I argue against Schroeder's explanation of pragmatic encroachment on knowledge. In section 1, I introduce pragmatic encroachment and point out that an explanation of it should avoid Pascalian considerations. In section 2, summarize the key aspects of Schroeder's explanation of pragmatic encroachment. In section 3, I argue that Schroeder's explanation faces a dilemma: it either allows for an objectionable form of Pascalian encroachment or it fails to be a fully general explanation of pragmatic encroachment.

In Part I, this paper argues that Duncan Pritchard’s version of safety is incompatible with closure. In Part II I argue for an alternative theory that fares much better. Part I begins by reviewing past arguments concerning safety’s problems with closure. After discussing both their inadequacies and Pritchard’s response to them, I offer a modified criticism immune to previous shortcomings. I conclude Part I by explaining how Pritchard’s own arguments make my critique possible. Part II argues that most modal theories of knowledge will run into problems similar to those found in Pritchard’s Anti-Luck Virtue Epistemology. I hence offer my own theory grounded in risk assessment and explain why and how it does much better.

This paper introduces a new argument for the safety condition on knowledge. It is based on the contention that the rejection of safety entails the rejection of the factivity condition on knowledge. But, since we should maintain factivity, we should endorse safery.

Can I be wrong about my own beliefs? More precisely: Can I falsely believe that I believe that p? I argue that the answer is negative. This runs against what many philosophers and psychologists have traditionally thought and still think. I use a rather new kind of argument, – one that is based on considerations about Moore's paradox. It shows that if one believes that one believes that p then one believes that p – even though one can believe that p without believing that one believes that p.

How does logic relate to rational belief? Is logic normative for belief, as some say? What, if anything, do facts about logical consequence tell us about norms of doxastic rationality? In this paper, we consider a range of putative logic-rationality bridge principles . These purport to relate facts about logical consequence to norms that govern the rationality of our beliefs and credences. To investigate these principles, we deploy a novel approach, namely, epistemic utility theory. That is, we assume that doxastic attitudes have different epistemic value depending on how accurately they represent the world. We then use the principles of decision theory to determine which of the putative logic-rationality bridge principles we can derive from considerations of epistemic utility.

Contextualism is the view that the word ‘knows’ is context sensitive and shifts according to the relevant standards in play. I argue that Contextualism is best paired with internalism about contexts. That is to say, an attributor’s context is completely determined by mental facts. Consequently, in the absence of awareness, external facts do not lead to contextual shifts. I support this view by appealing to the typical cases contextualists employ, such as DeRose’s Bank Cases and Cohen’s Airport Case. I conclude by reflecting on the nature of attributor’s themselves, and suggest this also supports the view that Contextualism is internalistic about contextual shifts.

This paper surveys our inescapable limits as cognitive agents with regard to a full world of fact: the well-known metamathematical limits of axiomatic systems, limitations of explanation that doom a principle of sufficient reason, limitations of expression across all possible languages, and a simple but powerful argument regarding the limits of conceivability. In ways demonstrable even from within our limits, the full world of fact is inescapably beyond us. Here we propose that there must nonetheless be a totality of fact, and that despite our limits we can know something of its general character. The world as the totality of fact must form a plenum, with a radically unfamiliar formal structure that contains distinct elements corresponding to each element of its own power set.

In this paper I defend a unified approach to knowledge and understanding. Both are achievements due to cognitive abilities or skills. The difference between them is a difference of aspects. Knowledge emphasizes the successful aspect of an achievement and the exclusion of epistemic luck, whereas understanding emphasizes the agent's contribution in bringing about an achievement through the exercise of one's cognitive skills. Knowledge and understanding cannot be separated. I argue against the claim that understanding is distinct from knowledge because the former is compatible with environmental luck. Achievements rule out environmental luck because abilities can be exercised only in their proper environment. I also reject the intellectualist claim that understanding requires the ability to explain what one intends to understand. The understanding of an item is reflected in our ability to solve cognitive tasks using that item. The more tasks one can deal with by using an item, the deeper is one’s understanding of that item. Being able to explain why a claim holds is not necessary for possessing understanding, even though it may be necessary for accomplishing some very specific tasks. Neither understanding nor knowledge require any kind of second-order cognition by default.

When you don’t believe a speaker’s testimony for reasons that call into question the speaker’s credibility, it seems that this is an insult against the speaker. There also appears to be moral reasons that count in favour of refraining from insulting someone. When taken together, these two plausible claims entail that we have a moral reason to refrain from insulting speakers with our lack of belief, and hence, sometimes, a moral reason to believe the testimony of speakers. Reasons for belief arising from nonepistemic sources are controversial, and it’s often argued that it’s impossible to base a belief on non-epistemic reasons. However, I will show that even if it is possible to base a belief on non-epistemic reasons, in the case of testimonial insult, for many or most cases, the moral reasons for belief don’t need to be the basis of our doxastic response. This is because there are, in many or most cases, either sufficient epistemic reasons for belief, or sufficient moral reasons for action that guide our response to testimony. Reasons from testimonial insult, in many cases, simply lead to overdetermination. Even if there are such moral reasons for belief, they are therefore practically unnecessary in many cases. There are, though, some cases in which they play an important role in guiding belief. This perhaps surprising conclusion is one unexplored way to defend epistemic over pragmatic reasons for belief.

This paper explores a particularist element in the theory of method of Imre Lakatos, who appealed to the value-judgements of élite scientists in the appraisal of competing theories of method. The role played by such value-judgements is strongly reminiscent of the epistemological particularism of Roderick Chisholm. Despite the existence of a clear parallel between the particularist approaches of both authors, it is argued that Lakatos’s approach is subject to a weakness that does not affect the approach of Chisholm.

In a recent issue of this journal, Miloud Belkoniene challenges explanationist accounts of evidential support in two ways. First, he alleges that there are cases that show explanatory virtues are not linked to the truth of hypotheses. Second, he maintains that attempts to show that explanatoriness is relevant to evidential support because it adds to the resiliency (stability) of probability functions fail. I contest both of Belkoniene’s claims.

This paper puts forth a functionalist difficulty for Sally Haslanger’s proposal for engineering our concept of ‘woman.’ It is argued that the project of bringing about better political function fulfillment cannot get off the ground in virtue of epistemic failure.

I have argued that Gettier cases are misleading because, even though they appear to be cases of knowledge failure, they are in fact cases of semantic failure. Atkins has responded to my original paper and I have replied to his response. He has then responded again to insist that he has the so-called “Gettier intuition.” But he now admits that intuitions are only defeasible, not conclusive, evidence for and/or against philosophical theories. I address the implications of Atkins’ admission in this paper and I again show that his attempts to revise Gettier’s original cases such that they do not involve semantic failures are unsuccessful.