Joe Salerno

The project is to demonstrate important connections between our metaphysical views and the logical principles we endorse. For example, acceptance of the principle that every proposition is determinately true or false commits us to certain "metaphysical realist" views about the existence of mind-independent facts. Conversely: acceptance of certain metaphysical views necessitate rejection of certain logics in favor of others. ;Current debates offer several views about the relation between logic and metaphysics. "Semantic anti-realism" says that truth is essentially constrained by possible knowledge: that is, that all truths are knowable, and that once this is acknowledged, we must revise classical logic. The arguments for this view are not compelling; I suggest that Dummett and others do not provide a precise enough account of the relation between logic and semantic anti-realism. My main positive contribution is a revised formulation of the purported connection. It establishes, in a way that others have not, that semantic anti-realism does issue a demand to revise our principles of logic. ;Another form of anti-realism suggests that certain indicative statements serve not to represent facts, but rather to express sentiments. Such "irrealist" metaphysical views are said to be compatible with classical principles of inference. I argue that this is a mistake: despite Blackburn's "quasi-realist" attempts to show that the expressivist can coherently accept Excluded Middle, the expressivist cannot consistently endorse the possibility of unknowable moral truths. So this form of metaphysical irrealism entails an important form of semantic anti-realism , and therefore, shares its consequences for logical revision. ;One general result is that various metaphysical theories necessitate radical revisions in our inferential rules. Another result is that endorsement of classical logic implies acceptance of various forms of metaphysical realism . Upshot: our logic really does, in some important sense, guide our metaphysics, and vice versa