Abstract

A metalinguistic negotiation occurs when conversational partners are seemingly having a descriptive dispute about the world, but instead they are having a normative dispute about what the words they are using should mean. Cappelen (2018) argues that if semantic externalism is true, the meanings of our words are beyond our control. If this is the case, however, metalinguistic negotiation cannot be successful unless by a complete fluke. I examine whether Cappelen’s conclusion follows the orthodox externalist views. I conclude that it does not straightforwardly do so, but more needs to be said about the relationship between metalinguistic negotiation and meaning control. I take on that task, introducing a framework that distinguishes between conversation-level and linguistic community-level metalinguistic negotiations, which I use to draw two conclusions. The first, and more modest one, is that Cappelen’s worry turns out to be nothing more than the platitude that one speaker cannot by herself change the meaning of a word for the whole linguistic community, which does not in any way show that individual acts of metalinguistic negotiation are not worth pursuing. The second, and less modest one, is that metalinguistic negotiation is (one of the) main driving mechanism(s) for meaning change under a social externalist picture. Finally, I reflect on what these two conclusions mean for conceptual engineering and how we should go about improving our representational devices.