MORAL PHILOSOPHY is often written as though the history of the subject were only of secondary and incidental importance. This attitude seems to be the outcome of a belief that moral concepts can be examined and understood apart from their history. Some philosophers have even written as if moral concepts were a timeless, limited, unchanging, determinate species of concept, necessarily having the same features throughout their history, so that there is a part of language waiting to be philosophically investigated which deserves the title “the language of morals” (with a definite article and a singular noun). In a less sophisticated way, historians of morals are all too apt to allow that moral practices and the content of moral judgments may vary from society to society and from person to person, but at the same time these historians have subtly assimilated different moral concepts-and so they end up by suggesting that although what is held to be right or good is not always the same, roughly the same concepts of right and good are universal.

In fact, of course, moral concepts change as social life changes. I deliberately do not write “because social life changes,” for this might suggest that social life is one thing, morality another, and that there is merely an external, contingent causal relationship between them. This is obviously false. Moral concepts are embodied in and are partially constitutive of forms of social life. One key way in which we may identify one form of social life as distinct from another is by identifying differences in moral concepts. So it is an elementary commonplace to point out that there is no precise English equivalent for the Greek word δικαιοσύνη, usually translated justice. And this is not a mere linguistic defect, so that what Greek achieves by a single word English needs a periphrasis to achieve. It is rather that the occurrence of certain concepts in ancient Greek discourse and of others in modern English marks a difference between two forms of social life. To understand a concept, to grasp the meaning of the words which express it, is always at least to learn what the rules are which govern the use of such words and so to grasp the role of the concept in language and social life. This in itself would suggest strongly that different forms of social life will provide different roles for concepts to play. Or at least for some concepts this seems likely to be the case. There certainly are concepts which are unchanging over long periods, and which must be unchanging for one of two reasons. Either they are highly specialized concepts belonging within stable and continuing disciplines, such as geometry; or else they are highly general concepts necessary to any language of any complexity. I have in mind here the family of concepts expressed by such words as and, or, and if. But moral concepts do not fall into either of these two classes.

So it would be a fatal mistake to write as if, in the history of moral philosophy, there had been one single task of analyzing the concept of, for example, justice, to the performance of which Plato, Hobbes, and Bentham all set themselves, and for their achievement at which they can be awarded higher or lower marks. It does not of course follow, and it is in fact untrue, that what Plato says about δικαιοσύνη and what Hobbes or Bentham says about justice are totally irrelevant to one another. There are continuities as well as breaks in the history of moral concepts. Just here lies the complexity of the history.

The complexity is increased because philosophical inquiry itself plays a part in changing moral concepts. It is not that we have first a straightforward history of moral concepts and then a separate and secondary history of philosophical comment. For to analyze a concept philosophically may often be to assist in its transformation by suggesting that it needs revision, or that it is discredited in some way, or that it has a certain kind of prestige. Philosophy leaves everything as it is-except concepts. And since to possess a concept involves behaving or being able to behave in certain ways in certain circumstances, to alter concepts, whether by modifying existing concepts or by making new concepts available or by destroying old ones, is to alter behavior. So the Athenians who condemned Socrates to death, the English parliament which condemned Hobbes’ Leviathan in 1666, and the Nazis who burned philosophical books were correct at least in their apprehension that philosophy can be subversive of established ways of behaving. Understanding the world of morality and changing it are far from incompatible tasks. The moral concepts which are objects for analysis to the philosophers of one age may sometimes be what they are partly because of the discussions by philosophers of a previous age.

A history which takes this point seriously, which is concerned with the role of philosophy in relation to actual conduct, cannot be philosophically neutral. For it cannot but be at odds with the view of all those recent philosophers who have wanted sharply to distinguish philosophical ethics as a second-order activity of comment from the first-order discourse which is part of the conduct of life, where moral utterances themselves are in place. In drawing this distinction such philosophers have tried so to define the realm of philosophy that it would be a conceptual truth that philosophy could not impinge upon practice. A. J. Ayer, for instance, has written about one particular ethical theory that it “. . is entirely on the level of analysis; it is an attempt to show what people are doing when they make moral judgments; it is not a set of suggestions as to what moral judgments they are to make. And this is true of all moral philosophy as I understand it. All moral theories . in so far as they are philosophical theories, are neutral as regards actual conduct.”1

My quarrel with this view will emerge from time to time in these essays. But what I hope will emerge even more clearly is the function of history in relation to conceptual analysis, for it is here that Santayana’s epigram that he who is ignorant of the history of philosophy is doomed to repeat it finds its point. It is all too easy for philosophical analysis, divorced from historical inquiry, to insulate itself from correction. In ethics it can happen in the following way. A certain unsystematically selected class of moral concepts and judgments is made the subject of attention. From the study of these it is concluded that specifically moral discourse possesses certain characteristics. When counterexamples are adduced to show that this is not always so, these counterexamples are dismissed as irrelevant, because not examples of moral discourse; and they are shown to be nonmoral by exhibiting their lack of the necessary characteristics. From this kind of circularity we can be saved only by an adequate historical view of the varieties of moral and evaluative discourse. This is why it would be dangerous, and not just pointless, to begin these studies with a definition which would carefully delimit the field of inquiry. We cannot, of course, completely avoid viewing past moralists and past philosophers in terms of present distinctions. To set out to write the history of moral philosophy at all involves us in selecting from the past what falls under the heading of moral philosophy as we now conceive it. But it is important that we should, as far as it is possible, allow the history of philosophy to break down our present-day preconceptions, so that our too narrow views of what can and cannot be thought, said, and done are discarded in face of the record of what has been thought, said, and done. We have to steer between the danger of a dead antiquarianism, which enjoys the illusion that we can approach the past without preconceptions, and that other danger, so apparent in such philosophical historians as Aristotle and Hegel, of believing that the whole point of the past was that it should culminate with us. History is neither a prison nor a museum, nor is it a set of materials for self-congratulation.