The epistemological status of logic

The dissertation evaluates the nature and strength of our warrant for axiomatic logical principles. I argue that logical principles are non-trivially a priori justified, if a priori warrant is understood in a way that is free of the inessential properties that have, historically, been associated with the concept. To defend this claim, I argue against the dominant, rival views of how logical principles are justified as well as some less familiar accounts.;Various empirical accounts of logic are considered but found to be inadequate because (1) it is crucially unclear how logical principles should be answerable to experience and (2) the need to presuppose some logical principles in the evaluation of others effectively precludes the possibility of showing all of logic to be empirically justified in a non-circular way.;Attempts to show that logic is analytic are examined, as are some more recent attempts to locate the genesis of knowledge of logic in meanings without taking logic to be straightforwardly analytic. I argue that none of these strategies succeed. Conventionalist accounts of logical truth are implausible for a number of reasons, and the alternatives trade on ambiguities that, once understood, eliminate their prospects for providing the requisite reduction of a priori justification.;Finally, I consider rule-circular arguments in support of deductive principles. But I find that these arguments fail to provide genuine justifications. So a satisfactory account of the justification of logical principles remains to be given.;An account of our first-order justification for logical principles based on the inconceivability of their invalidity under the idealization assumptions that govern their conditions for employment is defended. However, a full account of the justification of logical principles requires a metajustificatory component to show why these first-order considerations are veridical. I argue that the instrumentality of logical principles to the attainment of knowledge can provide the needed assurances. Finally, I take up and address some residual worries concerning the compatibility of my account with a naturalistic metaphysics. I show that my account is compatible with a sensible naturalism. Thus a naturalist can accept a non-reductive account of a priori justification for logical principles.