pattern which emerged was one in which speakers overwhelmingly switched at Intonation Unit boundaries. Using what I have termed the Switch-Boundary Intonation Unit (SBIU) as my unit of analysis and adapting the notion of completion points (Ford and Thompson, 1996; Sacks, Schegloff, & Jefferson, 1974), I examine intonation contour type, syntactic completion and constituency, and pragmatic completion in order to best characterize the codeswitching frame.