We speak of ‘obligations’ in many contexts. But what are obligations? I argue that obligations of an important type inhere in what I call ‘joint commitments’. I propose a joint commitment account of everyday agreements. This could explain why some philosophers believe that we know of the obligating nature of agreements a priori. I compare and contrast obligations of joint commitment with obligations in the relatively narrow sense recommended by H. L. A. Hart, a recommendation that has been influential. Some central contexts in which Hart takes there to be obligations in his sense are contexts in which there are obligations of joint commitment. Nonetheless different senses of ‘obligation’ appear to be at issue.

The idea that immoral behaviour can sometimes be admirable, and that moral behaviour can sometimes be less than admirable, has led several of its supporters to infer that moral considerations are not always overriding, contrary to what has been traditionally maintained. In this paper I shall challenge this inference. My purpose in doing so is to expose and acknowledge something that has been inadequately appreciated, namely, the moral aspect of nonmoral goods and evils. I hope thereby to show that, even if immorality can be admirable (and morality less than admirable), this poses no threat to morality.

Lately there has been a renewal of interest among political philosophers and theorists in the idea of cosmopolitanism. However, there is little consensus among contemporary theorists about the precise content of a cosmopolitan position. This article calls attention to two different strands in recent thinking about cosmopolitanism. One strand presents it primarily as a doctrine about justice. The other presents it primarily as a doctrine about culture and the self. Although both forms of cosmopolitanism have some appeal, each is sometimes interpreted in ways that render it untenable. This article attempts to distinguish between the more and the less plausible versions of each form of cosmopolitanism. In each case, the distinction turns on how the normative status of particular interpersonal relationships and group affiliations is understood.

Smith has defended the rationalist's conceptual claim that moral requirements are categorical requirements of reason, arguing that no status short of this would make sense of our taking these requirements as seriously as we do. Against this I argue that Smith has failed to show either that our moral commitments would be undermined by possessing only an internal, contextual justification or that they need presuppose any expectation that rational agents must converge on their acceptance. His claim that this rationalistic understanding of metaethics is required for the intelligibility of moral disagreement is also found to be inadequately supported. It is further proposed that the rationalist's substantive claims – that there are such categorical requirements of reason and that our actual moral commitments are a case in point – are liable to disappointment; and that the conceptual claim is fatally undermined by reflection on how we might best respond to such disappointment.

The distinction between killing and letting die is too simple. A third category – preventing people from being saved – must also be recognized. Like killing, preventing a person from being saved is a species of doing harm; like killing, it infringes one of the victim's negative rights. Yet preventing a person from being saved is morally on a par with letting die, which infringes one of the victim's positive rights. It follows that we cannot explain the moral inequivalence of killing and letting die by saying, as so many have, that negative rights are more stringent than positive rights. A more promising strategy is suggested at the end of the article.

This paper discusses what the so-called Priority View implies regarding possible people. It is argued that this view is plausible when applied to fixed populations, but that, when applied to the issue of possible people, it faces certain difficulties. If it is claimed that possible people fall within the scope of the Priority View, we are led to the repugnant conclusion (and other counter-intuitive conclusions) at a faster pace than we are by, e.g., utilitarianism. And if it is claimed that possible people do not fall within the scope of the Priority View, we shall have to combine this view with a different view in order to avoid certain counter-intuitive judgements. Such a combined view, however, leads to intransitivities. At the end of the paper, I discuss what these conclusions imply regarding both the Priority View and our moral obligations to possible people.

Moral pluralism of the kind associated with W. D. Ross is the doctrine that there is a plurality of moral principles, which in their application to particular cases can conflict, and that there is no further principle to determine which of these principles takes priority in cases of conflict. Two objections are commonly advanced against this kind of pluralism: that it proposes a rag-bag of moral principles lacking a unifying basis; and that it offers no way to adjudicate moral disputes when our intuitions about what to do conflict. The present paper replies to both of these objections, in particular by responding to versions of them advanced by Brad Hooker. The tying together and justification of different moral principles may be achieved by a general rational justification procedure, rather than by a further moral principle; and such a rational justification procedure can help to adjudicate moral disputes.

James Lenman is critical of my claim that moral requirements are requirements of reason. I argue that his criticisms miss their target. More importantly, I argue that the anti-rationalism that informs Lenman's criticisms is itself implausible.

The relationship between Bentham's ‘enunciative principle’ (that each person's sole ultimate motive is the maximization of their own happiness) and his ‘censorial principle’ (that it is the effects on the happiness of all affected which determines what they ought to do) is famously problematic. The problem's solution is that each person has an overwhelming interest in living in a community in which they, like others, are liable to punishment for behaviour condemned by the censorial principle (and in some cases rewarded for behaviour which it favours) either by the institutions of the state or by the tribunal of public opinion. The senses in which Bentham did and did not think everyone selfish are examined, and a less problematic form of psychological hedonism than Bentham's is proposed.

It is impermissible to violate a constraint, even if by doing so a greater number of violations of the very same constraint were to be prevented. Most find this puzzling. But what makes the impermissibility of such minimizing violations puzzling? This article discusses some recent answers (by Scheffier, Kamm and Nagel) to this question. The article's first aim is to make clear in what way these answers differ. The second aim is to evaluate the answers, along with Kamm's and Nagel's proposed solutions of what they see as the puzzle of constraints. The main thesis of the article is this: because defenders of constraints are not committed to any conception of valuable states of affairs, constraints do not conflict with maximizing rationality; but neither can they be accounted for in terms of impersonal values.

Bentham and J. S. Mill can be regarded as utilitarian tax-reformers distinguished from political economists who were simply averse to taxation. The purpose of this paper is to elucidate the difference between Bentham's and Mill's tax reform programmes. Bentham proposed the law of escheat and a tax on bankers' and stock dealers' profits, subject to the principle of least sacrifice of enjoyment. He also planned to correct the inequality of the land tax by extending it into a general income tax. Mill proposed an income tax on the basis of the principle of equal sacrifice of enjoyment. He also proposed a progressive inheritance tax and a variable land tax, regarding unearned income as a fit subject for a special tax. Consequently, Mill used Bentham's ideas and tools to take a step towards a more egalitarian programme.

In this paper I critically evaluate Ronald Dworkin's attempt in Life's Dominion to understand sacred value as a form of intrinsic value which is grounded in investment. I argue that there are two problems with Dworkin's conception of intrinsic value. First, it does not allow him to distinguish, as he must, between incremental and sacred values. Secondly, sacred value qua intrinsic value is not the kind of value which can be grounded in investment. I argue that both of these problems are an implication of a realist conception of intrinsic value. Nevertheless, investment does seem to play a role in our judgements about intrinsic value and in light of this I aim to develop an alternative account of intrinsic value whichcan make sense of this.

I distinguish two different arguments against cognitivism in Bernard Williams's writings on moral dilemmas. The first turns on there being a truth of the matter about what we ought to do in a moral dilemma. That argument can be met by appealing to our epistemic shortcomings and to pro tanto obligations. However, those responses make no headway with the second argument, which concerns the rationality of the moral regret that we feel in dilemma situations. I show how the rationality of moral regret can be explained on an ‘independent desire’ model. And I show how Williams's second argument only appears to have force because of a certain faulty way of conceiving the issue over cognitivism. But Williams's argument rightly alerts us to the rational role of desire in our moral thought.

Eldon Soifer and Béla Szabados argue that hypocrisy poses a problem for consequentialism because the hypocrite, in pretending to live up to a norm he or she does not really accept, acts in ways that have good results. They argue, however, that consequentialists can meet this challenge and show the wrongness of hypocrisy by adopting a desirefulfilment version of their theory. This essay raises some doubts about Soifer and Szabados's proposal and argues that consequentialism has no difficulty coming to grips with hypocrisy, whether or not one favours a desire-fulfilment account of the good.

In Life's Dominion Ronald Dworkin argues that disagreement over the morality ofabortion is about how best to respect the intrinsic value of human life, rather than about foetal rights as many people mistakenly suppose. Dworkin argues that the state should be neutral indebates about intrinsic value and thus it should be neutral in the abortion debate. Through a consideration of the notion of intrinsic value, it is argued in this article that Dworkin'sargument fails. On the interpretation of ‘intrinsic value’ which Dworkin seems to favour, it is shown to be implausible that such a notion accounts for different views about the value of human life. On an alternative interpretation of ‘intrinsic value’ it is argued that the state is not usually neutral on such matters, and thus there is no reason why it should be in the case of abortion.

Frances Howard-Snyder has argued that objective consequentialism violates the principle that ‘ought’ implies ‘can’. In most situations, she claims, we cannot produce the best consequences available, although objective consequentialism says that we ought to do so. Here I try to show that Howard-Snyder's argument is unsound. The claim that we typically cannot produce the best consequences available is doubtful. And even if there is a sense of ‘producing the best consequences’ in which we cannot do so, objective consequentialism does not entail that we ought, in this sense, to produce the best consequences.

Shelly Kagan and Leonard Katz have offered versions of hedonism that aspire to occupy a middle position between the view that pleasure is a unitary sensation and the view that pleasure is, as Sidgwick put it, desirable consciousness. Thus they hope to offer a hedonistic account of well-being that does not mistakenly suppose that pleasure is a special kind of tingle, yet to offer a sharp alternative to desire-based accounts. I argue that they have not identified a coherent middle position.

Geoffrey Scarre has recently argued that the version of qualitative hedonism which I attribute to Mill is unsatisfactory for various reasons. In his view, even if it is formally compatible with value monism, ‘the Mill/Riley line’ involves non-hedonistic elements and offers an implausible account of the relationship between ‘higher’ and ‘lower’ pleasures. In this paper, I show that his objections, which are similar in spirit to those pressed earlier by Bradley, Moore and others against Mill, are unfounded where not confused. The Mill/Riley line does not rely on non-hedonistic standards and has sufficient flexibility to account for many different kinds of pleasures and pleasing activities. It remains a coherent version of qualitative hedonism, worthy of further consideration and study.

Frances Howard-Snyder argues that objective consequentialism should be rejected because it violates the principle of ‘ought’ implies ‘can’ in asking us to do what we cannot. In this comment I suggest that Howard-Snyder does not take sufficiently seriously the chief defence of objective consequentialism, which reformulates it so that it applies only to actions we can perform. Nonetheless, I argue that there are arguments relating to ‘ought’ implies ‘can’ which discredit objective consequentialism even if it is thus reformulated. These arguments also cause problems for a reformulated version of subjective consequentialism.