It’s an intuitive and pervasive thought that although we could be mistaken about nearly everything, we’re always in a position to know about certain aspects of our own experience. Surely, the thought goes, our access to certain conditions, like our being awake, or feeling cold, or having a visual experience as of a red object, is so direct that whenever they obtain, we can know that they do. Call a luminous condition one that, whenever it obtains, one is in a position to know that it obtains. In Knowledge and Its Limits (2000), Timothy Williamson argues that there are no luminous conditions. I present a novel way for someone sympathetic to luminosity to reject Williamson’s anti- luminosity argument, without appeal to the luminosity of the very condition that’s at issue in the argument. Selim Berker’s (2008) and Amia Srinivasan (2013) discuss what sort of principle is needed to get Williamson’s controversial margin-of-error principle out of a safety condition on knowledge. I argue that the most reasonable such principles presuppose that conditions beyond the scope of the anti-luminosity argument are not luminous, an assumption that a luminist should reject.

This paper aims to show that judgments involving credibility excess can importantly disrespect their recipients qua knowers and, indeed, that recognizing this fact will help us understand more fully what it is to respect someone as a knower in the first place. The literature on epistemic injustice suggests that credibility excess either never disrespects agents qua knowers or that, if it ever does, it is because it renders the agents epistemically arrogant. I show that this seriously miscaptures the significance of credibility excess. In cases where a person grants another excessive credibility on the basis of some “positive prejudice” toward her social identity, the judgment is alienating, objectifying, and, ultimately, lacking in a basic form of respect. I show that epistemic respect requires that our judgments be properly responsive to our evidence about others qua particular, epistemic subjects.

The pragmatist about belief claims, minimally, that there are at least some practical reasons for belief. Although I am sympathetic to this picture, I here raise a new kind of problem for pragmatism. In particular, I present considerations that make it implausible that we could ever rightly believe on the basis of such practical reasons, even if there are such reasons for belief. The problem, in general terms, is that once an agent forms a belief, their having that belief will affect other people. In particular, we invite people to depend on the truth of our beliefs through the practices of testimony and assertion. I argue that when you testify the hearer ordinarily has a right to expect that you have epistemic reason for your belief. If you testify to a belief that you hold only for practical reasons you wrong the listener. They take you to have reason to think your belief is true, but you have reason to think believing will bring you some other good. Even if you can transfer practical justification to them through testimony, I argue that you treat them paternalistically. I consider whether some practical reasons to believe outweigh the potential of wronging a listener, and if not, what kind of burdens this places on the pragmatist.

While earlier arguments in support of Uniqueness, the thesis that there is only one doxastic attitude that agents are rationally permitted to hold toward any proposition given some total evidence, have been criticized on the grounds that they fail to establish the interpersonal version of the thesis, a new kind of argument has recently emerged with the explicit aim of filling this lacuna (Dogramaci and Horowitz 2016, Greco and Hedden 2016). These arguments share a distinctive strategy: in order to vindicate Interpersonal Uniqueness, they appeal to considerations regarding our social practice of evaluating one another’s beliefs. I discuss each of the two main arguments that use this strategy and argue that they fail. Dogramaci and Horowitz’s argument hangs on the false premise that promoting rationality in an impermissive context is a more efficient means to ensure the reliability of testimony in an epistemic community than it is in a permissive context. Greco and Hedden’s argument rests on an inadequate formulation of the deference principle that underlies our epistemic evaluations. I discuss an argument by Meacham (Meacham 2018) and build on it to show that the epistemic deference principle is neutral on the question of whether Uniqueness is true.

Tracing accounts of responsibility typically hold that, for an agent A to be responsible for φ, φ must be traceable back to a point at which A meets certain conditions with respect to her future φ-ing, typically conditions grounding A’s φ-ing in a voluntary action or omission. These conditions are often thought to be something like (a) the agent exercised voluntary control over φ-ing or over a series of actions or omissions that led to her φ-ing, and (b) the agent foresaw, or could have reasonably been expected to foresee, her φ-ing, or (c) the agent’s φ-ing or failure to φ is traceable back to a fair opportunity to exercise a rational capacity that the agent knows she possesses. Tracing conditions like these are often applied to responsibility for actions, but versions of them have also been applied to various attitudes, e.g., beliefs. This paper presents a counterexample to the general tracing strategy as applied to doxastic responsibility. The counterexample is not intended to apply to any particular conception of the tracing conditions. Rather, it is intended as a counterexample to the general claim that tracing conditions must apply to beliefs if agents are to be responsible for them.