Abstract

This thesis examines necessary truth and defends a normative, or linguistic, account of it. Roughly, it holds that necessary truths state or follow from conceptual norms (i.e., norms that determine patterns of correct concept use). While the thesis touches upon logical and mathematical truth, its primary focus are those necessary truths typically expressed using natural language. The thesis has three parts. In Part I, I criticise metaphysical accounts of necessity and present and defend a normative account of it. At no point do I give a history of normative accounts, but clearly their roots are to be found in the first half of the twentieth century – in the works of Wittgenstein and Carnap, for example. In Part II, I consider whether language can sustain the normative account. Some argue that the account requires language to be regimented in a way that it is not. I show that while it requires a distinction in kind between empirical and conceptual principles, it nevertheless makes room for indeterminacy regarding whether a given statement is an empirical claim or follows from conceptual norms. Finally, in Part III, I consider the relationship between the world and our conceptual scheme. I argue that denying our concepts answer to the world does not mean that they cannot be justified. The normative account does not say that we have no reasons for categorising things in a certain way, but rather that natural facts, in combination with our interests, are fit to provide them. The purpose of the thesis is to show that normative accounts of necessity can be much more robust than they are often given credit for and needn’t have the malign implications often associated with them.