“Taking up speech” in an Endangered Language: Bilingual Discourse in a Heritage Language Classroom Robert Moore University of Pennsylvania Faced with the “diversiication of diversity” (Vertovec, 2010) that seems to deine the contemporary world, some have called for a fundamental reorientation of sociolinguistics: from a focus on languages and speakers to a focus on resources and repertoires; from unitary, localized and countable ethnolinguistic communities to diasporized (or even virtual) ones; and from fully-luent ‘native speaker’ competence to “individuals’ very variable (and often rather fragmentary) grasp of a plurality of differentially shared styles, registers and genres” (Blommaert & Rampton 2011, 6). The ‘super-diversity’ that prompts such relections, I argue, can and should be discussed together with what seems to be its opposite: the seeming loss of diversity brought about by processes of language shift, obsolescence, and endangerment. Examination of classroom discourse on a US Indian Reservation suggests that in this community, at least, people have long since moved on from the idea that all the competences associated with “proiciency” in language need to coincide in a single person. These students are learning to speak (parler) rather than internalizing a complete grammar (langue); in this respect their project resembles that of (other) denizens of the “super-diverse” metropole. Introduction H ere I describe community efforts to document and teach Kiksht (Wasco- Wishram dialect of Upper Chinookan) in a heritage language classroom on the Warm Springs Indian Reservation (Oregon, USA). I argue that the discourse strategies adopted in the classroom create new strategies for the capture of speech (Certeau, 1997 [1968]) by younger generations interested in the revitalization of this severely endangered American Indian language. Attention to the participation frameworks and production formats (Goffman, 1981) of classroom discourse shows how participants pool their scarce resources and collaborate to produce new voicings of Kiksht in a new register, one that is emblematic of these adult students’ emergent status as new speakers in a context informed by globalization. Most important of all, these new voicings of Kiksht do not depend for their effectiveness on any assumption that all the forms of linguistic competence associated with luent native speakerhood need necessarily to coincide in the same person. Indeed, the register itself presupposes a distribution of communicative roles—speaker, (over-)hearer, translator/interpreter, repeater, and scribe, among others—that is instantly recognizable in the organization of other formal speech events in the local community. Working Papers in Educational Linguistics 27(2): 57-78, 2012 // www.gse.upenn.edu/wpel