Morality and Reality

I

“If we can depend upon any principle which we learn from philosophy, this, I think, may be considered as certain and undoubted, that there is nothing, in itself, valuable or despicable, desirable or hateful, beautiful or deformed; but that these attributes arise from the particular constitution and fabric of human sentiment and affection.” The issue thus raised by Hume and by those who, horrified by Hume, insisted that moral right and wrong are “real characters of action,” not “qualities of our minds,” was long thought to be the central issue of moral philosophy.

Perhaps most of those who approach moral philosophy for the first time still expect to find either some refutation or confirmation of doubts about the “objectivity of value,” and they are disappointed when they find the question not merely unanswered but not even debated, and in its place detailed discussions of logical or linguistic issues. These discussions appear to many to take Hume’s skeptical position for granted and merely to refine upon it, substituting references to the “speech acts” of “commending,” “grading,” or “prescribing” for the eighteenth-century references to “sentiments and affections.”

Many contemporary philosophers believe that the “objectivity” or “subjectivity” of moral value, the truth or falsity of what may be called ethical realism, is not an important or even a genuine issue. Even if there were constituents of reality corresponding to the concepts of moral right and wrong, good and evil (and not merely to descriptions of the conduct or states of affairs which are judged in these terms), the course of moral argument, moral agreement and disagreement, would be unaffected: no moral problem would be easier (or more difficult) to solve. So the alleged value constituents of reality, the existence of which cannot in any case be demonstrated, can be dismissed as explaining nothing.

Both John Mackie (of Oxford) and Gilbert Harman (of Princeton) unfashionably, and I think rightly, take the question of the objectivity of value as the starting point of their introductions to moral philosophy and devote many pages to it. They are right in doing this because the objection that the alleged objective values or ethical constituents of reality explain nothing is an argument (which Harman develops at length in a brilliant first chapter of his book) against the existence of “moral facts” and not a reason for not discussing the issue. Even if the course of moral judgment and argument would remain unaffected, the loss of the belief that they are backed by something more than human attitudes or policies is, and will continue to be, for many as profound as the loss of belief in God. And in our own day the sense of the tragic “absurdity” of life in a world which contains no values awaiting discovery has spilled over into philosophy itself. Existentialism is the philosophy par excellence of the disappointed objectivist.

In any case, as Mackie argues, the issue is of great importance to general philosophy. If we had to accommodate a belief in objective values or, as in the teleology of the ancient world, think of nature as containing intrinsically desirable ends of human conduct, this would make a radical difference to metaphysics, to epistemology, and to the psychology of human action.

II

“There are no objective values.” So runs the first sentence of Mackie’s book, proclaiming what he terms second-order moral skepticism: “second-order” because what is rejected here is not the practice of morality but a particular view of the status of moral values and of the way in which moral right and wrong, moral good and evil, fit into the world. In similar vein but in less firmly assertive and enthusiastic tones, Harman elaborates in the first chapter of his book reasons both for holding that moral judgments cannot have the kind of objectivity traditionally attributed to them, and for taking seriously what he calls moral nihilism—the view that there are no moral facts and so no knowledge of moral right and wrong, good or evil.

Though from these apparently similar starting points Mackie and Harman reach different conclusions, the books have other similarities. Both are addressed to (among others) the beginner and the general reader as introductions to moral philosophy. But in contrast with many examples of what has become a rather depressing philosophical genre, they do not merely survey and explain the merits of the familiar rival theories (intuitionism, Kantianism, utilitarianism, emotivism, prescriptivism) and their several variants, and then leave the reader to take his choice. Instead these authors make often novel, always interesting and illuminating criticisms of these much discussed theories, and they both defend a distinctive controversial position of their own.

Both books, particularly Harman’s, considered as introductions to the subject, set tough standards. The layman who can grapple unaided with Harman’s chapters on Kant and on the ontology of reasons for action will certainly be a gifted student of philosophy, and even amid Mackie’s generally most lucid undecorated prose there are too many philosophical technicalities (e.g., “illocutionary force,” “prescriptivism,” “the categorical imperative”) introduced sometimes with inadequate explanation and sometimes with none at all.

A more important similarity between these two books suggesting a general contemporary mood in Anglo-American philosophy is the critical attitude toward linguistic philosophy shown by both writers. Mackie is quite clear that although linguistic or conceptual analysis may have a place in philosophy, such analysis leaves untouched the central question of the objectivity or subjectivity of moral values. An account of “meaning” such as linguistic philosophy supplies, he insists, is not an account of what there is and cannot take its place; so it cannot either confirm or refute moral skepticism. At best, philosophical scrutiny of the so-called language of ethics can serve to make explicit what those who make moral judgments, the users of the language, believe about this central issue.

But what they believe may be confused and inconsistent, and even where it is clear and consistent may be false. Mackie argues that a belief in objective values is indeed false and so holds what he calls an “error theory”: that false beliefs about the status of morality are built into the structure of moral concepts and consequently into the meaning of moral judgments. Hence Mackie views the positive contribution which the moral philosopher can make after showing what is false in ordinary views as literally a constructive one: it is not to discover but to take advantage of what he terms the “malleability of ethics” by working out, in the light of our understanding of certain relatively permanent tendencies of actual human motivation, acceptable principles of constraint on action. Respect for these principles, in his view, will do most to counteract certain evils to which the human condition is exposed.

Harman’s theoretical conclusions about the nature of morality (as distinct from their practical conclusions) are very different from Mackie’s. He argues that, notwithstanding the reasons for skepticism which he explains in his first chapter, moral judgments can after all have a kind of objectivity, not absolute but relative, sufficient to dispel the nightmare of moral nihilism. I do not, for reasons which I explain below, find these conclusions illuminating or the argument for them convincing; in particular Harman’s dependence at certain crucial points on an analysis of the meaning of what we say both in making and in withdrawing moral judgments is extraordinarily perplexing in view of the anathema upon “linguistic philosophy” that he pronounces in his preface.

Here Harman seems to go far beyond Mackie: for him “the old linguistic philosophy” or “meta-ethics” as he calls it (“meta-linguistic ethics” would be a more accurate term) rests on an incoherent belief that questions of meaning can be distinguished from questions of substance, and has been worse than useless. It has, he believes, caused students, bored with “the series of questions about meaning” with which the linguistic philosophers had replaced the basic philosophical problems about morality, to turn to normative ethics and to grapple with the more interesting and “relevant” first-order practical moral issues of the day (war, abortion, equality), thus losing touch, in Harman’s view, with the philosophy of the subject. So Harman’s book is written with a partly pedagogic aim to reverse this malign trend and to redirect attention to genuine philosophical problems about morals, unhindered, as he puts it, by the meta-ethical “baggage” or “apparatus” of the linguistic philosophy which he says “had just about come to an end by 1960.”

III

Harman holds the basic philosophical problem about ethics to be the apparent fact that it is cut off from “observational testing.” His exposition of this point, though no doubt owing much to the philosophy of W.V. Quine, is fresh as well as illuminating, and it has the great merit of offering a definite view about what exactly is meant by a claim or denial of objectivity. Too often what is to count as objectivity is obscured by the use of metaphors or clichés, English and American, so that the question is said to be whether good and evil, right and wrong, are “part of the fabric of the universe” (English), or whether there is an ethical “fact of the matter” (American). Harman makes it clear that he means by “observation” the making of an immediate judgment in response to a situation without conscious reasoning or inference, and he insists that in this sense there are both moral observations as well as non-moral observations. So, to take his example, confronted with children pouring gasoline on a cat and lighting it, we may make both the non-moral observation that the children are doing just that and the moral observation that what the children are doing is morally wrong.

Both kinds of observation are “theory laden” in that in both cases what is perceived (the children doing what they do and the wrongness of this) depends on the background of theory, the concepts and beliefs possessed by the observer and brought to bear by him on the situation. But at this, the crucial point, ethics and science diverge. A scientist testing a scientific theory sees a vapor trail in a cloud chamber and thinks, “That is a proton.” The making of this observation confirms the physical theory because the assumption that there really was a proton, with the properties assigned to it by the theory, going through the cloud chamber is the best explanation of the vapor trail which the scientist saw and so explains his making the observation.

The case of a man making the simpler observation that the children are pouring gasoline on a cat is similar: the assumption that that is what the children are really doing is the best explanation of his making that observation and so it is confirmed by it. But the moral judgment that what the children are doing is wrong is different. In order to explain the making of this observation there is no need to assume anything about moral facts such as that it is really wrong to set cats on fire. Any such assumption would be irrelevant to the explanation of the making of the moral judgment by the observer. All that is needed to explain the making of the moral judgment is the “psychological set” of the observer viz. that he has more or less well articulated moral principles which are reflected in his judgments. So the making of the moral observation is evidence only for facts about the observer, his dispositions and moral sensibility; it is not explained by and is not evidence for the truth of the moral principle that such behavior is wrong.

So moral nihilism, the doctrine that there are no moral facts, no moral truths, and, most important of all, no moral knowledge, has to be taken seriously. Harman takes it seriously but thinks that it can still be avoided, and among the many different escape routes from it which he tries—and finds blocked—is the view that moral knowledge is not something which can be gained by observation and so it may be a mistake to bother about “observational testing.” Instead it may be that the knowledge of morality is a priori knowledge of principles of conduct by which any rational being will necessarily be motivated and are discoverable by “practical reason” because reason is their source. Harman discusses this Kantian doctrine and its post-Kantian variants but concludes that it cannot be shown that reason has such powers as Kant claimed for it. The only constraints that pure reason can place on the selection of principles of conduct is, as the existentialists claim, the constraint of consistency; reason cannot pick out from a plurality of consistent principles one uniquely rational set or find any special place for principles prohibiting the killing of innocent persons, lying, stealing, and prescribing some minimum care for others which are central elements in morality as we understand it.

Harman also considers various more or less sophisticated versions of emotivism, the moral theory that takes the function of moral judgments to be not that of stating truths but of expressing attitudes of approval or disapproval. The more sophisticated version of this theory, according to which approval or disapproval, to count as moral, must be attached to general principles of conduct, does allow us to say of some particular moral judgment “That is a fact” or “That is true.” But we can say this only because these expressions amount to little more than an endorsement or assertion of the correctness of the application of the general principle to the particular case.

This last position Harman terms “moderate” moral nihilism but for him it is not enough. Why? Because given only these weakened senses of “fact” and “truth” we cannot, he thinks, claim to know moral facts and any theory that does not allow this he believes must completely undermine morality as we ordinarily talk and think of it.

So the abyss of moral nihilism still yawns. Harman’s way back from the abyss is a social form of moral relativism. We cannot, he thinks, give up the idea that we have knowledge of moral facts, but we can (apparently) water it down; a conception of morality as a set of tacit social conventions varying from society to society will enable us, notwithstanding the points made in his first chapter, to claim knowledge of “moral facts” albeit relative moral facts. Such facts will help to explain the making of judgments and so such facts can rank as evidence for the truth of moral principles.

Harman therefore argues first for relativism and secondly from relativism to the existence of moral facts which explain the making of moral judgments and entitle us to speak of moral knowledge. The first step in this argument relies heavily and surprisingly (given the earlier repudiation of the “baggage” of linguistic philosophy) on a highly controversial analysis of the meaning of “ought” and “wrong” as used in moral judgments. It would, Harman says, be a “misuse of language” to say of hardened professional criminals that it is morally wrong of them to steal or that they ought not to kill, or to say of Hitler that it was morally wrong of him to order the extermination of the Jews.

This is not because “ought” or “wrong” are too weak to match such misdeeds, but because we cannot say that such men, not sharing our principles, have any reason not to do these things: they are, because of the meaning of these expressions, outside the reach of such moral judgments. Harman considers it a possible (though probably false) view of Stalin that he was not so far beyond the pale as Hitler. Hence, he writes, “it is not odd to say” that it was wrong of Stalin to do what he did (though they were equally bad things) “in the way in which it is odd to say” this about Hitler. Harman then uses these findings about meaning to correct what he says is “our ordinary view that we can make moral ‘ought’ judgments about anyone, no matter what their principles are.” With this the obstacle presented by “ordinary views” is cleared, making way for the social moral relativism which Harman holds to be the truth about the nature of morality.

This certainly seems a strange example of linguistic philosophy. If ordinary language does not reflect but, as Harman finds, is inconsistent with “our ordinary views” about morality, how is it established what the ordinary view is? It is not of course impossible that ordinary men when asked for a general view about some subject should give answers at variance with what is implicit in what they say about particular cases. But it is surely more likely that the ordinary view that we can make such moral judgments about others whatever their principles is a consequence of the fact that such judgments are frequently made. Harman gives no hint of the nature of the evidence that, in the case of morals, ordinary views are in conflict with ordinary language.

There is too a further difficulty. Harman claims that if “we” find that someone about whose conduct we have made such moral judgments does not share with us the relevant principles on which they are based, we withdraw our observations “but not as mistaken.” We withdraw them because in making our observations we have wrongly “presupposed” that the subject shares our principles. Since this turns out not to be the case we withdraw our judgment not as false but as (together with its negation) having no application to him. It is difficult to believe that Harman (or anyone else) has heard (or read) of many reallife examples of this phenomenon of withdrawing-but-not-as-mistaken. And if the ordinary view is, as he says, that moral judgments can be made about those who do not share the judger’s principles, why should the phenomenon ordinarily occur?

Finally there is Harman’s step from social moral relativism to the desired conclusion that there is knowledge of (relative) moral facts. Suppose it is the case that, because of their meaning, moral “ought” judgments can only be made about those who share the judger’s principles, and that, as Harman says, morality is essentially social, consisting of shared convictions and principles of conduct. How would this show that moral principles are, as he claims, after all like scientific theories in the crucial respect that they are subject to “observational testing,” so that we may properly claim to have “knowledge” of “moral facts” and the ghost of moral nihilism may at last be laid to rest?

Of course it is a fact that those who genuinely accept moral principles make moral judgments in accordance with them; but this is not a moral fact and not what Harman means. He appears to argue in his chapter on reasons (though I confess to fear that I have not understood him here) that if I pronounce some moral judgment about you in accordance with conventions or principles which we both share, e.g. (my example), if I say “you ought not to steal that book,” there is a “relative” moral fact—i.e., that there is a good reason for not stealing—which explains my making my moral judgment about you and so is evidence for the truth of the moral principle that it is wrong to steal.

But surely there is some philosophical double-counting here. My making my moral judgment about you is explained by what in his first chapter Harman calls my “psychological set.” I have accepted or internalized the principles in question and believe you and others also have done so, and I have acquired the disposition and capacity to judge particular cases in their light. The “moral fact” that relative to these principles there is good reason for not stealing only helps to explain my making my moral judgment because mention of it is required in the full relevant description of my “psychological set”; that is, in the specification of what my accepting the shared principles of a society amounts to in relation to this particular case.

If this is all that “knowledge of relative moral facts” amounts to it is not only a feeble nostrum against moral nihilism, but has the positive disadvantage that it may perpetuate a confusion between intersubjectivity and objectivity, by disguising the former as some thin or shadowy form of the latter. The case for social moral relativism would be better presented without this.

There is, however, a possible interpretation of the different strands of argument scattered through Harman’s book which would show his case to be more interesting, if not more plausible, than my criticisms have suggested. It may be (and this would be in accord with the strongly holistic form of theory which Harman, like Quine, favors) that his main reason for holding that morality is a matter of tacit social convention to which a maker of moral judgments subscribes is that this is a hypothesis which explains a wider range of phenomena than any other moral theory.

This range would include not only the (alleged) “oddness” of saying of hardened criminals, or Hitler, that it was wrong of them to do what they did and the (alleged) phenomenon of withdrawal-but-not-as-mistaken. It would also include the sense that moral requirements both have an external source and are self-imposed. Yet apart from doubts about the alleged linguistic phenomena there are, as Mackie shows, better explanations of these things and also of what this hypothesis fails to explain, namely that it is (as Harman asserts) our “ordinary view” that we may make moral judgments about anyone whatever their principles are.

It may also be that Harman’s chapter on reasons contains some answer to the charge of philosophical double-counting in his argument from relativism to the conclusion that there are (relative) “moral facts.” It may be said that my charge of double-counting ignores an important truth: that in explaining the actions of those known to share our desires and principles we make use of the hypothesis that they are moved by certain reasons as we ourselves would be if we were in their place and this is to regard such reasons as objective, though relative, facts. But here, alas, my lack of understanding is invincible: I cannot see how anyone worried about moral objectivity could be consoled by this truth or how the case for moral relativism is strengthened by it.

IV

Mackie treats as the central issue lying beneath discussions of the objectivity of value a question about reasons for action. Are there reasons for human action, or requirements to be met by it, which are not contingent upon any desires, purposes, or principles which men actually have but are independently and intrinsically authoritative? Such reasons and requirements may be called “transcendent” (not Mackie’s term): if they exist and can be discovered, it will be possible to demonstrate that there is something that a man, just as a rational agent, and whatever his contingent motivation may be, must (or must not) do. If such transcendent reasons do not exist, what a rational agent must as such do is limited to what is appropriately related to the desires, purposes, or principles by which he is actually motivated. Plato, Aristotle, and Kant held that there are such transcendent reasons or requirements, while Hume was the pioneer of the skeptical view that there are not, and his famous, often misunderstood, dictum that “reason is and ought to be the slave of the passions” simply summarizes the skeptical view.

Mackie agrees here with Hume and argues against the belief in objective value and transcendent reasons, which he holds to be not meaningless but false, but this is one of the few parts of this book where the discussion is, I think, too cursory. His main argument is that ends of human action which have such intrinsic desirability or “to-be-pursuedness” built into them, or courses of action which have an authoritative “not-to-be-doneness” built into them, are entities “of a quite different order from anything with which we are acquainted.” So their “queerness” is the main argument against their existence. So also the “queerness” of the special faculty of moral perception which would be required to account for human knowledge of them is the main argument against its existence.

But mere “queerness” of the alleged ethical object and the corresponding faculty seems in itself no better an argument against their existence than an argument against the existence of electrons from the “queerness” of particles moving from one point to another without passing through any intermediate point. The reliable argument here must surely be not one from “queerness” but the argument developed by Harman: that unlike the hypothesis of the “queer” electron the hypothesis of the “queer” ethical object explains nothing that cannot be better explained without it.

The ethical object has, so to speak, no vapor trails which it explains and so are evidence for its existence. This would open up for discussion the possibility that there are such “vapor trails.” It may be argued that the hypothesis that there exist intrinsically desirable properties of action or states of affairs attainable by human action which, when known, influence the deliberations and conduct of rational human beings independently of any contingent motivation in fact explains better than any other hypothesis morality as we know it: the way in which we acquire and maintain moral beliefs, their content and their effect on our lives. Mackie’s argument does not directly meet this point but his book contains convincing reasons for preferring other explanations.

Mackie argues for an “error theory” that false beliefs are built into the meaning of ordinary moral judgments. His evidence is presented in a detailed examination of the meaning of “good” and “ought” (often a philosopher’s misnomer for “must”), and the argument, in effect, contradicts Harman’s account and is to be preferred to it. There is, Mackie argues, a unifying thread of meaning common to the non-moral and moral uses of these terms, but just at the point of transition from their non-moral to their moral use the element of transcendent reasons or requirements, falsely believed to exist, emerges.

It is however a complexity and a virtue of Mackie’s account that interwoven with this argument he gives another explanation, based on what he calls an “institution,” of what it is we are doing when we make moral judgments and demands on those whom we know may not be motivated by our principles. An “institution”—for example, the institution of “promising”—exists whenever, as in the case of an established moral tradition, there is widespread acceptance of rules or principles as guides to conduct and standards of appraisal: one who subscribes to those may speak “from within the institution” using the languages of “ought” or “must” simply to express, or demand, compliance with the requirements of the institution without any certainty that those they judge or address accept its principles, though usually, no doubt, they hope that they do, or that they may be stirred to do so. In so speaking we are projecting our principles on to others rather than objectifying them. This it seems to me is the element of truth exaggerated in familiar theories that take the “speech act” of “prescribing” as the paradigm of moral judgment. But when an institution is supported by wide, diffuse, or anonymous social pressures, projection and objectification may be inextricably mixed. Neither, as Mackie says, can be taken as disclosing “the” meaning of moral judgment.

V

If there are no transcendent reasons, anyone reflecting critically on the constraints of morality will ask: what reasons are there, so far as the desires or purposes by which he is or is likely to be actually motivated are concerned, for submitting to such constraints? For an important group of moral constraints, this involves asking what is their object or point. In his answer to this question Mackie adapts for moral philosophy the main idea of Hobbes’s contractarian political theory. Human beings are competitive with and vulnerable to one another; living in a world of scarce resources they need a “Sovereign” to enforce a system of restraints. This is a device which by counteracting human egoism provides some of the essentials of human welfare: security from violence and from deception, and respect for keeping promises and for property and fair dealing.

There is a similar but more fundamental need for an internal stabilizing analogue of the Sovereign’s coercive power, the invisible checks and ties of morality, the sense of moral obligation, supporting similar mutual restraints. Without this, as the “Prisoners Dilemma” (and other models of games theory) shows, there will be no sufficient ground for general confidence that sufficient conformity will be forthcoming to prevent men from abandoning cooperation for the disastrous course of pursuing without restraint their separate uncoordinated interests.

It is a virtue of Mackie’s account that the bare and indeed brittle Hobbesian bones are fleshed out with an important addition. In order that the “device” of morality should work, those who work it must regard it and be known to others to regard it as more than a device, and must have acquired a disposition to act on moral grounds without calculating the pay-offs. “A humane disposition is a vital part of the core of morality.” But dispositions cannot be switched on or off to suit particular calculations of self-interest and there are limits to the fineness of the discriminations they can incorporate. So though most men at most times have reason enough in self-interest to welcome the moral device, its practice inevitably extends beyond self-interest and indeed transforms it.

Any philosopher offering to explain in such ways the “point” of morality is open to the charge that he is presenting the special moral theory of utilitarianism in the guise of a general account of morality. Mackie has two answers to this charge. His picture of morality as a device for protecting human welfare from certain specific evils arising from human egoism is not utilitarianism but the core of good sense which the utilitarian has distorted and blown up into an “ethic of fantasy”: the conception that all that matters morally is the “maximization” of the total (or average) general happiness or satisfaction of desire. The chief vice of this conception, in addition to its impracticability, is that it allows and indeed requires not only the replacement of the satisfaction of one desire by another but the replacement of one person’s happiness by that of others. This would strip away all independent significance from forms of activity, special relationships, or principles of distribution which human beings actually value. Mackie’s critique of utilitarianism along these and other lines is an important contribution to contemporary moral theory.

In any case for Mackie morality identified as a set of constraints on egoism (“morality in the narrow sense”) is not the whole of morality though a necessary foundation for the rest. In the broad sense morality is the total set of priorities by which men conduct their lives and which embody some more or less articulate conception of the good life. If we accept that there are no transcendent reasons to dictate what the good life must be, the good life must be one that men are actually disposed to welcome; and since men for cultural, political, and religious reasons will have different views, there are few generalizations to be made about its content. Yet Mackie finds some very Aristotelian things to say about its structure; he allots a very central place both to the satisfaction (within the constraints of morality in the narrow sense) of egoistic and only relatively altruistic aims and to steady states of character or dispositions of informed choice in which, for Aristotle, the virtues consist. But there can be no reason, if the main arguments of the book are accepted, either for forcing one’s own conception of the good life on others or for sacrificing it in order to maximize the general welfare—however much utilitarianism may have conspired with Christian ideas of the brotherhood of man to make this appear a plausible ideal.

VI

I have considered only some of the main arguments in these two densely packed books. Both authors touch off interesting new lines of thought over a wide range of other topics: Harman most notably on egoism and its social transformation; Mackie in an original three-stage analysis of the alleged “universalizability” of moral judgments. But whereas Harman’s central argument seems to me to fail and Mackie’s to succeed, the latter’s book is open to a more general criticism. Its provoking subtitle “Inventing Right and Wrong” and its great stress on the idea that morality is something to be made, or “malleable,” diverts attention from the complex connections between morality and the emotions, and the conflicts and uncertainties that attend those connections. Mackie says too little about the emotions either as subjects of moral assessment or as its accompaniments. Even if they are not incorporated in their meaning, the emotions are generally present underneath the surface of moral judgments, especially self-judgment, and are often the chief guide to the interpretation of what is judged. Morality would be a very different thing from what it is if indignation, admiration, contempt, guilt, or remorse were not the standard accompaniments of the moral use of “must,” “ought,” or “good” studied here.

There is also, I think, a too easy assumption implicit in this book that if a cool head is kept, no ordinarily intelligent person will meet with moral conflicts that cannot be resolved by some sensible “trade-off,” which though it may ruffle, will not profoundly disturb, an intelligently “invented” morality. Even a beginner should perhaps be warned that morality is not likely to have so much cool consistency or easy “malleability.” We may need, as J.L. Austin said, “the grace to be torn between simply disparate ideals—why must there be a conceivable amalgam—the Good Life for Man?”

These last matters apart, Mackie’s book manages astonishingly well to sustain careful, detailed, and informative argument throughout its wide range and to defend some strong central theses which, accepted or not, are a powerful source of illumination.