What justifies belief? : probability, normalcy, and the functional theory

Abstract

‘What justifies belief?’ This question is arguably one of the most important questions in contemporary epistemology. The first part of this study looks at two very different answers to the above question, but ultimately finds both of them wanting. According to probabilistic accounts of justification, the property that makes a belief justified is some property along the lines of being highly probable. I call this picture of justification the Lockean View. In contrast, according to the most prominent non- probabilistic accounts of justification, the property that justifies belief is some property along the lines of being true in all normal worlds. I call this non-probabilistic picture of justification the Normalcy View. However, as we will see, both families of views turn out to be problematic. While probabilistic accounts are incompatible with an attractive principle called multi premise closure (MPC), non-probabilistic accounts, I argue, are too demanding and therefore too stingy. This leaves us in a dilemma; neither probabilistic nor non-probabilistic accounts of justification seem to be wholly satisfactory. I call this the (MPC)-Stinginess Dilemma.

The second part of this study is concerned with how we should respond to this dilemma. After considering but rejecting some initial options, I argue that the dilemma can be avoided if we reject the almost universally accepted monist assumption that there is only one way for a belief to be justified; or, that there is only one property that can make a belief justified. Subsequently I develop and defend a novel, pluralist, theory of epistemic justification. I call it the Functional Theory of Justification. One upshot of the functional theory is that it makes room for the idea that there is more than one way for a belief to be justified; or, more precisely, that depending on our epistemic environment, justification can be realized by different properties.