David Lewis's unified theory of the contents of de se and de dicto attitudes faces a problem. Whether or not it is adequate for representing beliefs, it misrepresents the content of many of our desires, which rank possible outcomes in which the agent with the desire does not exist. These desires are shown to playa role in the rational explanation of action, and recognising them is important in our understanding of ourselves.

My purpose in this essay is to clarify and evaluate Arthur Schopenhauer's grounds for the view that happiness is impossible. I shall distinguish two of his arguments for that view and argue that both of them are unsound. Both arguments involve premises grounded on a problematic view, namely, that desires have no objects. What makes this view problematic is that, in each of the two arguments, it conflicts with Schopenhauer's grounds for other premises in the argument. I shall then propose a way of fixing both arguments. The solution involves substituting the view that desires have no objects with the view that we have a desire to have desires. The latter view, I shall argue, can do the grounding work that the former does in Schopenhauer's arguments but, unlike it, the view that we desire to desire is consistent with Schopenhauer's grounds for the rest of premises in those arguments.

It is sometimes suggested that there are two kinds of reasoning: inferential reasoning and non-inferential reasoning. However, it is not entirely clear what the difference between these two kinds of reasoning is. In this paper, I try to answer the question what this difference is. I first discuss three answers to this question that I argue are unsatisfactory. I then give a different answer to this question, and I argue that this answer is satisfactory. I end by showing that this answer can help to resolve some disagreements in which the difference between inferential and noninferential reasoning plays a role.

Traditional approaches to epistemology have sought, unsuccessfully, to define knowledge in terms of justification. I follow Timothy Williamson in arguing that this is misconceived and that we should take knowledge as our fundamental epistemological notion. We can then characterise justification as a certain sort of approximation to knowledge. A judgement is justified if and only if the reason (if there is one) for a failure to know is to be found outside the subject’s mental states; that is, justified judging is possible knowing (where one world accessible from another if and only if they are identical with regard to a subject’s antecedent mental states and judgement forming processes). This view is explained and defended.

There has recently been considerable interest in accounts of fiction which treat fictional characters as abstract objects. In this paper I argue against this view. More precisely I argue that such accounts are unable to accommodate our intuitions that fictional negative existentials such as “Raskolnikov doesn’t exist” are true. I offer a general argument to this effect and then consider, but reject, some of the accounts of fictional negative existentials offered by abstract object theorists. I then note that some of the sort of data invoked by the abstract object theorist in fact cuts against her position. I concludle that we should not regard fictional characters as abstract objects but rather should adopt a make-believe theoretic account of fictional characters along the lines of those developed by Ken Walton and others.

There is a traditional conception of sensory experience on which the experiences one has looking at, say, a cat could be had by sorneone rnerely hallucinating a cat. Disjunctivists take issue with this conception on the grounds that it does not enable us to understand how perceptual knowledge is possible. In particular, they think, it does not explain how it can be that experiences gained in perceptionenable us to be in ‘cognitive contact’ with objects and facts. I develop this challenge to the traditional conception and then show that it is possible to accornrnodate an adequate account of cognitive contact in keeping with the traditional conception. One upshot of the discussion is that experiences do not bear the explanatory burden placed upon thern by disjunctivists.

This paper explores the relation between understanding and forgiving. A number of people have argued against the old adage that to understand is to forgive, for in many instances understanding leads to excusing rather than forgiving. Nonetheless, there is an interesting connection to be found between forgiving and understanding. I identify three ways in which understanding can lead to forgiveness ofunexcused wrongdoing: It can do so by changing our interpretation of the actor, by changing our interpretation of the action, and by engaging self-Iove.

An influential line of thought in metaphysics holds that where common sense discerns a tree or a dog or a baseball there may be just many microparticles. Provided the microparticles are arranged in the right way -- are “treewise” or “dogwise” or “baseballwise” arranged -- our sensory experiences will be just the same as if a tree or dog or baseball were really there. Therefore whether there really are suchfamiliar objects in the world can be decided only by determining what more is needed for microparticles dogwise arranged actually to compose a dog. This paper argues that this line of thought sets up the wrong agenda. Composition is trivial; dogwise (etc.) arrangement is tricky. Dogwise arrangement will obtain in the wrong regions unless we stipulate that there are dogs, and that dogwise arrangementobtains only within their borders. The bearers of dogwise arrangement, moreover, will have to be dogs themselves, not their microparticles. Thus allowing that dogwise arrangement obtains at all is allowing that there are dogs.