Title: Cognitivism and Non-Cognitivism about Logical Inference

Is there a special relationship between logic and reasoning? Some people think there is in virtue of the fact that reasoners often perform so-called logical inferences—i.e, deductive inferences that are somehow governed by logical principles. How to make sense of inferences of this kind, however, is not so clear. In this talk, I focus on one puzzle regarding logical inference, known as the Adoption Problem. The Adoption Problem says that if someone does not already use certain fundamental logical principles in their reasoning, then they cannot rationally begin to do so, because beginning to use these principles requires that one already be using them; hence, the adoption of these principles by any individual is either unnecessary or impossible. This is meant to be a problem for anyone who thinks that logical-inferential habits are subject to rational evaluation and revision. After reconstructing the argument for the Adoption Problem and providing an overview of recent responses in the literature, I show that the puzzle is in fact a useful tool for making explicit different philosophers’ views on logical inference, which are usually left implicit. I propose a classification of different responses to the Adoption Problem as either cognitivist or non-cognitivist about logical inference. This classification, I argue, allows us to understand the fundamental issues surrounding the nature of logical inference more clearly. In the final section of the talk, I present an argument against cognitivism about logical inference—and inference more generally—and I suggest that a species of rationalist non-cognitivism about Iogical inference is likely the most fruitful view to pursue.