An analysis of conventionalism in early analytic philosophy

This dissertation is a critical examination of the doctrines of conventionalism that are defended by philosophers in the early Analytic tradition. The burden of this dissertation is to establish that Quine's argument that the fact-convention distinction dissolves in the face of epistemological holism depends on an artificially narrow conception of a convention. I wish to show, first, that Poincare uses the term 'convention' to designate not only those principles that can be maintained in the face of empirical evidence, but also definitions that make it possible to generate evidence for or against our abstract scientific theories in the first place. Second, I will argue that Schlick defends an account of the revision of such definitions that makes it clear how these revisions differ from the changes that we are compelled to make to the empirical components of our theories. Finally, I will maintain that Carnap's position that traditional metaphysics be replaced by the 'logic of science' is not itself a proposal to be evaluated on the basis of its pragmatic merits, but the result of his careful analysis of the function of philosophical claims within the framework of our theoretical knowledge. All three philosophers use the term 'convention' to designate not only those principles that we refuse to revise, but those definitions whose special status is due to the unique role that they play in the organization of our theoretical knowledge. It is this fact that renders Quine's argument inapplicable to the conventionalist doctrines of Poincare, Schlick, and Carnap.;Key words: conventionalism, a priori knowledge, analyticity, Poincare, Schlick, Carnap, holism, philosophy of science, epistemology.