There are three elements that are essential to morality: a moral community, a set of moral values, and a moral code. All three are necessary, and none on its own is sufficient.

First, it is as impossible to have a purely private morality as it is to have a purely private language, and for very similar reasons. Secondly, the moral life of the community consists of the shared pursuit of non-material values, such as fairness, truth, comradeship, freedom. It is the nature of the values pursued that distinguishes morality from economics.

Thirdly, this pursuit is carried out within a framework that excludes certain prohibited types of behaviour. It is this that distinguished morality from aesthetics, which is also a pursuit of non-material values. The answer to the question, ‘Who does the prohibiting’, is that it is the members of the moral community: membership of a common moral society involves subscription to a common code.

So let us take a look at these three points in more detail.

According to the once dominant prescriptivism of Richard Hare, the moral legislator was essentially an individual. All that was needed for a moral judgement was that an individual should issue a prescription that could be applied to any other moral agents in similar situations – that was, in Hare's terms, ‘universalizable’. There was no restriction on the content of the prescription and no need for anyone else to concur in the judgement; the only logical requirement was that the totality of the individual's prescriptions should be consistent.

Of course, the prescription was applicable to people other than the prescriber: that was what was meant by universalizability. But it applied hypothetically to other individuals in the same circumstances as the prescriber. Whether anyone was ever in fact in those circumstances was neither here nor there. In this sense, one could say that a morality of Hare's type was a private morality.

I have said that it is as impossible to have a purely private morality as it is to have a purely private language, and for similar reasons. Wittgenstein argued that it was impossible to give meaning to a term by a private ostensive definition because with such a procedure there would be no difference between something's seeming right and its being right. So too with a private morality. An individual's sincere universal prescription cannot be faulted by any other individual, and so again for the prescriber there is no difference between being right and seeming right.

For there to be a morality there has to be a moral community, and moral communities are not put together by individual choice: we are all born into and brought up within moral communities. To use a Kantian metaphor, we are not volunteers but conscripts in the kingdom of ends. But it is by no means an uncontroversial matter to determine the extent of the moral community to which one belongs. Shall we say ‘the whole human race’? There have been many times and places in which the effective moral community was much smaller than this: the inhabitants of a city, the members of a tribe, the citizens of a nation.

The Stoics held that each of us stands at the centre of a series of concentric circles. The first circle surrounding my individual mind contains my body and its needs. The second contains my immediate family, and the third and fourth contain extensions of my family. Then come circles of neighbours, at varying distances, plus the circle that contains all my co-nationals. The outermost and largest circle encompasses the whole human race. If I am virtuous I will try to draw these circles closer together, treating cousins as if they were brothers, and constantly transferring people from outer circles to inner ones. Hierocles, who expounded this account of moral progress, was writing just a few years after Jesus had preached that the neighbour whom I should love as I love myself could belong to a different sect or group or nation.

Since the time of Montaigne, some people have extended the moral community beyond humankind to other sentient beings, claiming that animals have an equal claim with human beings. The principle philosophical source of the idea is the utilitarianism of the cat-loving Jeremy Bentham. But by making the supreme moral criterion a matter of sensation (happiness, he said, was identical with pleasure) he made it appropriate to consider animals as belonging to the same moral community as ourselves since animals as well as humans feel pleasure and pain.

This, in the long term, proved to be one of the most significant consequences of Bentham's break with the classical and Christian moral tradition, which placed supreme moral value in activities not of the sense but of the reason, and regarded non-rational animals as standing outside the moral community. In my view, the Christian tradition was correct: only the kinds of beings who have duties can have rights, and non-human animals have no duties. A good dog has virtues of its own, but they are not moral virtues.

So much for my first point. Secondly, I argue that the moral life of the community consists of the shared pursuit of non-material values such as fairness, truth, comradeship, freedom. It is the nature of the values pursued that distinguishes morality from economics. But the individual values are not pursued in isolation from each other. The pursuit of these values is the same thing as the exercise of the moral virtues, and these virtues form a unity. Several values are pursued as elements of a good life, for the individual and for the community. It is in tying together the individual values that the notion of happiness plays a crucial role in morality.

My third point is that the exercise of the virtues, the pursuit of the non-material goods, was carried out within a framework that excludes certain prohibited types of behaviour. It is this that distinguishes morality from aesthetics, which is also a pursuit of non-material values.

Many people are suspicious of the idea that some classes of action are absolutely prohibited. Where, they ask, do these absolute prohibitions come from? Religious believers see them as coming from God; but they do not seem to be able to agree among themselves which are the actions that God has prohibited absolutely. Even if they did, how could they convince unbelievers of this? On the other hand, can there be an absolute prohibition without a prohibiter?

These prohibitions, I maintain, do not need to emanate from a source external to the community, such as a transcendent Godhead. The moral community, I suggest, creates moral laws in a manner similar to that in which the linguistic community creates the rules of grammar and syntax. Moral rules, like linguistic rules, may change as society changes; but unless a set of such rules is in operation society collapses into anarchy as language, without some form of grammar or syntax, collapses into incoherence.

Even the most devout do not believe that God promulgates linguistic rules in the way that they believe him to promulgate moral commands. In Genesis, God teaches Adam the names of the animals; but even the Bible does not suggest that he taught him grammar and syntax. Even if Chomsky is right that there are innate constraints on the kind of languages there can be, the most that God has written in our hearts is the set of rules of depth grammar. The linguistic laws that we obey are laws that we have legislated for ourselves, by our daily conformity to the conventions that they embody.

The idea that moral rules originate within the moral community itself was picturesquely sketched by Immanuel Kant in his Groundwork of the Metaphysic of Morals. Every human being, he said, is a member of a Kingdom of Ends, a union of rational beings under common laws. My own will is rational only in so far as its maxims – the principles on which it makes its choices – are capable of being made universal laws. The converse of this is that universal law is law that is made by rational wills such as mine.

A rational being is thus subject only to laws which are made by himself and yet are universal. In the Kingdom of Ends, we are all both legislators and subjects. Kant left room for a divine being that was not just a member but a head of the Kingdom of Ends: a legislator not subject to law. No doubt it was pious and prudent for him to do so – but it was not, if I am right, a necessary part of his system. Instead of a Kingdom of Ends, we can make sense of a Republic of Ends, in which the members are citizens rather than subjects. Such a republic is a direct, not a representative, democracy: all are legislators and no one is above the law.

Such, then, is the moral community. It is that community that issues, and enforces, the prohibitions that constrain the activities of its members. Within those constraints, the members of the community pursue the non-material values that constitute the community's moral life. Thus, the three elements that we identified at the beginning of this essay come together to constitute a morality.