T. M. Scanlon
T. M. Scanlon; drawing by David Levine

Moral philosophy in English-speaking countries is an inquiry now pursued principally in universities. This is accidental rather than essential; the great reforming philosophers, Jeremy Bentham and John Stuart Mill, were not professors. But still the present fact determines much of the style and form of contemporary moral philosophy, and determines specifically the very high standards of clear argument, apt for education in seminar and lecture rooms, which are imposed on moral philosophy. The sentimental seductions and unashamed fantasies of Rousseau and of other writers of his century, however influential within the discipline, are not now encouraged.

T.M. Scanlon is a scrupulous, astringent, relentlessly exact writer, without any of the fuss and flutter that come from the desire to please. His book is pure philosophy, unadorned. But from the beginning he faces an enormous difficulty. What is to be included under the heading of “morality”? On the one hand it seems that we must circumscribe and delimit some part of the vast area of human interest that might plausibly be categorized as moral concerns. On the other hand it seems that no decision on this question of what should be included can be morally neutral, because any decision must presuppose a particular moral outlook, which will exclude related interests as being other than moral, or less than moral.

If my failure or mistake on a particular occasion was a moral failure or mistake, it can scarcely have been a trivial failure or a trivial mistake. This much sense the word seems to carry with it through all its interpretations; and Scanlon has an interesting section on the concept of importance when we apply it to moral decisions. But does the word “moral” carry the further implication that if a failure or mistake is a moral failure or mistake, it is worse than any other kind of failure or mistake? If there is a moral argument against putting up a new building on a particular piece of land (for example, it tramples on individuals’ rights), does that argument necessarily override all other concerns in the matter (economic efficiency, aesthetic considerations, etc.)? The overridingness of morality is a crucial issue here, and I am not entirely clear where Scanlon stands. He seems to me to have good reasons within his own argument to be uncertain about it.

He is a strict circumscriber of morality, at least for the purposes of his book. By morality, as he proposes to expound it, he means the whole set of principles that govern, or should govern, an individual’s relations with other people within a community. When the search for a definition of justice begins in the early pages of Plato’s Republic, a definition of justice very similar to Scanlon’s notion of morality is suggested to Socrates by a high-minded, unphilosophical man: to be just is to repay what one owes. Scanlon claims that there is a “central core” of morality, discernible in almost any society, which is constituted by fair dealing among fellow citizens, the recognition of explicit and implicit contracts, the absence of freeloading and of any other forms of cheating or taking unfair advantage.

His project is to extend the range of this implied social contract based on reasonableness, in accordance with a fixed principle, until the contract covers a very large part of morality, as it is ordinarily understood in a civilized community. The fixed principle requires that in any serious situation our conduct ought to be attributed to a reason, or reasons, which any reasonable person would accept as being reasonable or not unreasonable. We have therefore a kind of test, by design a very loose test, of whether our action is, in the particular circumstances of its time and place, morally wrong. It is wrong if the supporting reason for the conduct fails this test.

The looseness comes in part from Scanlon’s notion of the reasonable, which is to be sharply distinguished from the rational. He rejects Kant’s discovery of practical rationality in his universal formula, the Categorical Imperative. For Scanlon, there exists only the community of persons who respect their shared deliberative processes as they weigh the pros and cons in their search for good and sufficient reasons in making any important decision. Even if a harmony of judgment among the reasoners is not to be guaranteed by such a process, at least moral wrongness will be thrown into relief; it will always be seen as a flouting of principle and a repudiation of judgment by a person who wrongly excepts himself from fulfilling the implied contract of reasonableness. Loyalty to the implicit contract of consensual reasoning is an unchanging social good, he argues, which does not vary from society to society, in obvious contrast with the many customs and traditions that vary as a result of differing social histories.


Scanlon therefore believes that he has found a bedrock for morality in the narrow sense of what we owe to others, and bedrock is what moral philosophers are typically looking for. Salvation comes from the word “reason,” which excludes not only the divisiveness of the passions of different individuals, but also the divisiveness of particular conventions, traditions, tastes, and fashions. Scrape away these historical accretions, according to Scanlon, and you will be left with the naked Homo cogitans, looking for universally acceptable principles of conduct.

Scanlon seems to me to stop short here of making the essential point: that we can appeal to universality in the formal processes of moral reasoning but not to any universality in the content and the conclusions of the reasoning. He requires that priority be given to his contractualist morality, “the central core,” over all other moral values; and “what we owe to each other” is not normally to be overridden by the pursuit of other objectives which we consider unquestionably good, for example, the objectives of art or science or religion. Scanlon calls such objectives “impersonal values,” and writes, “Impersonal values do not provide, in themselves, reason for rejecting principles of right and wrong. Considered in themselves, they represent a category of value, or of morality in the broader sense, that is distinct from what we owe to each other.” We have to take these impersonal values into account in determining what we owe to each other, and in deciding what principles we can justifiably reject. But they cannot be part of the bedrock, which is always constituted by interpersonal relationships, based on reasonable principles of mutual respect, as if in our moral reasoning we each become, for the occasion, citizens of an ideal community of reasoners in the style of Rousseau and of his General Will.

Never dogmatic, always tentative, Scanlon immediately admits that “there still may be cases in which we have to choose between impersonal values and what we owe to each other.” The method of resolving the conflict is to ask what, in the last analysis, we agree we owe each other, which might occasionally lead us to exalt the impersonal values over our apparent obligations. There is evident uncertainty in Scanlon’s argument here, and the word “reasonable” offers little guidance. Scanlon is relying on the immense weight, accumulated over centuries of Western thought, of the notion of reason and reasonableness, which he thinks will provide him with some sort of confirmation of standard moral judgments and some protection against skeptical and relativist philosophies. He writes repeatedly of the “authority” of moral judgments and of “their special force.” With characteristic fairness he acknowledges the diversity of moral ideals and of moral ambitions and of moral imagination. But he also writes:

The possibility of a broader range of variation in standards of moral appraisal emerges when we take into account the fact that the term “moral” is commonly used to cover much more than the morality of right and wrong that contractualism [his theory] seeks to characterize. Many of the forms of variation that anthropologists have studied, and that may have shocked the moral sensibilities of Europeans when they were first reported, involve such things as variations in sexual practices and in marriage and kinship structures. They are part of “morality” only in the broader sense, and some might say that they do not raise moral issues at all.

This is a surprising statement, because sexual practices and kinship structures and the obligations of marriage are surely at the root of “what we owe to others” in the private sphere, together perhaps with various prohibitions against violence.

Scanlon claims that “uniform moral standards are applied to the practices of different societies and to the conduct of individuals in those societies,” and again, “the central core [of morality] contains those judgments of right and wrong that hold everywhere.” But the cost of this achievement of universality will be the exclusion from the central core of private morality of much of what we owe to each other. It is evident that sexual practices and family relationships, and kinship relationships more generally, are immensely variable, in different places and at different times, not only in their demarcation as moral or immoral, but also in the moral weight attached to them. The regulation of libido, together with the prohibition of murder, belongs to the origins and essence of morality, but the specific question about sexuality, “With whom and under what conditions?,” belongs to the more conventional and the more historically conditioned superstructure of morality.

How does this conflict between claims of universality and particular variable conventions come about in Scanlon’s account of morality? I think it arises because of the inherited mythology of “reason.” Scanlon starts his argument with the premise that any acceptable justification of past or present conduct must pick out the reason that gives rise to the conduct; and the reason will have the form of a principle that any reasonable person will accept. He writes,


The fact that I desire something does not itself provide me with a reason to pursue it. Being an object of rational or informed desire may be correlated with the presence of such reasons, but these reasons are provided not by this hypothetical desire, but by the considerations that would give rise to it, or make it “rational.”

This seems to me to be false, insofar as it blurs a necessary distinction, which we all habitually recognize. An example: an amateur collector of Italian Renaissance bronzes “falls in love,” as we say, with a particular bronze sculpture on sale, and feels that he must have it in his collection, even though, being of the wrong period and of doubtful provenance, it adds nothing to the distinction or value of his collection. The intensity of his desire is the reason that he would give for spending so much money and his justification also, and not, for instance, any further calculation or thought.

We surely all have such immediate experiences, such enchantments, even if they are more often concerned with food, drink, sexual desire, or with particular localities and memories than with bronzes. Such experiences are swept under the carpet in this book, because of the primacy in its argument of having a reason that can be anyone’s reason, and that therefore introduces the possibility of moral harmony and moral agreement, as well as “co-deliberation”—the process of arriving at agreement—potentially as wide as humanity itself.

My collector may certainly not subscribe to a universalizable principle that anyone is permitted to buy anything that he “falls in love” with. He makes no universalizable claim. His entire feeling is directed toward this object here and now, and this feeling is his only “justification.” But then the bedrock provided by reason is missing and therefore the possible moral harmony with others is missing. We know that desires and interests and emotional attitudes are unavoidably divisive, being tinged with personal histories and being fiercely peculiar, and are at best typical of a particular time and place. They may explain my conduct, but how can they justify it?

Rather than deny the known and actual power of the diversity of sentiments among men and women, it would surely be better to look elsewhere for the universal aspect of morality, the bedrock: in the procedures that unavoidably arise from conflicts of sentiment in individual minds and in society. In The Republic, Plato suggested that the need for justice arises from an individual’s experience of inner conflict and that morality enters as a negative force which prohibits unworthy desires. With their domain extended to social conflicts, justice and morality retain their essentially negative character; they act, according to Plato, as a shield against the disruption and the chaos of uncontrolled conflicts in the city. For the tyranny of Plato’s philosophic wisdom we should substitute the fairness in public argument which always hears both sides in adversary reasoning before deciding between them. In all concerns about what we owe to others it is just and reasonable to be open to both sides in a conflict and to balance conflicting moral claims against each other.

What Scanlon insists we owe to one another is a comparatively small, fenced-off part of the vast domain of morality, and it seems to presuppose a fairly stable social system into which my station and its duties, in the old Victorian phrase, can be intelligibly fitted. What Scanlon presents is a personal morality within an established civil society. But this fencing off, which is convenient for moral theory, will almost certainly lead to a faulty description of morality when larger public and political issues arise and when the great evils of human experience have to be discussed. Toward the end of his book Scanlon writes:

There are many questions of right and wrong that seem clearly to have correct answers. There is no doubt, for example, that murder, rape, torture, and slavery are wrong. No system of rules could be a system that people had reason to accept as an ultimate, normally overriding standard of conduct if it permitted these practices.

This seems to me true and important, but it is surprising that he includes slavery in his short list of the great, undeniable evils, in a chapter designed to refute the cruder forms of relativism and of skepticism about moral values. Slavery, unlike murder and rape, is exactly the evil that relativists love to cite as being perceived as evil and wrong only after sensibility changed; it was not generally perceived as a moral evil in any of the centuries before the ages of Jefferson and Lincoln. Slavery and torture were generally regarded as evils only in the sense that physical weakness and disease are evils; they were in past times evils that arise from the normal flow of things in the human world, not undeniable offenses against moral principles. (Machiavelli in a letter writes about being tortured himself without much vividness, as if it were a common illness, deplorable but not monstrous or unnatural.)

Was the change in general attitudes toward slavery a change in sensibility, as I contend, or is it better characterized as having been a change in the reasoned principles of conduct that constitute an overriding standard of conduct within our communities, as Scanlon contends? This is the eighteenth-century battle between the Sentimentalists, of whom Hume and Adam Smith were, and remain, the leaders, and the reasoners, led finally by Kant. The Sentimentalists can claim that the great man-made evils, slavery among them, needed to be directly perceived and felt as evil before the reasons were found to explain and justify the view of them as evil. As an individual person, I am disgusted by the thought of owning another person, as in chattel slavery, and I ask myself why I am disgusted. Perhaps the disgust is superstitious, a mere taboo, and has no considered judgment and argument to support it.

A possible analogy: I may find some sexual practices repugnant, but on reflection I decide that there is no argumentatively supported reason for considering them evil or wrong, and that this repugnance is a matter only of my particular sexual temperament. I will therefore have no justification if I try to have these practices prohibited. So I reflect and ask myself what precisely is truly repugnant and shocking in the ownership of a person. Suppose that my reflection does not lead me to a generalizable reason or principle, as in the sexual case, but that I am still left with my original perception and feeling as strong as before. It is as if I cannot get behind my feeling and cannot provide it with a respectable derivation. I am satisfied after reflection that it is not a subjective response, but that it arises directly from the nature of the object under consideration. Of course I can further explain and amplify my feeling, and assign it a place among sentiments that you may perhaps share.

In these circumstances I think it would be a mistake to try to justify to myself my perceptions and feelings with some general principles about freedom and the autonomy of the individual. In reflecting, I realize that it is the highly specific relationship between master and chattel slave, with its detailed implications and consequences in actual life, that I find morally repugnant and that Jefferson probably did not perceive in the same way. Abstract principles of freedom of the individual and autonomy are introduced later as rationalizations of one’s repugnance toward the countless humiliations and injustices and deprivations that are involved in the daily life of a slave. It is the concrete details of the master-slave way of life that are appalling.

Many of the undeniable and horrible man-made evils are not natural subjects for consensual deliberation and argument, unlike the duties and obligations involved in what we owe to other people in a civilized community. It is true that my feelings, including my moral feelings of repugnance, are naturally divisive and will probably lead to conflict. I know that styles of domination and obedience that appall me are felt to be natural and acceptable by others, while forms of irreverence and impiety that are found repugnant by others seem entirely natural and acceptable to me. I stand by my feelings, just as firmly as I stand by the principled and community-endorsed moral judgments that Scanlon endorses. I am as much the man who strongly feels so-and-so as the man who clearly thinks so-and-so.

Child labor in the mines and factories of Britain in the 1840s did not appall many of Mr. Gladstone’s contemporaries or strike them as being evil, and the conditions of child labor in India and elsewhere at present do not appall many of our contemporaries. They accept the facts as the natural way of the world, and perhaps they feel more strongly about the evil of state interference with the workings of a free market. There are reasonable persons who are horrified by abortion as I am horrified by some conditions of child labor. When it comes to the public arguments, which it surely will in a democracy, my political enemies will deploy a variety of reasons to explain and support their revulsion, and I will do the same for my revulsion.

If we are candid, we will not pretend that our reasons are such that no reasonable person will reject them. We know, or ought to know, that persons who are reasonable in the normal sense of the word, in the sense that they can evidently follow and evaluate an argument, will remain unconvinced by the arguments, because their moral conclusion emerges, in these particular cases, from their moral feelings or passions, and not principally from the force of arguments. Evidently there are moral and political conflicts in which I am swayed entirely to one side or the other by the force of argument alone. But many of the deepest and strongest moral convictions are not formed in harmony with a community that shares them. They are more likely to be formed in passionate opposition to some practices and attitudes prevailing in the encircling society in which I was brought up.

Morality naturally arises from prohibitions and always keeps some of its negative character amid the chaos of competing moral ideals. But there is a possible resting point in a potentially shared ideal of rationality, as Scanlon and philosophers generally have hoped. I suggest that rationality in this context is justice and fairness in the procedures that resolve the conflicts caused by conflicting moral ideals. Procedural justice in conflict resolution is, in Scanlon’s terminology, an impersonal value that everyone, whatever his positive moral ideals, may recognize and respect. It embodies the principle of hearing both sides of a disputed case, a principle that we are all bound to follow in our private reasonings.

If we add to this “the principle of established practices,” as Scanlon names it, requiring that public procedures of conflict resolution should always be customary, we have reached a universal, minimum morality, a bedrock of fairness. But this can only happen if it is conceded that the motives that justifiably guide conduct are often not principles that no reasonable person can reject, but rather passions that many reasonable men do not share. Consequently we shall always find ourselves caught up in conflicts and oppositions, of argument and counterargument, of devotees of a religion versus skeptics and naturalists, of right versus left, and so forth. Scanlon’s “inner core” of morality, within which an implied contract between reasonable persons determines rights and duties in a civilized community, is always under threat and at risk from countervailing passions.

Because of his use of the concept of reasonableness, Scanlon must deny that reason is, and ought always to be, the slave of passions, a saying that seems to me both subtle and supremely right. If emancipated, reason (our powers of argument) runs free with its calculations without the constraints of natural feelings of sympathy; it will construct theologies and theories of religious and racial superiority that have for centuries engendered religious and tribal and national wars. Humanity is only restored to itself when such theories are undermined and finally put aside.

I conclude that What We Owe to Each Other characterizes too small a part of morality to be taken as representing a microcosm, which is what Scanlon hopes it can be. I think it presents a too regulated and too peaceful picture of our moral conflicts, in the soul and in the community, both of how conflicts are and how they ought to be resolved. It represents wrongdoing principally and typically as lapses from reasonable standards of integrity and fair dealing. But speaking expansively, and with less than Scanlon’s exactness, must one not say that history is principally a record of “undeniable evils,” of hate-filled killings, of invasions and destructions, and of a merciless lust for domination? Scanlon’s book is a very clear and careful reconstruction of the possible moral foundations of a settled society, or of a society that believes itself settled; but I think most of the more contentious and painful problems are beyond its scope: in particular politics. Politics has always been a running battle between groups of moral friends and their moral enemies, a type of moral conflict that has to be bypassed in Scanlon’s scheme.

This Issue

April 22, 1999