Following Kelly (2010), Meacham (2014) and Schoenfield (2014), credence permissiveness seems highly plausible. Following Pettigrew (2016a; 2016b), measuring epistemic disutility through strictly proper scoring rules (like the Brier score) also seems highly plausible. However, both claims are incompatible. How do we solve that problem? In this paper we argue that 1- credence permissiveness and strictly proper scoring rules are incompatible; 2- that holding uncalibrated scoring rules can lead an agent to hold inconsistent credences; and 3- that credence permissiveness conflicts with a basic rationality requirement, namely, being responsive to new evidence.

Testimony is encoded into memory, changing the information of that memory. The testimony changes the memory into misinformation, and other times it fixes it to produce true beliefs. These cases are called “helpful incorporation.” A worry about cases of helpful incorporation is the accuracy of those beliefs will largely be subject to luck. Kourken Michaelian argues we can count helpful incorporation as knowledge because testimony is generally reliable. He argues people are generally sincere, meaning they don’t lie. And we generally perceive people as being credible. These two bias work together to produce reliable testimony. I pose Fricker-style objections to show testimony is actually not nearly as reliable as Michaelian argues it to be. Negative stereotypes about groups of people make it less likely that their true testimony will be seen as sincere. This has the odd effect of making prejudiced people’s beliefs more conservative, and less likely to incorporate false testimony into their memory. Positive stereotypes, however, make it more likely the speaker will be seen as credible when they are not. Given our prejudices, we’re not reliably good at picking out credible testimony. As a result, we should not trust that helpful incorporation is more reliable than luck.

The epistemic injustice literature has focused on wrongful epistemic harms. I argue that such harms can sometimes be permissibly inflicted as epistemic punishments. When agents are epistemically culpable for their false testimony, we have epistemic reasons to epistemically harm them. In addition to being interesting in their own right, epistemic punishment helps to give substance to views which appeal to epistemic blameworthiness or excuse. In section 2, I present a catalog of epistemic offenses and epistemic punishments. In section 3, I argue that the things I call epistemic punishments really are. In particular, they are (i) epistemic harms, (ii) inflicted in response to (alleged) epistemically culpable wrong-doing that (iii) would otherwise epistemically wrong someone subject to them. Section 4 concludes by briefly discussing how my view fits in with other views of punishment, and points out interesting new questions for the future.

Both luck and disagreement are widely discussed topics in contemporary epistemology. And yet there are almost no sustained treatments of the relationship between luck and disagreement. Why has so little been written on the subject? Are there simply no interesting connections between luck and disagreement? I maintain, on the contrary, that luck and disagreement are conceptually and normatively linked and that this linkage should be of interest to anyone concerned with either luck or disagreement. Moreover, I argue that evidence of peer disagreement just is evidence of knowledge-precluding luck, that the normative significance of peer disagreement reduces to the normative significance of knowledge-precluding luck, and that one’s theory of luck bears directly on one’s approach to peer disagreement. The ultimate upshot is that luck and disagreement are closely related and that their relatedness matters.

Lisa Miracchi and Christoph Kelp have argued against indirect virtue epistemology (IVE), called “indirect” since “it aims to analyze knowledge in terms of a competence to do something other than know”. Kelp’s and Miracchi’s arguments turn on the putative inability of IVE to accommodate widespread intuitions about crucial cases. For Kelp, IVE cannot simultaneously accommodate our intuitions regarding both epistemic Frankfurt cases and Fake Barn cases. For Miracchi, IVE is unable to explain why we fail to know in certain Gettier cases.

In this paper, I argue that John Greco’s version of IVE has the resources to (1) explain away our intuition about Kelp’s epistemic Frankfurt case while preserving the intuition about Fake Barns; and (2) explain why one fails to know in Miracchi’s Gettier cases. Hence, there is at least one brand of IVE that can withstand Kelp’s and Miracchi’s objections.