Moore is taken to have followed Sidgwick in his arguments against naturalism and in his consequentialism. I argue that there are differences on both issues. Sidgwick's arguments against naturalism do not rely on a controversial view of analysis, and one of his arguments for consequentialism gives him greater resources against critics of consequentialism such as T. M. Scanlon.

In many recent discussions of the morality of actions where both good and bad consequences foreseeably ensue, the moral significance of the distinction between intended and foreseen consequences is rejected. This distinction is thought to bear on the moral status of actions by those who support the Doctrine of Double Effect. According to this doctrine, roughly speaking, to perform an action intending to bring about a particular bad effect as a means to some commensurate good end is impermissible, while performing (...) an action where one intends only this good end and merely foresees the bad as an unintended sideeffect may be permissible. Consequentialists argue that this is a distinction which makes no moral difference to the evaluation of the initial act in the two cases, given that the overall consequences are the same in each case. In this paper we aim to show that a standard consequentialist line of argument against the moral relevance of the intention/foresight distinction fails. Consequentialists commonly reject the moral relevance of this distinction on the grounds that there is no asymmetry in moral responsibility between intending and foreseeing evil. We argue that even if this claim about moral responsibility is correct, it does not entail, as many Consequentialists believe, that there is no moral asymmetry between acts of intended and foreseen evil. We go on to argue that those consequentialists who do concede the moral relevance of the intention/foresight distinction at the level of agent evaluations cannot consistently make such a concession, and that such a position is in any case untenable, because it entails a complete severance of important conceptual connections between act and agent evaluations. (shrink)

Something can be said to be good for a particular person, whether or not it is good for anyone else, let alone good ‘overall’ or ‘good simpliciter ’. Sometimes we speak of ‘John's good’ as well as of things that are ‘good for John’. What is ‘good for John’ is whatever enhances his ‘good’ or, to use an apparently synonymous term, his ‘well-being’. But what is a person's well-being: in what does it consist?

The essay argues that while there is no general agreement on whether moral realism is true, there is general agreement on at least some of the moral obligations that we have if moral realism is true. Given that moral realism might be true, and given that we know some of the things we ought to do if it is true, we have a reason to do those things. Furthermore, this reason is itself an objective moral reason. Thus, if moral realism (...) might be true, then it is true. (shrink)

In Recent years noncognitivist ethical theories have been supported by an argument which has come to be widely accepted among moral philosophers.1 According to this argument, an ethical term like ‘good’ has both a commending function and a describing function, but between these functions there is the important difference that the commending function alone is invariant while the describing function varies greatly. For many and different things may be called good—hammers, sunsets, paintings, missionaries, cannibals—but despite these differences in the descriptive (...) criteria for applying the word to objects, ‘good’ retains a common meaning in all these uses, for in each case the word is being used to commend. The conclusion drawn, then, is that because of its being the sole invariant or common feature of every use of ‘good’, the commendatory function rather than the varying descriptive function must be the primary meaning of ‘good’. I shall refer to this as the commendatory invariance argument. (shrink)