Interactional problems, adaptation and code-switching in interactions involving Malay-English bilingual speakers with Traumatic Brain Injury: A conversation analysis study

Background: This thesis is an exploration of talk-in-interaction involving Malay-English bilinguals who have suffered a traumatic brain injury (TBI) and their conversation partners. Current TBI studies focus on a monolingual population, which may omit the possible influence of bilingualism on spontaneous communication. TBI studied from a conversation analysis (CA) perspective is an emerging area compared to impairment or discourse based approaches. However, compared to other acquired neurogenic disorders such as aphasia and dementia, CA studies in TBI are lacking. Aims: The key aims are: 1) to understand the problems in conversations involving persons with TBI; 2) to understand how conversation partners adapt their practices of speaking in interactions involving persons with TBI; and 3) to understand the function of code- switching behaviour in persons with TBI. Methods: Conversation analytic methods enabled a systematic analysis of talk-in-interaction using naturalistic data. Three participants with TBI of differing severities (mild, moderate and severe) and their conversation partners contributed to 95 minutes of self-recorded audio interactional data. These included a combination of dyadic and multiparty interactions recorded in the home, clinic and outdoor settings. The interaction data were sampled at at least two time points (five months apart), with one participant (with severe TBI) contributing to data over four years. Two neurotypical dyads contributed to 20 minutes of data as a form of comparison when studying code-switching behaviour. Results: The data showed that there were interactional problems in conversations involving persons with TBI. These problems included atypical physical and verbal actions (e.g. touching, singing) and problems with understandability on the part of the person with TBI or their conversation partner. The data also revealed adaptive behaviour by conversation partners of persons with TBI within talk-in-interaction. These included displaying “teacher- like” behaviours (e.g. asking test questions) and the adoption of particular interactional roles within multi-party interactions (e.g. clarifying for the person with TBI or for the conversation partner). Lastly, code-switching functions in conversations involving persons with TBI and neurotypical dyads revealed similarities (e.g. using code-switching to emphasise a point), and differences (e.g. the use of code-switching as a scaffolding tool by the conversation partner). The conversations also revealed inappropriate code- switching i.e. using a language unfamiliar to the conversation partner. Theoretical implications: There are three key theoretical implications from this study. Firstly, findings from monolingual TBI literature were replicated. These include the presence of inappropriate touching and singing, self-initiation of repair resulting in delays in TCU progressivity, understandability problems resulting from agrammatism, and the use of test questions by conversation partners. Secondly, this study details multi-party interaction in more depth than previous studies, demonstrating the collaborative behaviour of familiar and unfamiliar conversation partners in helping the bilingual with TBI navigate multi-party conversations. Lastly, using CA methodology in analyzing the interaction between TBI and neurotypical bilinguals allowed the establishment of examples of typical (e.g. to emphasize a point) and atypical (e.g. code-switching into a language not understood by the conversation partner) code-switching behaviour.