Jarrett Leplin’s research while affiliated with University of North Carolina at Chapel Hill and other places

Reliability theories of epistemic justification and knowledge originated in the late 1970s and flourished through the 1990s. Although still influential, reliabilist epistemology is widely thought to be seriously defective, and has been largely superceded by such (purported) rivals as evidentialism and virtue theory. This book takes exception to this development. I contend that reliabilism remains an important part of the true story of justification. I will develop a new reliability theory free from the burdens that discredited earlier theories. But the solutions my theory offers to counterexamples and objections raised against other theories are incidental to its motivation and development.

According to my theory, paired down, a method of belief-formation is justificatory if it is counterfactually sensitive to falsity. The method is not required to be sensitive to truth. Thus, mine is not a tracking theory; belief-formation does not covary with, or track, truth-value.This is unfortunate with respect to our epistemic goal. But it was to be anticipated. It reflects the greater relative importance of falsity-aversion over truth-acquisition within this goal. Believing falsely is epistemically worse than failing to believe truly. No general method can guarantee the acquisition of truth and nothing but.

Is the reliability of the process by which a belief is formed necessary for the belief’s justification? Are there not justified beliefs that would have been formed though false, by the method that did form them, even under normal conditions? Consider a person who mistakes certain bushes for trees, but is a reliable identifier of redwood trees. Observing a redwood, he believes justifiedly that there is a tree before him and this belief is justified, but he might have held this belief falsely by observing a bush.

I propose to explicate truth-conduciveness in terms of reliability. In this chapter I shall be concerned both with the reliability of processes or methods by which beliefs are formed, and with the reliability of the formation of beliefs. These are distinct notions. Each incorporates a condition unique to the theory of justification I shall propose: The reliability of methods introduces a condition of normalcy, and the reliability of belief-formation introduces a condition of intentionality. Much of this chapter is devoted to explaining these conditions and applying them to problems raised by the notion of reliability.

Truth-preserving inference from reliably formed beliefs is a reliable method of forming beliefs. Beliefs formed by truth-preserving inference from beliefs formed by a method that would not, under normal conditions, have produced them had they been false will not, under normal conditions, be false. For, truth-preserving inference from truths yields truths under all conditions. Thus, truth-preserving inference from reliably formed beliefs satisfies the first clause of condition A of my theory. If the beliefs from which one infers also satisfy the second clause of A and are thereby justified, then the beliefs inferred will also satisfy this clause and be justified. For a justified belief incompatible with an inferred belief must also be incompatible with beliefs that support the inference. Hence, truth-preserving inference from beliefs justified under A yields beliefs justified under A. By condition B, a good reason to believe that one’s belief is formed by truth-preserving inference from beliefs justified under A justifies holding the belief.

According to the paradox of the preface, one is justified in disbelieving the conjunction of one’s beliefs. CJC cannot allow this (supposing one believes justifiedly), but what is to prevent it? One’s own experience, like that of anyone else, cautions against confidence in one’s belief-system as a whole. It would seem to be justified, indeed compelled on pain of hubris, to expect some among one’s present beliefs to turn out false. This expectation is not localizable; it impugns no belief in particular. Hence one’s (presumed) justifications for one’s beliefs, taken individually, are unaffected by it. But the justification that CJC supplies for their conjunction is defeated.

My theory does not require for one’s justification of a belief that one possess a general capacity or faculty for discriminating truths from falsehoods across a range of propositions to which the belief belongs. Such a faculty is a kind of intellectual virtue. It abstracts away from particular cognitive states or acts to an abiding feature of intellectual life, much as moral virtue abstracts away from particular right acts to stable traits of moral character. This chapter explains why some philosophers require intellectual virtue for justification, and why it is a mistake to do so.

I can now give a preliminary statement of my theory. The theory is in two parts: A. A belief is epistemically justified if it is reliably produced or sustained and no incompatible belief is epistemically justified. B. A person is epistemically justified in believing a proposition that he has good reason to believe is an epistemically justified belief.

The theory of epistemic justification I shall advance is based on an assumption: epistemic justification is justification that promotes the epistemic goal of believing truths without believing falsehoods. This chapter explains why an assumption is needed, motivates my choice of what assumption to make, and clarifies what is being assumed.

This book proposes an original theory of epistemic justification that offers a new way to relate justification to the epistemic goal of truth-conducive belief. The theory is based on a novel analysis of reliable belief-formation that answers classic objections to reliability theories in epistemology. The analysis generates a way of distinguishing justified belief from believing justifiedly, such that inerrant belief-formation need not be justificatory whereas systemic deception could be. It thereby respects the intuition that standards for justification must be accessible to the believer, while maintaining the essential connection of justification to truth. The analysis shows how justification relates to, but is distinct from, evidence, rationality, and probability. It provides a unifying treatment of issues central to current debate in epistemology, including epistemic paradoxes, epistemic closure, skepticism, contextualism, virtue theories, the effect of luck on knowledge and justification, the interpretation of subjunctive conditions for justification, the conflict between internalism and externalism, and metaphilosophical evaluation of epistemological theories. There are further applications to metaphysics, the philosophy of language, the philosophy of science, and ethics. The book will engage philosophers working in epistemology or related fields, and their graduate students.