UC Berkeley

We are no less judgmental about what others ought to believe than we are about what others ought to do. Yet our judgments about what agents ought to believe can seem mysterious in ways that judgments about what agents ought to do are not. My dissertation provides an account of how to make sense of these deontic judgments about beliefs with an eye to some of the puzzles that they naturally generate. I argue that we can make sense of genuinely prescriptive epistemic ‘ought’ judgments, and in a way that allows us to largely preserve the broad similarities between epistemic ‘ought’ judgments and practical ‘ought’ judgments.

Chapter 1 provides an overview of the dissertation by setting up and motivating the questions to which the later chapters respond. In the first substantive half of the dissertation (Chapters 2 and 3), I examine some foundational questions about the nature of the epistemic norms that underlie prescriptive epistemic ‘ought’ judgments. In Chapter 2, we examine the nature of evidential norms that enjoin agents to proportion their beliefs to the evidence. If what agents ought to believe depends on what the evidence suggests, what explains why this is so? According to a widespread view among epistemologists, evidential norms are derivatively normative, owing their normative status to some role they play in getting us to form true beliefs. Against this view, I argue that evidential norms are not derivatively normative. In Chapter 3, we take up a related challenge. Here we ask: Why does it seem that epistemic norms are specially privileged with respect to non-epistemic norms when it comes to assessing doxastic attitudes? I argue against a number of prominent positions here that aim to answer this question by identifying some unique property that epistemic norms have that gives them their overriding normative status vis-à-vis non-epistemic norms. I then argue, in a more deflationary vein, that the priority of epistemic norms is grounded in the fact that the only deliberative questions that can be live for us, when it comes to deliberating about what to believe, are questions that can be settled by appealing to epistemic reasons.

In the second substantive half of the dissertation (Chapters 4 and 5), we examine the nature of our judgments about what agents ought to believe. An influential view in epistemology holds that judgments about what agents ought to believe cannot be genuinely normative, since agents lack control over what they believe, and agents can be required or obligated to do something only if they have some control over whether or not they can bring it about (in other words, “ought implies can”). In recent years, many epistemologists have rejected this view by arguing that agents do have a form of control over what they believe: agents can form and revise their beliefs for reasons. And the Ought-Can principle should be understood in terms of reasons-responsiveness. In Chapter 4, I explore different ways of formalizing the Ought-Can principle in terms of reasons-responsiveness and argue that the no version of Ought-Can is plausible. I argue, instead, that we should reject Ought-Can, and I offer an error theory for Ought-Can that makes this surprising rejection more palatable.

In Chapter 5, I develop my positive proposal about how to make sense of epistemic ought judgments in a way that is compatible with our rejection of Ought-Can. The view I develop and defend, contextualist evidentialism, holds that epistemic ought judgments are true relative to the body of evidence that is relevant at a context of use. I motivate and defend this view against the contemporary orthodoxy in epistemology that holds that epistemic ought judgments are true relative to the body of evidence available to the subject of the judgment. The view I offer builds directly off of recent work in the philosophy of language, and aims to apply this work to contemporary epistemology in a way that opens up new possibilities in the normative landscape that have remained as yet unexplored.

I conclude, in Chapter 6, by exploring further questions that come out of my positive proposal.