The emergence of “translanguaging” as a concept referring to bilingual practices has challenged the appropriateness of “code-switching” – the term that has been most influential in studies of bilingualism and language mixing. Reassessing the literature on Cantonese-English mixing in Hong Kong, this paper suggests that the kind of spontaneous code-switching in peer talk, largely intra-sentential (or intra-clausal) and intra-turn, can indeed be recast as translanguaging, where speakers transcend language boundaries between Cantonese and English for the purpose of meaning-making. Nevertheless, Hong Kong speakers do constantly draw language boundaries by marking words as English or Cantonese, both in metalinguistic judgment and in real-time language production. Revisiting an unpublished dataset of radio talk, this paper further illustrates a number of sequences in which Cantonese speakers may “languagise” the code-switched words or expressions as “English”. It is concluded that, in a Conversation-Analytic understanding, the difference between “translanguaging” and “code-switching” boils down to “languagising”, and the contrast between the two notions may have been overstated.