This paper argues that a functional approach to language can provide the foundation for an explicit account of the status of linguistic facts, bridging the 'internalist' and 'externalist' views of language. This position is seen as a development of the position of Searle on social facts and of Keller on language change.