Discourse markers in contrast: But, actually and well in native-nonnative English conversations between friends

Ordinary conversation remains a relatively unexplored area in Second Language Acquisition. Little is known about how learners acquire and carry out the basic discursive practices in social life outside the classroom. This study examines interactional disharmony arising from nonnative-like use of discourse markers in dyadic conversations between friends---a Korean learner and a native speaker of English. Approximately ten and a half hours of conversational interaction was audiotaped or videotaped and transcribed. Results reveal that interactional disharmony between the two participant groups is particularly noticeable in the use of three contrastive markers---but, actually and well---in topic management moves and oppositional talk. From a discourse analytic perspective, each marker is analyzed in terms of sequential environments and accomplishment of interactional task. Findings suggest that discourse markers play an important role in socioculturally situated and interactionally motivated use of language. Native speakers use discourse markers to create and reinforce positive social relationships, specifically in relating themselves to different points in conversation and in dealing with interactional difficulties. In contrast, Korean learners are found to be unaware of the context-specific and social use of discourse markers. They appear to be mainly concerned with maintaining topical flow and textual coherence of the conversation in progress. They use discourse markers to structure and organize their utterances in order to be intelligible, orderly and coherent. Consequently, learners fail to meet native-speaker expectations in locally managed interpersonal relationships. This study concludes with the suggestion that discourse markers be regarded as interactional features, rather than cohesive devices, that emerge from dynamic meaning making and relationship management processes. Furthermore, this study provides empirical evidence that pragmatic failure may not always be explained as an incorrect form-function mapping or a miscalculated assessment of the given social context. Additionally, it is equally important to use linguistic features in correspondence to the interactional expectations and contingencies for the development of communicative competence.