Both consequentialist and non-consequentialist ethical reasoning have difficulties in accounting for the value of consequences. Taken neat, consequentialism is too fierce in its emphasis on success and disregard of luck, while non-consequentialism seemingly over-values inner states and undervalues actual results. In Uneasy Virtue Julia Driver proposes a form of objective consequentialism which claims that characters are good if they typically (but not invariably) produce good results. This position addresses the problems moral luck raises for consequentialism, but requires some form of realism about traits of character. However, if our knowledge of mental states is ascriptive, this form of objective consequentialism may make excessive demands. Non-consequentialists may gain in so far as the theories of action to which they are typically committed are less demanding, and are built to take account of the typical or systematic connections between states of character and results of action.

The following symposium was featured at the seventy-seventh meeting of the Pacific Division of the American Philosophical Association on the occasion of J. B. Schneewind's retirement from the Chair of the Board of Officers. It was my honour to serve as the chair of this session on 28 March 2003.

Utilitarianism continues to vex its critics even in the absence of generally respected arguments in its favour. I suggest that utilitarianism survives largely because of its welfarism. This explains why it survives without the backing of respected arguments. It survives without such arguments because justifying the value of welfare requires no such argument.

This article is concerned with the shape of the story of seventeenth- and eighteenth-century moral philosophy as told by J. B. Schneewind in The Invention of Autonomy. After discussion of alternative possible shapes for such a story, the focus falls on the question to what extent, in Schneewind's account, strands of empiricist voluntarism and rationalist intellectualism are interwoven in Kant. This in turn leads to consideration of different types of voluntarism and their roles in early modern ethical theory.

In this article we argue that the concept of need is as vital for moral theory as it is for moral life. In II we analyse need and its normativity in public and private moral practice. In III we describe simple cases which exemplify the moral demandingness of needs, and argue that the significance of simple cases for moral theory is obscured by the emphasis in moral philosophy on unusual cases. In IV we argue that moral theories are inadequate if they cannot describe simple needs-meeting cases. We argue that the elimination or reduction of need to other concepts such as value, duty, virtue or care is unsatisfactory, in which case moral theories that make those concepts fundamental will have to be revised. In conclusion, we suggest that if moral theories cannot be revised to accommodate needs, they may have to be replaced with a fully needs-based theory.

What outcomes are good, and what there is reason for one to do, is not generally determined by what one thinks or even what one has reason to think. But is a similarly ‘externalist’ account of the distinctively moral concepts, the concepts of moral duty or obligation, of moral wrongness, blameworthiness and guilt, appropriate? I argue not; and on that basis I suggest that an externalist account is not appropriate for the concept of a virtue either.

In this article I argue against Ronald Dworkin's rejection of the labour auction in his ‘Equality of Resources’. I criticize Dworkin's claims that the talented would envy the untalented in such an auction, and that the talented in particular would be enslaved by it. I identify some ways in which the talent auction is underdescribed and I compare the results for the condition of the talented of different further descriptions of it. I conclude that Dworkin's deviation from the ‘envy test’ criterion results in an inequality between the talented and the untalented which cannot be justified in egalitarian terms.

In The Invention of Autonomy, Schneewind argues that a main development in early modern ethical thought is the transition from a conception of morality as obedience to a conception of morality as self-governance. I consider the presuppositions implicit in the latter conception and ask whether they can be maintained.

Julia Driver's Uneasy Virtue offers a theory of virtue and the virtues without being an instance of virtue ethics. It presents a consequentialist challenge to recent virtue ethics, but its positive views – and especially its interesting examples – have great significance in their own right. Driver's defence of ‘virtues of ignorance’ has force despite all the challenges to it that have been mounted over the years. But there are also examples differing from those Driver has mentioned that favour the idea of such virtues. Perhaps certain virtues of religious faith and the virtue necessary for dealing as best one can with moral dilemmas both require ignorance. However, some of the examples Driver does discuss raise the question whether virtue status is based solely on consequences, rather than perhaps having (in addition) a motivational component.

This essay is a rejoinder to comments on Uneasy Virtue made by Onora O'Neill, John Skorupski, and Michael Slote in this issue. In Uneasy Virtue I presented criticisms of traditional virtue theory. I also presented an alternative – a consequentialist account of virtue, one which is a form of ‘pure evaluational externalism’. This type of theory holds that the moral quality of character traits is determined by factors external to agency (e.g. consequences). All three commentators took exception to this account. Therefore, the bulk of my response focuses on defending the externalist account of virtue presented in the final chapters of Uneasy Virtue.

J. B. Schneewind's Sidgwick's Ethics and Victorian Moral Philosophy was the single best philosophical commentary on Henry Sidgwick's Methods of Ethics produced in the twentieth century. Although Schneewind was primarily concerned to read Sidgwick's ethical theory in its historical context, as reflecting the controversies generated by such figures as J. S. Mill, F. D. Maurice, and William Whewell, his reading also ended up being highly neo-Kantian, reflecting various Rawlsian priorities. As valuable as such an interpretation of Sidgwick surely is, Schneewind's approach has always been in some key respects too narrowly conceived in its construction of Sidgwick's philosophical and cultural context, failing to grapple with such troubling, philosophically relevant issues as the possible racism of Sidgwick's ethical and political views, or the sexual politics manifest in his collaboration with such figures as John Addington Symonds.

‘The Second Mistake’ (TSM) is to think that if an act is right or wrong because of its effects, the only relevant effects are the effects of this particular act. This is not (as some think) a truism, since ‘the effects of this particular act’ and ‘its effects’ need not co-refer. Derek Parfit's rejection of TSM is based mainly on intuitions concerning sets of acts that over-determine certain harms. In these cases, each act belongs to the relevant set in virtue of a causal relation (other than marginal contribution) to a specific harmful event. This feature may make an act wrong, in a fashion consequentialists could admit. That explication of TSM does not rely on the questionable assumption that the set of acts is what harms here. Independently of this, there are several other reasons to prefer it to the ‘mere participation’ approach.

J. B. Schneewind's Sidgwick's Ethics and Victorian Moral Philosophy surpassed all previous treatments of Sidgwick's The Methods of Ethics by showing how Sidgwick's work follows a coherent plan of argument for a conception of ethics as grounded in practical reason. Schneewind offered his interpretation as the product of a historical rather than a critical study. This article undertakes a critical study of Sidgwick's work based on Schneewind's interpretation. Its thesis is that the conception of ethics for which Sidgwick argued is incoherent. As a result, it is argued, the coherent plan of argument in the Methods that Schneewind disclosed masks a deep incoherence in the argument itself.

The aim of this article is twofold. First, it is argued that while the principle of ‘ought implies can’ is certainly plausible in some form, it is tempting to misconstrue it, and that this has happened in the way it has been taken up in some of the current literature. Second, Kant's understanding of the principle is considered. Here it is argued that these problematic conceptions put the principle to work in a way that Kant does not, so that there is an important divergence here which can easily be overlooked.

James Lenman argues that consequentialism fails as a moral theory because it is impossible to predict the long-term consequences of our actions. I agree that it is impossible to predict the long-term consequences of actions, but argue that this does not count as a strike against consequentialism. I focus on the principle of indifference, which tells us to treat unforeseeable consequences as cancelling each other out, and hence value-neutral. I argue that though we cannot defend this principle independently, we cannot do without it in practical rationality. Thus abandoning the principle of indifference would involve abandoning all of rationality, not just consequentialist reasoning. I suggest that we should understand the principle as P. F. Strawson understands inductive reasoning – as being part of rationality.

Adams's suggestion that there must be one really right way of presenting the history of moral philosophy seems implausible to me, so I reject – with additional reasons – his charges against the structure of Invention of Autonomy. Skorupski's way of stating the ‘equal moral abilities’ thesis is not, I argue, very Kantian; a more Kantian version is not open to his objections. I am unconvinced by Schultz's claim that Sidgwick did not really hold that thesis. Deigh raises questions I cannot reply to here, but I do offer some texts that seem to show that Sidgwick is not guilty of the confusion about ‘ought’ with which Deigh charges him.

This article considers the reach of arguments for saving the greater number without interpersonal aggregation, and argues that interpersonal aggregation is useful to encompass the proper respect due to each separate person. I first give a precise definition of interpersonal aggregation, which many non-utilitarians try to avoid. Then, I show that consequentialism and Scanlon can justify the case for the greater number without interpersonal aggregation. However, I propose the Aggregation Approach, which justifies the case for the greater number in some cases and the case for tossing a fair coin in other cases. I conclude that interpersonal aggregation does not disrespect the separate person.

This article seeks to diagnose a serious defect in a highly influential supposed counterexample to utilitarianism: Bernard Williams's case of Jim and the Indians. Discussing this, Williams argues that, according to utilitarianism, it is obviously right to say that Jim should kill an Indian. But as this is not obviously right, Williams takes the example to furnish a forceful counterexample to utilitarianism. I note here that the force of the supposed counterexample is in fact very doubtful as the utilitarian can readily enough explain the non-obviousness of the claim that Jim should kill with reference to the non-obviousness of utilitarianism itself.

Many egalitarians believe that social inequalities are worse than natural ones. Others deny that one can coherently distinguish between them. I argue that although one can separate the influence of these factors by an analysis of variance, the distinction is morally irrelevant. It might be alleged that my argument in favour of moral irrelevance attacks a straw man. While I think this allegation is incorrect, I accommodate it by distinguishing between four claims that are related to, and sometimes confused with, the claim that social inequalities are worse. These are: that one has a stronger reason to eliminate inequalities that obtain between, or are produced by, members of one's own society; that inequalities that result from unfair treatment are worse; that inequalities that we make, rather than merely allowing to exist, are worse; and that it is bad if people are treated unfairly.

I will argue that Mill used the concepts tendency and intention as technical terms a proper understanding of which is vital in interpreting his moral philosophy. I examine two interpretations of tendency, offered by Brian Cupples and Fred Berger, and proceed to show weaknesses in both. I will also sketch an interpretation of my own in which tendencies have an important place in Mill's understanding of not only science but moral philosophy as well. I will then show how my interpretation of tendency can help in understanding Mill's foreseeable consequences utilitarianism that relies heavily on Mill's unorthodox use of the notion of intention.