This paper examines Kant’s accounts of friendship and marriage, and argues for what can be called an ideal of “moral marriage” based on Kant’s notion of moral friendship. After explaining why Kant values friendship so highly, it gives an account of the ways in which marriage falls far short, according to Kant, of what friendship has to offer. The paper then argues that many of Kant’s reasons for finding marriage morally impoverished compared with friendship are wrong-headed. The paper further argues that a few of Kant’s views about friendship are false. The main point is that, when we slightly revise Kant’s account of friendship and jettison Kant’s misguided notions about marriage, we see that marriages can aspire to much of the same moral richness as friendships. Finally, the paper argues that this friendship model of marriage does not obscure the important ways in which marriages and friendships differ.

The author presents and defends a general view about belief. and certain attributions of belief, with the intention of providing a solution to Saul Kripke’s puzzle about belief. According to the position developed in the paper, there are two senses in which one could be said to have contradictory beliefs. Just one of these senses threatens the rationality of the believer; but Kripke’s puzzle concerns only the other one. The general solution is then extended to certain variants of Kripke’s original puzzle, which have to do with belief attributions containing empty names and kind terms.

The central problem for a realist about mental causation is to show that mental causation is compatible with the causal completeness of physical systems. This problem has seemed intractable in large part because of a widely held view that any sort of systematic overdetermination of events by their causes is unacceptable. I account for the popularity of this view, but argue that we ought to reject it. In doing so, I show how we thereby undermine the idea that mental causes must be naturalizable in order to be legitimate. Thus I argue that a non-naturalist conception of mental causation is compatible with a plausible kind of physicalism.

Donnellan famously argued that while one can fix the reference of a name with a definite description, one cannot thereby have a de re belief about the named object. All that is generated is meta-linguistic knowledge that the sentence “If there is a unique F, then N is F” is true. Donnellan’s argument and the sceptical position are extremely influential. This article aims to show that Donnellan’s argument is unsound, and that the Millian who embraces Donnellan’s scepticism that the reference-fixer cannot secure the relevant de re belief faces a serious problem: Millianism about names plus scepticism about the reference-fixer’s de re belief conflicts with what seems to be an analytical thesis about the relationship between semantic content and understanding. The upshot is that the Millian has good reason to seek an alternative to scepticism.

Hector-Neri Castaneda claimed in several papers that a proposition expressed by an indexical sentence (in a context of utterance) can be re-expressed by means of an oratio obliqua clause (in a sentential context) that contains a quasi-indicator. Robert M. Adams and Rogers Albritton have presented a counter-argument that is accepted by Castaneda himself. I will argue that the Adams/Albritton argument is not convincing: The argument uses several assumptions which could be disputed. The paper tries to develop a more direct argument against Castaneda’s central claim. If Castaneda’s thesis is false, what then is achieved by quasi-indexicals in oratio obliqua? Adams and Castaneda answer this question with a picture: the quasi-indexical clause portrays an indexical proposition. I use Perry’s idea that quasi-indicators could be seen as expressions that bind special sense variables to give a less metaphorical account of the functioning of quasi-indicators. Finally, I explore the consequences of this account for iterated knowledge-ascriptions with quasi-indicators and for truth-conditional theories of meaning.

Most contemporary internalists are fallibilists, denying that there need be anything about which we are infallible for us to have knowledge or justified beliefs. At the same time, internalists standardly appeal to ‘internal twins’ in arguing against externalism and motivating internalism---a Cartesian demon can ruin the ‘external’ relations we have to the world, but one is equally well justified in one’s beliefs whether or not one is subject to such deception. Even if one doesn’t motivate one’s internalism by appeal to internal twins, any internalist must agree that internal twins are equally well justified in their beliefs. I argue that the internal twins argument for, or commitment of internalism, commits one to the claim that the conditions in virtue of which one is justified must be ones about which a believer is infallible. The basic argument is that for anything about which one can be mistaken, one has an internal twin who is mistaken, but is equally well justified---and so, not in virtue of that about which one can be mistaken. If the argument can be resisted, this should tell us something useful about how to properly understand both internalism in general, and the idea of internal twins in particular.