Abstract Using data from Rwandan multilingualism and applying a CA-informed qualitative methodology, this paper contributes to the debate about the nature of translanguaging, understood as the use of linguistic features from different languages without any orientation to language boundaries. To do this, the paper explores two empirical questions. First it is observed that, in bilingual conversation, practices which can be correctly described as translanguaging (may) co-occur with those which cannot be so described (referred to globally as ‘code-switching’). Therefore, the paper raises the question of the logic behind speakers’ decision to translanguage or not to translanguage. In the article, it is proposed that a bilingual conversation is by definition a site where two conversational preferences are at work, namely the preference for progressivity and the preference for noticing the languageness of the linguistic features used. Analysis of the data shows that code-switching and translanguaging are alternative strategies that bilingual participants use in managing this situation of multiple preferences. Secondly, because, in interaction, a linguistic feature can be oriented to as translanguaging just as it can be oriented to as code-switching, a question arises as to how the actual status of a linguistic feature is arrived at. In the paper, I demonstrate that the status of a linguistic feature either as translanguaging or as code-switching is an achieved status, i.e. is locally negotiated between participants at the point of its use.
Joseph Gafaranga (Sun,) studied this question.