Coping with problems of understanding : repair sequences in conversations between native and non-native speakers

Abstract

The present study investigates how the speakers cope with problems of understanding in conversations between native and non-native speakers. It focusses on repair sequences, since the problems surface mostly in them. The data set consists of five conversations between native and non-native speakers of Swedish. The repair sequences in them were investigated through an analysis based on ethnomethodological conversation analysis (Garfinkel 1967; Heritage 1984b) and on the analyses of verbal interaction developed by Goffman (1967 & 1981) and Gumperz (1982ab). The research questions investigated firstly, how speakers from different linguistic and cultural backgrounds cope with problems and contribute to creating a shared world, and, secondly, how conversation analysis that has been developed for analysing conversations between equals can be used when the speakers lack adequate linguistic means to express themselves. The repair sequences were divided into three categories according to the degree of explicitness in the second speaker's behavior that initiates the repair: lack of response, ambiguous responses and explicit requests for repair. The complexity of the accomplishment of the repair depended on the explicitness of the repair initiation. Microanalysis demonstrated that the speakers often had problems in identifying the trouble source, which led to new problems. Both the native speakers and non-native speakers exhibited non-native features in their communication. The problems of understanding were jointly resolved in most repair sequences. The analysis suggests that a micro-analysis sheds light on features and actions that cannot be obtained by methods used in second language acquisition studies. The present data challenges conversation analysis, since there is a discrepancy between the speakers' production and intended actions. Furthermore, analyses of second language data are hindered by the restrictions in shared knowledge of participants and analyst. A study of incomplete language raises questions about the necessary prerequisites of conversation and problematizes the assumption regarding shared knowledge in conversations between equals.