Abstract

In this paper I present a novel approach to resolving a problem for inferentialist metasemantic theories that arises from a naturalistic point of view. According to normative inferentialists, expressions have their semantic content in virtue of the inferential rules that govern them. This poses a problem because a rule is a naturalistically suspicious normative notion that is used by normative inferentialists as an explanatory tool. I propose a reconciliation of naturalism and inferentialism by adopting a quasi-realist view on inferential rules. According to this approach, inferential rules viewed externally are considered non-normative constitutive rules that are implicit in the dispositions of a linguistic community to allow and prohibit certain inferences. Conversely, from the internal perspective on linguistic practice, inferential rules are treated as regulative and carry a normative force, i.e., they have an action-guiding character because of the fact that they are individually accepted by members of a linguistic community.