People engaged in teaching and research in philosophy departments are under strong and constant -pressure to believe that what they are doing makes sense. Their belief may well be rooted in an initial project, formed autonomously under the guidance of teachers and fellow students; even so, it will continue to be shaped and confirmed by the requirements of peer approval and student satisfaction. The hungry sheep keep looking up. Sheer humanity requires us to show enthusiasm for what we dispense as nutrition, and a convincing display of conviction requires that we start by convincing ourselves. But in retirement all this falls apart. Often, perhaps usually, an element of belief and devotion is maintained, because the formed ego of the retiree is that of self-as-teacher or self-as-philosopher. But if, as may well be, the retiree's philosophizing was undertaken merely as gainful employment, things rapidly come to pieces. In these circumstances, a request for a paper on metaethics or any other loosely defined topic may bewilder. What, really, could be at issue here? Or so it seems. But there is something factitious in this bewilderment.

It seems to be generally agreed that a foundationalist view of any area of justified beliefs is the affirmation that there are some (basic) beliefs which are to some degree credible for a person independently of reflection on logical relations to any others of his beliefs, and that any other beliefs of his are justified because of appropriate logical relations to these basic beliefs — thus contrary to the coherentist thesis that beliefs can only be justified by appeal to their relation to other beliefs of a person, independently of whether these other beliefs are themselves independently credible. Thus, for the area of moral beliefs, foundationalism is the view that there is at least a subset of a person's moral beliefs which are either justified independently of logical relations to any other beliefs, or are justified by their logical relations to some other beliefs, either independently credible moral beliefs, or independently credible non-moral ones.

I am here restricting ‘ethical’ beliefs to moral beliefs, as distinct from beliefs that something is a good thing or intrinsically good or good for a person.

Professor Foot's Hart Lecture, now published (1995), is largely devoted to an attack on people she calls ‘subjectivists’ and ‘noncognitivists', among whom she includes myself, although she is so good as to allow me, in a footnote, to reject the names. She seems to imply thereby that this is a mere matter of nomenclature or terminology. But in truth her use of these terms makes one suspect that she has not fully understood either the issues or what I have said about them.

It may therefore be useful to explain yet again why I reject these descriptions. I have done so already in a paper to a conference in Moscow which she heard, and to which she kindly refers (Hare 1993a). These explanations would not be important if she were the only person to be confused about this matter; but the confusions are so widespread, even among professional philosophers who should know better, and are so often taught to succeeding generations of students, that it is worth while to make yet another attempt to clear them up - though the confusions are so insidious that I have not much hope that they will ever be eradicated.

Theorizing is a distinctive activity. Within almost any discipline, for example, one finds a recognized distinction between theorists and nontheorists, theoretical work and non-theoretical work. To be sure, the distinction is often a matter of degree. Typically, theorists are identified by their commitment to developing an understanding of their subject matter that possesses certain characteristic features- they aim for generality, comprehensiveness, coherence, well-foundedness, and explanatoriness. They aim to formulate and justify systematic representational systems.

Theorizing is not to be confused with reflection, which can be more or less lacking in theoretical ambition. Nor is theorizing to be confused with criticism, which can be piecemeal. Theorizing is a form of reflection, and it can be critical, but theories can be vindicatory, even apologetic.

Why are traditional ‘objectivist’ theories of morality, such as those put forward by Aristotle, or Kant, or even Bentham, commonly thought not to pass ‘scientific muster’ insofar as they are not ‘naturalist’? My interest in this question is based on my being a moral objectivist, but answering this question is one that moral skeptics should be as interested in as I. The view that the commitments of science preclude us from accepting such theories is the basis of the moral skeptic's position. Yet showing what is wrong with a moral objectivist position is surprisingly difficult. It involves reflecting on what ‘scientific muster’ is supposed to be, and on why a theory is commonly thought to be disreputable unless it passes it. It also involves locating the ‘queer’ element in objectivist moral theory that makes it scientifically disreputable. Yet, as I hope to show in this article, there is no commonly accepted statement of what makes a theory scientifically acceptable or unacceptable, and (perhaps even more surprisingly) no rigorous account of what the queer component of objectivist moral theory is that makes any such theory scientifically unacceptable.

John McDowell (1979) suggests that virtuous agents have a perceptual sensitivity allowing them to determine reliably what to do in any specific context of deliberation. Moreover, the reliable perception yields an accurate depiction not only of the facts of the situation but of the morally right act to do. Finally, the reliably kind behavior of a kind person

In the first part of this paper I want to consider the relation between two familiar philosophical views that have not to my knowledge been considered in any depth together, even by philosophers who are well known for defending each separately. These views have a certain natural affinity, in that each has been attractive to philosophers of a generally naturalistic bent. Thus, since I intend to argue that there is a difficulty in reconciling them, I will be pointing out a difficulty, not, I think, in philosophical naturalism, but nevertheless in one package often accepted by naturalists. One of the views is that the problem of evil is, at the very least, a serious theoretical difficulty for theism. The other is nihilism about value, the thesis that there are no real values in the world and that statements ascribing values to things are never true. I hope it is obvious why there is at least a prima facie difficulty in combining these views: how can someone who thinks that nothing is really good or evil also think that we find in the world more evil than we would expect if theism were correct?

'Internalism’ in ethics is a cluster of views according to which there is an ‘internal’ connection between moral obligations and either motivations or reasons to act morally; ‘externalism’ says that such connections are contingent. So described, the dispute between internalism and externalism may seem a technical debate of minor interest. However, the issues that motivate it include deep problems about moral truth, realism, normativity, and objectivity. Indeed, I think that some philosophers view externalism as undermining the ‘dignity’ of morality. They might say that if morality needs an ‘external sanction’ - if the belief that one has an obligation is not sufficient motive or reason to do the right thing- then morality is debased in status. Even an arbitrary system of etiquette could attract an external sanction under appropriate conditions.

Although I believe that the more interesting internalist theses are false, there are important truths that internalism is attempting to capture. The most important of these is the fact that moral judgments are intrinsically ‘normative’ or ‘choice-guiding,’ that they are, very roughly, relevant to action or choice because of their content.

Metaethics is the philosophical study of what morality is. It differs from ethical theory, which attempts to systematize (and possibly ground) moral judgments, and also from practical or applied ethics, which reflects on particular moral issues or problems. As it has been done in this century, metaethics has usually involved three interrelated projects: a metaphysical investigation into the nature of moral facts and properties, a semantic inquiry into the meaning of moral assertions, and an epistemological account of the nature of moral knowledge. In all three areas, the questions raised by twentieth-century metaethics have apparently been radical, and the dominant position was even openly nihilistic. In metaphysics it was antirealist, maintaining that there are no moral facts, in epistemology noncognitivist, denying that there is moral knowledge, and in semantics emotivist or prescriptivist, holding that moral assertions aren't assertions at all, but are speech acts utterly devoid of truth conditions.

The Paradox of Postmodernism

Modernism is, roughly speaking, the Enlightenment belief in a single unified rational perspective, founded on some indubitable evidence given in human experience – either innate concepts à la Descartes and the rationalists, or sensations à la Locke and the empiricists – and elaborated according to reliable logical rules. This view was first attacked for its ‘foundationalism.’ Philosophers, such as Nietzsche, Dewey, Heidegger and the later Wittgenstein, denied that there is any indubitable given upon which truth can be founded. There is no experience, no testimony of the senses or of reason that blazons forth the undeniable truth. Rather the ‘given’ is, so to speak, constructed - which is to say, not given to us, but made by us. Some experience or other evidence is interpreted as this or that with this or that epistemological status, on the basis of beliefs that one already has about, say, space or mathematics or sense perception or the nature of what is ultimately real. Postmodernism is an intensification of this attack, with a distinctive political spin.

Most of the essays collected here are essays in metaethics seeking in exacting and interesting ways to resolve problems raised by the familiar options in metaethics we outlined in our Introduction. Richard Brandt, for example, forcefully argues, going much against the at least modestly holistic grain of our time, for a foundationalism (noncognitivist though it be) which would be foundational in both metaethics and normative ethics. R.M. Hare makes a brief but systematic defense, which is both spirited and clear, of his prescriptivism (a species of what we, following tradition, have called ‘noncognitivism,’ but which he argues should instead be called ‘nondescriptivism’). His arguments here for his position - call it nondescriptivism or noncognitivism- are directed forcefully against ethical naturalism (descriptivism) and specifically against the naturalism of Philippa Foot. Nicholas Sturgeon and David Copp contribute elaborate and rigorously argued defenses of ethical naturalism, or, as they might prefer to call it, ‘moral realism.’