| Applying Conversation Analysis (CA) to research in Second Language Acquisition (SLA) has spurred substantial scholarly activity (Firth and Wagner 2007; Markee and Kasper, 2004). A central difficulty that remains is the convincing demonstration of language development (acquisition/learning) by means of the data and the analytic procedures that CA applies to the analysis of talk-in-interaction (Kasper, 2006). A study in pragmatic learning, this paper presents demonstrable instances of language learning over time. The data consist of task-related, dyadic L2 talk on the telephone produced by 40 advanced learners of German. Comparing telephone opening sequences of two data sets collected from the same learners two months apart, the findings suggests that longitudinal CA studies are fully capable of demonstrating language learning. However, without additional data that is not bound to the constraints of talk-in-interaction, CA data may preclude inferences about language development that is not visibly reflected in the talk. Accordingly, it is argued that CA provides a heuristically viable methodology to study language development, particularly in pragmatics; yet, if applied exclusively, significant constraints inhere in the approach. Research on language development that acknowledges both its social-interactive dimension as well as the cognitive processes within a singular learner’s brain requires cross-disciplinary studies that gainfully link CA data to data and insight from other research procedures. |
