Translanguaging is a theory of multilingual practice developed most influentially by Ofelia Garcia and Li Wei (2014). It holds that multilingual speakers do not operate with separate, bounded language systems but draw on a single integrated linguistic repertoire. The term was coined by Cen Williams (trawsieithu) in the 1980s to describe a Welsh bilingual education practice of switching the language of input and output within a lesson. Garcia and Li Wei transformed it into a broader theoretical framework that challenges the very notion of "named languages" as discrete objects.
Translanguaging is not code-switching by another name. Code-switching describes alternation between identifiable languages and assumes those languages exist as separate systems. Translanguaging rejects that premise: a bilingual's linguistic resources form one repertoire, and what observers label "switching" is simply the speaker deploying features from that repertoire as the communicative situation demands. The distinction matters because code-switching still operates within a monoglossic ideology (two separate codes), while translanguaging adopts a heteroglossic one (one system with socially named subsets).
Researchers distinguish two modes of translanguaging in educational settings:
These are not dichotomous categories but points on a continuum: well-designed pedagogical translanguaging creates conditions for productive spontaneous translanguaging.
Translanguaging raises practical questions for English language classrooms:
The strong form of translanguaging, that separate languages do not exist, creates tensions. Teachers must still teach "English" as a named language. Multi-competence offers a more pragmatic alternative: it acknowledges that languages are distinct systems that interact dynamically within one mind, without denying their separateness. Critics also note that translanguaging can be co-opted to justify unsystematic L1 use that lacks pedagogical purpose.