Who Can Tell Right from Wrong?

1.

This is a powerful, original, and deeply interesting work and many will find it a disturbing one. For it is calculated to unsettle more radically than has been done before the belief, cherished for so long and by so many, that philosophy can furnish or discover rational foundations for ethical thought, or can at least provide some acceptable systematic theory that could determine how, in thinking ethically, we should think if we are to think correctly.

Williams differs from most of his predecessors and contemporaries in emphasizing the irreducible variety and complexity of ethical thought, and the extent to which it is obscured by fictions and myths and marked by conflict and uncertainty, leaving much to personal decision. These are things often obscured by philosophers in their pursuit of unity and system, and in their ambition to impose on ethical thought an inappropriate structure of discursive rationality and impersonal principle. So in Williams’ view, philosophy has not only done for ethical thought far less than it has often claimed to do, but it has also done much to damage it.

Williams writes with great understanding of the history of moral philosophy and is sensitive to aspects of it not seen or so imaginatively explored by others. He has an enviable grasp of its increasingly complex literature and in the course of this short book he throws a fresh and revealing light on a very wide range of philosophical thought. He does this equally well for philosophers with whom he has most sympathy, and in particular for Aristotle, as for others, like his contemporary R.M. Hare and the nineteenth-century philosopher Henry Sidgwick, whose different defenses of utilitarian ethical theory, dissected here with much acumen and vivacity, seem to him to rest upon and to foster some distinctive aberrations of modern moral philosophy. So in reading this book I had the exhilarating sensation of the mist clearing from many patches in this well-trodden, but still obscure, philosophical terrain.

But this is not an easy book, and I find it difficult to endorse the claim on the jacket of the English edition that it can be followed by readers without philosophical training. It is true that Williams writes without unexplained technicalities and often with clarity and wit, and his considerable philosophical scholarship does not intrude upon his text but appears in twenty pages of valuable discursive notes at the end of the book. Nonetheless much that he writes needs, as well as deserves, to be read more than once. Often this is because the slant of his attention and the insights he offers are novel; sometimes it is because he writes in an extraordinarily condensed, almost epigrammatic, style which leaves important implications to be worked out by the reader, not always (for me at least) with adequate guidance.

But there is also a difficulty of a quite different kind which many may encounter in reading this book. This arises not from any omission on the author’s part, but from his austere conception of the limited powers of philosophy to guide the course of ethical thought. This frequently calls for the suppression on the reader’s part of general questions that may spring to his lips about the direction in which the ethical practices and beliefs to which we find ourselves subscribing are to develop if we abandon, as Williams thinks we should, the construction of ethical theories to guide their course. Moreover, it is not only the invitation to abandon ethical theory that may prompt questions that philosophy, on Williams’s view, should not attempt to answer.

For within the general sphere of the ethical, which is left broadly defined as the range of those considerations which, in contrast with simple egoism, take account in various ways of the needs, desires, and aspirations of others, Williams distinguishes what he terms “the peculiar institution of morality,” which focuses narrowly on a special conception of obligation as the central type of ethical consideration. This “sub-system” of the ethical conceives of obligation—in contrast with what is ordinarily called obligation—as having an absolute priority over the agent’s desires and inclinations and as necessarily having to win in any conflict with other kinds of ethical consideration. Accordingly, blame, directed to the voluntary action of an agent endowed with free will, is conceived as the central type of ethical reaction to the conduct of others.

Though this type of ethical thought was given its purest expression by Kant, it is not in Williams’s view a philosophical invention. It is, he says, part of the outlook of us all. Yet his conclusion is that we should be much better off without it, for there are mistakes woven into it that spring “from a deeply rooted misconception of life.” One of these is “to require a voluntariness that will be total and will cut through character and psychological or social determination.” Of this Williams says that “it is an illusion that this demand can be met.”

Again at this point it is necessary to repress a natural impulse to ask for some general account of what should replace this misconceived form of morality; but to press this question would be to miss much of the point of this book. We can, Williams claims, only find out better ethical practices to replace those we have by “reflective living” in the light of as much clarity of thought and understanding of ourselves and our social life as we can muster. Philosophy can help in this, but only as other disciplines can, to detect and rid ourselves of fictions, to strip from our sentiments the disguise that makes them appear other than they are, and to articulate possibilities. What philosophy cannot do and should not attempt to do is to set up some standpoint outside all of our ethical practices to identify their proper course.

2.

I find it impossible without misleading oversimplification to summarize this rich and subtly argued book. Instead in order to convey something of its power and originality I shall consider Williams’s comprehensive discussion of the idea of objectivity in ethics that winds through the book. He shows that the view that moral claims may be objectively correct or incorrect may take either of two radically different forms, both of which he opposes. One view equates objectivity with realism and holds that moral claims are propositions (and therefore true or false); hence their justification depends on their relation to the world, so that belief in them, like belief in the propositions of science, may constitute knowledge of truths about the world.

The alternative view connects objectivity in ethics not with knowledge or truth but with practical reason, on the grounds that we are committed simply as reflective rational agents to accepting the claims upon us of a determinate form of ethical thought, which can therefore be said to have an “objective grounding” in practical reason.

Williams considers two forms of the argument that ethical thought has such a grounding in practical reason: one of them is Kant’s and the other is Aristotle’s. Aristotle’s starting point is the rational agent’s concern that his life as a whole should be one of well-being, a life worth living for beings constituted as men are. His aim is to show that for them the ethical life consisting of the acquisition and exercise in society of specific virtuous dispositions to think, feel, and act in certain ways is such a life.

To the exposition and criticism of this famous ancient theme Williams brings some welcome new ideas; but his conclusion is that without Aristotle’s metaphysical and teleological assumptions, or without an account of human nature which is merely a disguised form of ethical theory, no case can be made out for believing with Aristotle that one kind of ethical life in accordance with a determinate set of virtues is the uniquely correct and satisfactory development of human nature.

But though in this respect Aristotle’s enterprise is a failure, Williams shows how much there is to be learned from his broad and fruitful conception of the ethical as based on internalized dispositions to think, feel, and act in certain ways and how much this conception is needed to correct the narrow focus on action, obligation, and blame, which impoverishes much modern moral philosophy. Williams’s discussion of this part of Aristotle’s thought brings to light considerations of great importance for understanding claims to ethical objectivity. For, as Aristotle saw, the ethical dispositions which we have acquired are not only part of what we are, from which we start in deliberating about conduct, but they also constitute an internal perspective from which we see many other things in the world, other persons’ interests and the requirements of justice among them, as having ethical value independently of ourselves and our motivations. Yet we also command an external perspective; we can stand back from our dispositions and the internal perspective which they form. When we do so we can see that all that has to exist in the world for those things to have ethical value is, as Williams puts it, “people’s dispositions”; so that “there is a sense in which they are the ultimate supports of ethical value.”

For Aristotle the availability of these two perspectives was no reason for skepticism since for him there was nothing in the view of things from the outside which could conflict with the view of things from the inside. The dispositions that give us our ethical view of the world could, given his assumptions, be seen as the best development of human nature. But for us, who cannot share Aristotle’s assumptions, the agent’s internal perspective is only one of many compatible with human nature, and a gap may open between the two perspectives. So the thought that human dispositions are “the ultimate supports” of ethical value has for us skeptical implications.

Moreover there is for us another difficulty which did not and perhaps could not have presented itself to Aristotle. Even if our ethical thought is objectively grounded in practical reason as Aristotle thought, this would not show that ethical statements are true, but only that we are justified in accepting and acting on them. Yet that is not how such statements appear in the perspective of ethical thought itself, which cannot manifest the fact that it rests merely on human dispositions. For in that perspective they present themselves as truths about the world. So unless the realist conception of ethical objectivity can be established, ethical thought has, as Williams says, “no chance of being everything it seems.”

Kant’s form of the argument from practical reason to ethical objectivity starts not from the conception of a full human life, but from the bare idea of rational action. It purports to show that simply as free rational agents we are committed to conforming to the requirements of morality. Much in this argument depends on special features of Kant’s metaphysics, which brings together the notions of rational agency, freedom, and morality in the mysterious conception of the agent’s “noumenal” self, which escapes the drive of external causes only when the agent acts on impartial principles of morality, accepted by him as applicable to himself and others, whatever his and their desires may be.

But, as Williams shows, it is possible to construct a more mundane argument to show that ethical considerations are presupposed by rational agency. This argument is Kantian in spirit yet it does not depend on Kant’s metaphysics but only on a salient feature of human action as we experience it; namely that a rational agent does not merely act in accordance with psychological laws relating to desires and beliefs but acts on reasons. In doing this he takes the circumstances of his action not as mere causes but as considerations he must take into account before acting. So when he deliberates about what he should do, he stands back from his desires, seeing them not from their standpoint or from that of other agents but from the standpoint of impartial principles applicable to the conduct of all rational agents.

Though the starting point of this argument is sound, it fails because, as Williams shows, it rests on a mistaken assimilation (which Kant accepted) of the rationality of practical deliberation to the rationality of theoretical belief. Given some evidence on some question of fact, I may ask “What on this evidence should I believe?” But it is an important feature of this question that it is not primarily one that I ask about myself; for what I should believe on this evidence is the same as what anyone should believe on the same evidence; and putting the question in the first person is a mere derivative from the primary question, “What should anyone believe on this evidence?”

This is so because our factual beliefs aim at truths about the world which are the same for all; and in deliberating about their truth we implicitly invoke impersonal standards which harmonize all true beliefs. But practical deliberation is different in a crucial respect; the question, “What should I do in these circumstances” is essentially “first-personal” and not a mere derivative of and replaceable by “What should anyone do in these circumstances?” For the “I” of practical deliberation that stands back from my desires and reflects upon them is still the “I” that has those desires, and, unless I am already committed to the motivations of an impartial morality, reflective deliberation will not lead me to it. To hold otherwise is to confuse reflection with detachment; and that confusion has encouraged the mistaken idea that if our moral beliefs are to be more than mere prejudices they must be regulated by some general ethical theory au dessus de la mêlée of our ethical practices.

I find convincing Williams’s criticisms of these Aristotelian and Kantian arguments that morality can be derived from practical reason, but it is to be noted that he assumes without any discussion here that a person’s reason for acting is something that depends (though not necessarily directly or simply) on his subjective motivations, desires, or dispositions. But many philosophers ancient and modern have believed that a person’s reasons for acting and particularly his moral reasons may be objective or “external” in the sense of existing independently of his internal subjective state. On this view of what may constitute a person’s reason for acting or not acting, when I refrain from stealing out of fear of punishment I also have an objective or external reason for not stealing which depends only on moral considerations. Williams does not believe that there are reasons of this latter kind—any reasons that do not “speak to any motivation that the agent already has.”

Williams defended his subjective account of reasons for acting some years ago in an excellently argued article,¹ but his defense is not repeated here nor is the existence of controversy on this subject noticed except in a note on the penultimate page of the book. This is unfortunate, since the nonspecialist reader may believe the subjective view to be uncontroversial; the only clear reference in the book to the idea that ethical considerations may still be a reason for an agent to act even though he gives no weight to them, occurs in Williams’s discussion of the “peculiar institution of morality,” where the idea is treated as a mere fiction involved in the practice of blame.

3.

Williams’s discussion of the “realist” conception of ethical objectivity—the view that moral claims, like the propositions of science, may be knowledge of the world—proceeds mainly by way of a comparison between scientific and ethical thought. Williams defends with much subtlety the view that whereas science has at least a chance of being what it seems, namely an account of how the world really is, our reflective ethical thought has no such chance. It is tempting to suppose that the difference in this respect between scientific and ethical thought is shown quite simply by the fact that whereas there is a striking degree of convergence among scientists on the answers to scientific inquiries, there is far less agreement on questions of ethics. But such an account of the difference would be mistaken, because what is crucial is not the facts of convergence or of failure to converge, but the explanation of convergence where it does occur.

So, even if there were a complete convergence of ethical outlook between human beings, as might happen, the crucial differences between their scientific and their ethical thought could still remain. For what sustains the claims of scientific thought to objectivity is that the best explanation of the convergence of scientific beliefs is that such beliefs are “world-guided,” that is, guided by how things in the world actually are, whereas no such explanation seems possible in the case of reflective ethical thought.

In some of the most interesting, most difficult, and most controversial pages of his book, Williams confronts two diametrically opposed objections to his contrast between ethical and scientific thought. The first insists that the kind of objectivity that Williams ascribes to scientific thought cannot belong to it or to any human thought; the second insists that such objectivity can be ascribed to ethical as well as to scientific thought.

The first objection springs from a form of the pragmatism which is a constantly recurring feature of American as distinct from European philosophical thought.² It rejects the idea that there is an independent world with features that can be specified independently of our perspective on it, and to which our thought strives to be adequate. It also treats the conception that the history of science has been one of progress in discovering what the independent world is like not as a truth but as a cultural artifact, found convenient in various ways.

Williams’s main response to the pragmatist is to defuse his claims by a concession. It is true that the world cannot be thought of, or present itself to us, uncategorized in human terms; but this truth still allows us to think that there is an independent world to be described. It also allows us to suppose that the success of the description of it in terms of the categories of scientific thought is determined by that world. For though the perspective of scientific thought is a human perspective and not an “absolute,” perspectiveless conception, it is largely independent of the peculiarities of human observers and yields not only a description of the world but an explanation of its own success; that is, an explanation of how it is that human observers can observe and explain a world of that description.

Williams’s response to the pragmatist’s objection to his contrast between ethical and scientific thought seems to me effective. His argument can be met, if at all, only by some radical revision that I cannot now envisage of the distinction between the world and our thought about the world which his argument employs. But his reply to the second objection, which claims that ethical thought has as good a title as scientific thought to be considered objective and a form of knowledge, seems to me more controversial and less clear. This objection turns attention away from those general and abstract concepts such as “right,” “ought,” and “good” with which moral philosophy has been preoccupied, to specific and substantive concepts such as cruelty, lying, brutality, treachery, and gratitude, in terms of which much ethical thought is expressed.

These substantive, or “thick” concepts, as Williams calls them, are relevant to the argument about objectivity because they appear to express a union of fact and value in that their application is both determined by the facts of a situation and yet also involves an evaluation which in many cases provides reasons for action. Their application to the situations we face can therefore be, as Williams puts it, “world-guided” and “action-guiding”; it ranks both as knowledge of truths about the world and as evaluation. This union of fact and values cannot be explained away as if such concepts were simple conjunctions of a neutral descriptive element with an independent abstract “all-purpose” evaluative element. For if they were, their application to the world would be exclusively the function of the descriptive element and some other purely descriptive concept could pick out the same features of the world.

But in fact the application of a thick concept like cruelty to the world as it proceeds from one instance to another depends on its specific evaluative point and on the practical interest that has shaped its use. No purely neutral descriptive concept could discharge this function. Although only those who share the evaluative interest which thus guides the application of such thick concepts are in a position to assert statements applying them to particular situations, others with sufficient sympathetic insight may, as “outsiders,” grasp their evaluative force and so understand them and assess their truth.

At the end of a complex set of arguments Williams concludes that the use in ethical statements of such thick concepts and the kind of knowledge that their use may yield does nothing to show that reflective ethical thought can have the world-guided convergent character in which the objectivity of scientific thought consists. For it is only the unreflective, uncritical application of thick concepts, such as might be found in the ethical practices of a simple traditional society, which can be said to constitute knowledge of truths about the world. Members of this sort of society use such concepts in their ethical practices to guide conduct and to assess it and so to find their way around their own social world, but they do not compare their practices with alternatives or evaluate them as right or good by any standard that transcends them.

The limited scope of such unreflective use of thick concepts that is found in a traditional society does not reach to the concerns of critical and reflective ethical thought which is a pervasive feature of societies such as our own. Critical and reflective ethical thought, as Williams suggests, needs for its expression just those abstract “all-purpose” evaluative terms like “right” or “ought” which Williams is at pains to distinguish from the thick concepts whose unreflective use, in his view, constitutes ethical knowledge. At the reflective level there are, he believes, no thick concepts and so no hope of a world-guided convergence of reflective ethical thought to constitute knowledge at that level.

Williams seems to go even further than this in making the provocative remark that “reflection may destroy knowledge.” This seems to suggest that what was gained in the unreflective use of thick concepts may be shown by reflection on such concepts not to have been knowledge. But his point, more accurately expressed, is only that the further acquisition of such knowledge will not be possible once those who use such concepts reflect upon them and are led by such reflection to question them by reference to some general standards that transcend them.

Williams’s arguments are much richer in instructive and imaginative detail than I can show. But perhaps more than he gives us is needed to explain the conception of reflective ethical thought that figures in his arguments. For he writes here as if it could find expression only in those abstract all-purpose evaluative concepts of “ought” and “right” which enter into the mistaken analysis discussed above of those concepts as simple conjunctions of purely descriptive and purely evaluative terms. Thus, he says discussions with the ambition of arriving at the truth about the ethical will necessarily employ the most general and abstract concepts such as “right” which do not display world-guidedness.

But much that Williams has written in criticism of the pretensions of ethical theories purporting to establish general tests for the correctness of ethical beliefs suggests that reflective ethical thought could not possibly take this wholly abstract form. For example, he rejects the idea that “the reflective theorist can make himself independent of the life that he is examining,” and there is, he thinks, no reason “to believe that the theoretical reasonings of the cool hour can do without a sense of the moral shape of the world of the kind given in the every day dispositions.” These words suggest that there is need to consider the conception, developed notably by John McDowell that, to be intelligible, reflective ethical thought must be indirectly world-guided since it cannot dispense with all thick concepts but must proceed piecemeal relying on some of those concepts while criticizing others, much as in a famous image, which Williams himself adopts, sailors repairing their ship at sea have to rely on some parts of it while repairing others.

There is another related feature of Williams’s account of thick concepts which I do not understand. He envisages that some may “survive” at the reflective level, and presumably examples of this are our own use of such concepts as “brave” or “promise” as evaluative or as providing reasons for action. Our reflective use of these concepts is different from their unreflective use in a closed traditional society. Their force in evaluating conduct is not exhausted in guiding us around our society, as the evaluative force of our use of the concepts or rules of a ceremonial ritual or game may be. When we appraise conduct conforming to our thick concepts as right or good, we do not speak merely from within our social institutions or social world. We think of them as right or good in a wider, more abstract sense, which could be used to criticize these social institutions but is not, in Williams’s view, world-guided. But if this is so, then thick concepts reflectively used seem to be conjunctions, not of purely “descriptive” and purely evaluative elements as in the misleading account of thick concepts discussed above, but of two evaluative elements, one world-guided, the other not: calling someone brave is in part determined by facts in the world, but thinking of bravery as good in the more abstract sense discussed above is not world-guided. Yet Williams writes that judgments made with thick concepts that survive reflection can “straightforwardly be true”; but given the strong connection which he makes between truth and world-guidedness it is not clear to me how judgments made with concepts only partially or up to a point world-guided can “straightforwardly be true.”

4.

Finally, what bearing on practice will and should these arguments have? Why should anyone be disturbed if Williams is right in his claim that there is no independent rational foundation for ethical thought and no acceptable systematic theory to tell us what is morally right or wrong? In fact fears of two different kinds have been excited by such skeptical thought. The first is that if it becomes widespread we shall have nothing—or not enough—to say to the immoralist, whether he is the coldly selfish egoist of private life, or the brutal advocate of oppression in public life. But there is surely something laughable in the idea that anything we could draw from philosophy could weigh with such characters bent on having their way at others’ expense. Why should it matter to them that there is a philosopher’s proof that, in acting as they do, they are irrational, inconsistent, or flying in the face of some moral truths? As Williams says, “What will the professor’s justification do when they break down the door, smash his spectacles, take him away?”

The second kind of fear is more realistic. It is concerned not with the reactions of the immoralist or the egoist but with more ordinary people who with various degrees of conviction, difficulty, and backsliding manage to live up to the moral standards they have acquired and developed in their social life and to transmit them to their children. Williams himself does not suppose that a general acceptance of the view that there is no objective foundation for ethical thought, either in practical reasoning or in truths about the world, can leave ethical thought and practice unaffected. For once it is accepted it will be no longer possible to interpret ethical experience, as many do, as a demand coming from outside. The sense of necessity (the moral “I must”) in which the recognition of moral obligation often terminates, will have to be seen as coming not from outside, but from what is most deeply inside us even though it is normally also supported by others who share our practices and beliefs. The fear is that this will not be enough and that when we come to think of our moral standards as resting on no further foundation, we shall disregard them whenever they stand in the way of our getting or doing what we want.

How likely this is, is a question of moral psychology about which we know little enough, but it is plain that those most likely to abandon their moral beliefs when these are shown to have a subjective source are those whose moral sentiments have been formulated apart from concrete situations and kinds of conduct, and have become focused on general principles and theories or on the divine will or on whatever is taken to be a general authoritative source of all moral right and wrong. The moral monster who thinks there is nothing morally wrong in torturing a child except that God has forbidden it, has a parallel in the moralist who will not treat the fact that the child will suffer agony as in itself a moral reason enough.

Such a person may think that restraint in deference to that fact has no moral worth if it is not seen, in moral deliberation, as the requirement of some wider general principle or theory or “moral law.” Conversely, those whose moral education or self-education has not led them into this mode of moral thinking and who find their moral reasons at the ground-floor level of particular concrete situations are least likely to be disturbed by the revelation that their moral practices, and the feeling of constraint and necessity that accompanies them, are reflections of concerns which lie deep in their character.

A belief in the objectivity of ethical thought is only one of the conceptions that may make our ethical practices appear other than they are. Scattered through this book are pointers to many others, some of which underpin practices such as moral blame, which could not survive in its present form if the confused notion of voluntary action as somehow independent of social and psychological conditions on which it relies were generally seen to be what it is. On the other hand there is also in this book much to reassure those whom it might scare; for Williams argues that nothing in its sustained critique of the philosopher’s constructions of ethical theories and the modes of moral thinking which inspire them means that our ethical practices are mere unexamined prejudices or that there is any presumption in favor of egoism and Realpolitik as natural and rational, or that we have no ideas to support freedom and social justice. But as Williams foresees in a brief final chapter expressing his tentative hopes for the future, a large task awaits the philosopher. Instead of directing his efforts to the construction of ethical theories he should be engaged in marshaling whatever social knowledge, including history, is available which can give critical insight into the working of our institutions and increase our self-understanding. To that task this valuable book with its many insights may be seen as a prolegomenon.