Many studies in recent years identify and discuss Dutch ethnolects. Generally this work takes linguistic phenomena as directly reflective of speakers' ethnic identity. But if ethnicity is an inherent speaker feature, the absence of white ethnolect descriptions is difficult to explain. In this paper, therefore, I wish to judge the appeal of the notion of ethnolect against its usefulness for explaining language use. I argue that ethnolect can usefully label everyday ethnicisations of language, but that such evaluations generally compress a more complex reality in which so-called ethnolectal features are recruited for other purposes than (un)marking one's ascribed ethnicity. Crucial to unpacking this reality is the distinction of an intermediary step, the construction of interactional stance, between the use of linguistic features and their association with common-sense identity categories. ; SCOPUS: ar.j ; info:eu-repo/semantics/published