This essay explores the obligations that may arise from benevolently intended interventions that go awry. The author argues that even when the intervening agent has acted with good intentions and in a non-negligent manner, she may be required to continue aid in cases where her initial intervention failed. This is surprising because it means that persons who perform supererogatory acts run the risk of incurring additional heavy obligations through no fault of their own. The author also considers deflationary accounts that run counter to her own analysis.

Evidentialism implies that, for epistemic purposes, belief should be responsive only to evidence. Focusing on our reactive attitude such as resentment or indignation, I construct an argument that the beliefs or judgments accompanying those attitudes are constrained in advance by circumstances to be full, rather than being open to the whole range of partial beliefs. These judgments or beliefs imply strong claims to justification. But the circumstances in which those attitudes are formed allow only very limited evidence. Nevertheless, we cannot opt out regularly since the formation of such attitudes is so central a feature of a minimally content human social life.

Hume’s moral philosophy makes sentiment essential to moral judgment. But there is more individual consistency and interpersonal agreement in moral judgment than in private emotional reactions. Hume accounts for this by saying that our moral judgments do not manifest our approval or disapproval of character traits and persons “only as they appear from [our] peculiar point of view ... ” Rather, “we fix on some steady and general points of view; and always, in our thoughts, place ourselves in them, whatever may be our present situation” (T 581-82), in order to “correct” our situated sentiments. This seems to create two serious difficulties for Hume’s theory. First, moral evaluations become inductive, empirical beliefs about what we would feel if we really occupied the imagined common point of view, and hence are the deliverances of causal reason; this contradicts Hume’s claim that the making of a moral evaluation is not an activity of reason but of sentiment. Secondly, given Hume’s thesis that the passions do not represent anything else, he cannot say that our moral evaluations will better represent the object being judged if they are made from the common point of view. This leaves no clear reason to adopt it, rather than making judgments from our real position. Hume says that left to our particular points of view, we will encounter contradictions and be unable to communicate, but it is hard to see why.My interpretation resolves these two difficulties. I argue that every time we reflect upon someone’s character from the common point of view, we feel an actual sentiment of approbation or disapprobation, which may alter and merge with the situated sentiment or may fail to do so, leaving two different feelings about the same character. Furthermore, whenever we make moral evaluations we also simultaneously make objective, causal judgments about the love and hatred, pride and humility that the trait will produce. We routinely take up the common point of view in order to achieve truth and consistency in our causal judgments, to avoid grave practical problems.

Externalism holds, and internalism denies, that the individuation of many of an individual’s mental states (e.g., thoughts about the physical world) depends necessarily on relations that individual bears to the physical and/or social environment. Many philosophers, externalists and internalists alike, believe that introspection yields knowledge of the contents of our thoughts that is direct and authoritative. It is not obvious, however, that the metaphysical claims of externalism are compatible with this epistemological thesis. Some (e.g., Burge, 1988; Falvey and Owens (F&O), 1994) have sought to dispel the worry that there is a conflict, though they admit that if such a contlict exists, it spells trouble for externalism (see, e.g., F&O, 1994, p. 108). Boghossian has argued that there is indeed a conflict between externalism and introspective knowledge of content. Surprisingly, however, he also argues that there is a conflict between internalism and introspective knowledge of content. I will defend Boghossian’s claim that there is a conflict between externalism and knowledge of content, but criticize his claim that there is a conflict between internalism and knowledge of content.

Recent Descartes scholarship has revised the traditional view of the Cartesian project as one of strictly deductive rationalism. This revision has particularly stressed the role of science in Descartes’ thought. The revisionist conception of Descartes also downplays the significance of the sceptical arguments offered in the First Meditation, seeing them as tools for ‘turning the mind away from the senses’ in the interest of Cartesian science, rather than as reflecting genuinely epistemological concerns. This paper takes issue with this aspect of the revisionist reading of Descartes. It argues that seeing scepticism as critically important for Descartes is independent of interpreting him as a canonical rationalist. In fact, it is precisely Descartes’ scientific thought and practice which make scepticism such a problem for him.

The thesis that practical and epistemic justification can diverge---that it can be reasonable to believe something, all things considered, even when believing is epistemically unjustified, and the reverse---is widely accepted. I argue that this acceptance is unfounded. I show, first, that examples of the sort typically cited as straightforwardly illustrative of the “divergence thesis” do not, in fact, support it. The view to the contrary derives from conflating the assessment of acts which cause one to believe with the assessment of believing itself. I argue, too, that the divergence thesis cannot be rescued by appeal to the possibility of doxastic voluntarism. Finally, I argue that the general acceptance of the divergence thesis rests on a conception of justification, both practical and epistemic, which is seriously flawed.

This article argues, first, that the fundamental structure of the skeptical argument in Kripke’s book on Wittgenstein has been seriously misunderstood by recent commentators. Although it focuses particularly on recent commentary by John McDowell, it emphasizes that the basic misunderstandings are widely shared by other commentators. In particular, it argues that, properly construed, Kripke offers a fully coherent reading of PI #201 and related passages. This is commonly denied, and given as a reason for rejecting Kripke’s reading of Wittgenstein’s text. Second, it is pretty universally accepted that Kripke’s Wittgenstein is a ‘non-factualist’ about ascriptions of meaning. The article argues that, when Kripke’s discussion is rightly understood and the content of ‘non-factualism’ is clarified, there is an important sense in which the skeptical solution is not committed to non-factualism.