This presentation provides an overview of my research and relates it to the currently popular concept in the study of bilingual language practices, namely translanguaging. Translanguaging came about as a specific language teaching pedagogy, but it has since been claimed to be a theory of language and of bilingual language practices in particular. As a theory of bilingual language use, its key concern is whether, in actual interaction as in their mental grammars, bilinguals keep their languages separate. Its answer to this concern is the claim that the bilingual’s linguistic repertoire consists, not of languages, but of linguistic features which are not differentiated in terms of named languages. Just like translanguaging, my work started from a dissatisfaction with the notion that, in interaction, Rwandan bilinguals ought to keep Kinyarwanda and French separate and therefore that Kinyafrancais, the actual everyday language practice of bilingual Rwandans, is a problem which needs to be eradicated. However, recognising that the notion of language is an appropriate ‘scheme of interpretation’ (ethnomethodological sense) for some of the bilingual language choice practices, unlike in translanguaging theory where the notion of language is rejected outright, in my work I propose only to suspend it, to approach the study of language choice practices with an attitude of ‘indifference’, with the attitude of a language-blind observer. The result is, as I show in the presentation, a more complete account of language choice practices among bilingual speakers.