Those who endorse a knowledge-first program in epistemology claim that rather than attempting to understand knowledge in terms of more fundamental notions or relations such as belief and justification, we should instead understand knowledge as being in some sense prior to such concepts and/or relations. If we suppose that this is the correct approach to theorizing about knowledge, we are left with a residual question about the nature of those concepts or relations, such as justification, that were thought to be first but are now second. Jonathan Jenkins Ichikawa has recently proposed that we understand justification in terms of potential knowledge. Ichikawa combines his view of knowledge and justification with what initially seems to be a natural complement, epistemological disjunctivism. While Ichikawa focuses on hallucination, I shift the focus to illusion. I argue that the combination of justification as potential knowledge and epistemological disjunctivism entails that perceptual beliefs that arise from illusions are not justified.

The aim of this paper is to reexamine the importance of Rip van Winkle’s case for the problem of cognitive dynamics. First I shall present the main problem of cognitive dynamics. Then I shall explain the relevance of Rip’s case to this problem. After that I shall provide a short presentation of the main solutions to this problem. I shall explicate the problem concerning the manner in which philosophers who propose those solutions defend their response to the question of Rip’s case. My argument shall be that they defend their response either in overly dogmatic or in circular way. Finally, I shall suggest a way out of that problem.

How strongly an agent beliefs in a proposition can be represented by her degree of belief in that proposition. According to the orthodox Bayesian picture, an agent's degree of belief is best represented by a single probability function. On an alternative account, an agent’s beliefs are modeled based on a set of probability functions, called imprecise probabilities. Recently, however, imprecise probabilities have come under attack. Adam Elga claims that there is no adequate account of the way they can be manifested in decision-making. In response to Elga, more elaborate accounts of the imprecise framework have been developed. One of them is based on supervaluationism, originally, a semantic approach to vague predicates. Still, Seamus Bradley shows that some of those accounts that solve Elga’s problem, have a more severe defect: they undermine a central motivation for introducing imprecise probabilities in the first place. In this paper, I modify the supervaluationist approach in such a way that it accounts for both Elga’s and Bradley’s challenges to the imprecise framework.

In its most basic form, Simple Reliabilism states that: a belief is justified iff it is formed as the result of a reliable belief-forming process. But so-called New Evil Demon (NED) cases have been given as counterexamples. A common response has been to complicate reliabilism from its simplest form to accommodate the basic reliabilist position, while at the same time granting the force of NED intuitions. But what if despite initial appearances, Simple Reliabilism, without qualification, is compatible with the NED intuition? What we can call the Dispositionalist Response to the New Evil Demon problem is fascinating because it contends just that: Simple Reliabilism is fully compatible with the NED intuition. It is claimed that all we need to do to recognize their compatibility is appreciate that reliability is a dispositional property. In this paper I shall critically evaluate the Dispositionalist proposal.

Kopec and Titelbaum collect five alleged counterexamples to Uniqueness, the thesis that it is impossible for agents who have the same total evidence to be ideally rational in having different doxastic attitudes toward the same proposition. I argue that four of the alleged counterexamples fail and that Uniqueness should be slightly modified to accommodate the fifth example.

This paper contains a critical examination of the prospects for analyses of knowledge that weaken the factivity condition so that knowledge implies only approximate truth.

This paper proposes a functional approach to characterize epistemic and nonepistemic values. The paper argues that epistemic values are functionally homogeneous since (i) they act as criteria to evaluate the epistemic virtues a hypothesis ought to possess, and (ii) they validate scientific knowledge claims objectively. Conversely, non-epistemic values are functionally heterogeneous since they may promote multiple and sometimes conflicting aims in different research contexts. An incentive of espousing the functional approach is that it helps us understand how values can operate in appropriate and inappropriate ways in scientific research and inappropriate influences can eventually be prevented. The idea is to argue that since non-epistemic values are functionally heterogeneous, they cannot provide objective reasons for the acceptance of a hypothesis. However, their involvement is necessary in certain research contexts and the problem is the involvement of these need not be always legitimate. By analyzing a case from chemical research, I demonstrate that how non-epistemic values might influence scientific research and, then, I go on to demonstrate that how a proper understanding of the functions of different kinds of values might promote the attainment of multiple goals of a particular research in a legitimate and socially relevant way.

The Enkratic Principle enjoys something of a protected status as a requirement of rationality. I argue that this status is undeserved, at least in the epistemic domain. Compliance with the principle should not be thought of as a requirement of epistemic rationality, but rather as defeasible indication of epistemic blamelessness. To show this, I present the Puzzle of Inconsistent Requirements, and argue that the best way to solve it is to distinguish two kinds of epistemic evaluation – requirement evaluations and appraisal evaluations. This allows us to solve the puzzle while accommodating traditional motivations for thinking of the Enkratic Principle as a requirement of rationality.

One defense of the “steadfast” position in cases of peer disagreement appeals to the idea that it's rational for you to remain deeply agnostic about relevant propositions concerning your peer's judgment, that is, to assign no credence value at all to such propositions. Thus, according to this view, since you need not assign any value to the proposition that your peer's judgment is likely to be correct, you need not conciliate, since you can remain deeply agnostic on the question of how the likelihood of your peer's judgment bears on the likelihood of your own. This paper argues that the case for deep agnosticism as a response to peer disagreement fails. Deep agnosticism (as a general thesis) implies that it is sometimes permissible to withhold judgment about whether there is a non-zero chance of a proposition's being true. However, in cases of disagreement where deep agnosticism is supposed to support the steadfast position, such withholding isn't rational. This is because of constraints placed on rational credence by objective probability or chance, which ensure that rational credence adequately reflects strength of evidence.

Epistemic contextualism (EC) is the thesis that the standards that must be met by a knowledge claimant vary with (especially conversational) contexts of utterance. Thus construed, EC may concern only knowledge claims (“Semantic EC”), or else the knowledge relation itself (“Substantive EC”). Herein, my concern is with “Substantive EC.” Let’s call the claim that the sorts of linguistic evidence commonly cited in support of Semantic EC also imply or support Substantive EC the “Implication Thesis” (IP). IP is a view about which some epistemologists have equivocated. Keith DeRose is a case in point. Herein I argue that IP is false, and that it is false for interesting reasons. To this end, I consider two other terms which DeRose investigates, “free will” and “potency” in his efforts to demonstrate the alleged inability of distinctly philosophical or skeptical doubts to infect ordinary epistemic discourse. I describe how and why these two examples speak against, rather than for, DeRose’s recommendation of Substantive EC.

A growing cohort of philosophers argue that inference, understood as an agent-level psychological process or event, is subject to a “Taking Condition.” The Taking Condition states, roughly, that drawing an inference requires one to take one’s premise(s) to epistemically support one’s conclusion, where “takings” are some sort of higher-order attitude, thought, intuition, or act. My question is not about the nature of takings, but about their contents. I examine the prospects for “minimal” and “robust” views of the contents of takings. On the minimal view, taking one’s premise(s) to support one’s conclusion only requires focusing on propositional contents and putative epistemic support relations between them. On the robust view, taking one’s premise(s) to support one’s conclusion also requires knowledge (or being in a position to have knowledge) of the attitudes one holds toward those contents. I argue that arguments for the Taking Condition do not entail or sufficiently motivate the robust view. Accordingly, contra several philosophers, the Taking Condition does not illuminate a deep relationship between inference and self-knowledge.

In this essay, Crispin Wright’s various attempts at solving the so-called McKinsey paradox are reconstructed and criticized. In the first section, I argue against Anthony Brueckner that Wright’s solution does require that there is a failure of warrant transmission in McKinsey’s argument. To this end, a variant of the McKinsey paradox for earned a priori warrant is reconstructed, and it is claimed that Wright’s putative solution of this paradox is best understood as drawing on the contention that there is a transmission failure in the argument in question. In section II, I focus on Wright’s views in the second part of his pivotal article on the McKinsey paradox (published in 2003). It is argued that the solution to the paradox proposed there by Wright is convincing if his theory of entitlements is accepted. In the third section, however, I raise an objection against Wright’s account of entitlements. Finally, in section IV, Wright’s views in his most recent essay on the McKinsey paradox are examined. It is shown that his new solution to this problem does not work any better than his earlier attempts at solving it.

There is a very plausible principle linking abilities and possibilities: If S is able to Φ, then it is metaphysically possible that S Φ’s. Jack Spencer recently proposed a class of counterexamples to this principle involving the ability to know certain propositions. I renew an argument against these counterexamples based on the unknowability of Fitch propositions. In doing so, I provide a new argument for the unknowability of Fitch propositions and show that Spencer’s counterexamples are in tension with a principle weaker than the one linking abilities and possibilities.