T. M. (“Tim”) Scanlon is best known for his advocacy, especially in his book What We Owe to Each Other (1998), of the moral theory of contractualism. Contractualism is broadly the idea that morality is based on a social agreement or “contract.” It can in principle refer to any contract-based moral theory, within a certain range to be described in a moment, but in practice it refers to Scanlon’s theory unless the context makes clear that something else is meant. The Stanford Encyclopedia of Philosophy (SEP) article on “Contractualism” by Elizabeth Ashford and Tim Mulgan says that contractualism is distinguished from contractarianism by being grounded in the equal moral status of persons. Contractarianism, especially of the sort identified with Hobbes and Gauthier and Buchanan, tries to derive morality from an agreement that individuals make based on their own self-interest. A contractarian theory imagines people forming an agreement which each sees as maximizing his own personal self-interest and nothing else, and in particular without regard to the interests of anyone else. By contrast, contractualism imagines that people are deciding mutually agreed-upon principles from a position in which each person accepts every other person as a rational autonomous agent of equal moral importance with himself. Scanlon claims (5) that this conception of the social contract can be traced back to Rousseau.

The key concept of contractualism seems to be justifiability to others. In what follows, I shall explain what “justifiability to others” means in Scanlon’s contractualism and why I think it lies at the root of a serious deficiency of the theory.

As a first pass, “justifiability to others” means that other people are treated by contractualism as being already moral agents to whose moral judgment one’s proposed principles of action must be submitted for approval or rejection and who are concerned—in the same way that you are—with finding moral principles acceptable to all (and thus agreed to in the “contract”). Scanlon writes:

My view … holds that thinking about right and wrong is, at the most basic level, thinking about what could be justified to others on grounds that they, if appropriately motivated, could not reasonably reject. On this view the idea of justifiability to others is taken to be basic in two ways. First, it is by thinking about what could be justified to others on grounds that they could not reasonably reject that we determine the shape of more specific moral notions such as murder or betrayal. Second, the idea that we have reason to avoid actions that could not be justified in this way accounts for the distinctive normative force of moral wrongness. (5; cf. 191)

There are several points to note here.

First, justifiability to others is constitutive of right and wrong. It’s not that an action is right, independently, and therefore it can be justified to others. Rather, it’s that an action can be justified to others and therefore it is right. Contractualism about right and wrong is thus like pragmatism about truth and falsehood.

Second, justifiability to others is an ideal standard; it is not a question of getting others to actually accept something (154–155). The reason, Scanlon says, is that “there are obviously cases in which acting morally requires one to resist the prevailing standard of what is and is not justified.” Thus, right and wrong are a matter of what an ideal rational agent would recognize as truly justifiable on the basis of the full set of correct applicable reasons. This goes with what frequently seems to be implied, if never explicitly stated by Scanlon, that reasons are objective. Reasons, he says, are something we recognize, as though they were “there anyway,” not products of our subjectivity. For example, he upholds what he calls the “universality of reason judgments” (73–74), which is that if you judge that in light of a given set of factors G, you have reason to Φ, then you must judge that anyone else who stands in the same relation to G has the same reason to Φ. This is because reason-giving grounds are reason-giving grounds, the same for everyone. (Note that besides being objective, reasons are also primitive, according to Scanlon. That is, reasons ultimately are not explained by anything further; they simply exist. We have discussed this point previously.)

Third, Scanlon says “reasonably reject” rather than rationally reject specifically to set his view apart from contractarianism with its emphasis on rationality driven by the satisfaction of our (usually selfish) ends. “Reasonableness” is a matter of having “reasons,” which as we have noted are primitive and have nothing necessarily to do with achieving something of value.

Fourth, this scheme is supposed to account for our motivation to be moral (for “the distinctive normative force of moral wrongness”). We care about morality and are concerned to stay within its bounds because we “have reason to want to act in ways that could be justified to others,” and when you recognize something as a reason, you are thereby motivated to act on it (154). The “reason” he claims that we have to want to act in ways that could be justified to others is expressed in a phrase Scanlon quotes from Mill: “the importance for us of being ‘in unity with our fellow creatures’” (163). He emphasizes that the unity he’s talking about does not derive from a psychological need for togetherness or an emotion of fellow-feeling but from seeing the “reason” we have to live together with other rational creatures in a community structured by reasons. “[W]hat is particularly moving about charges of injustice and immorality is their implication for our relations with others, our sense of justifiability to or estrangement from them” (163). (For Scanlon’s most concentrated discussion of these issues, see chapter 4, sections 3–5, pp. 153–168.)

In several places, Scanlon cites what he sees as the phenomenological accuracy of his account of moral reasoning as a major point in its favor (e.g., 155, 187). And I have to say that the theory does strike me as phenomenologically accurate. When I deal with people and ask myself whether some course of action I am contemplating is okay, I do ask myself whether my action could be justified to those who would be affected. I ask whether they could reasonably object to my action. I’m tempted to say that I ask whether my action “would piss them off,” except that it’s not quite that simple, because they might be unreasonably pissed off, but that should not deter me. The question is whether my proposed action is defensible, including being defensible on grounds that they should accept if they are being at all reasonable—or, as Scanlon says, on grounds that they could not reasonably reject.

Scanlon thinks we are motivated to care about justifiability to others because it is fundamental to maintaining our relationships with them. What is the source of our sense of obligation to comply with the demands of morality? To answer this question, we need to know (a) what are the reasons that justify those demands, and (b) why we should care about those reasons. For example, suppose I have promised a friend not to reveal her secret of having an extramarital affair. It is not normally hard to keep such a promise, but what special conditions or circumstances would justify breaking it? Or again, suppose I see a golden opportunity to embezzle money from my company. What are the reasons why I shouldn’t? To be adequately motivating, we need adequate reasons—as it were, reasons that seem like genuine reasons—and “it would be wrong” doesn’t qualify. That is, although it may indeed be wrong to embezzle the money or betray my friend’s confidence, and although I may already be disposed to find morality strongly motivating, merely to repeat that an action would be wrong is not explanatory. Again, we want to know (a) why it would be wrong and (b) why I should care about whatever considerations provide the answer to (a). But this is a difficult pair of questions to answer, because reasons that seem adequately motivating usually do not seem to be moral reasons. For example, citing my self-interest—saying that if I break my promise, I might lose my reputation as trustworthy and a good friend, or that if I get caught embezzling, I might go to prison—might be adequately motivating, but doesn’t seem to have that much to do with why these actions would be wrong. On the other hand, citing the harm these actions would do to others might be appropriately related to their wrongness, but not very strongly motivating. Scanlon calls this problem Prichard’s dilemma (after H.A. Prichard, who describes a similar problem in his well-known essay, “Does Moral Philosophy Rest on a Mistake?,” 1912).

Scanlon’s answer to Prichard’s dilemma is his central principle of justifiability to others. He claims that, whereas in thinking about the civil and criminal law,

the reasons that guide us in thinking about what the law should be are commonly very different from the “sanction” that moves us to obey it (whether this is fear of punishment or a sense of obligation). In the case of the morality of right and wrong, however, these two kinds of reasons flow from the same more general reason: the reason we have to live with others on terms they could not reasonably reject… Because we have this reason we have reason to attend to the question of which actions are right and wrong, that is, to try to determine what would be allowed by principles that others could not reasonably reject, and we also have reason to govern our practical thought and our conduct in the ways that these principles require. (153–154)

That is, because we are powerfully motivated to live with others on terms that are justifiable to them, we have reason (a) to ask what those terms are and (b) to comply with those terms. And, since to ask what are the justifiable terms is to ask what is right and what is wrong, we see that this (a) and (b) correspond exactly to the previous (a) and (b), and that the same fundamental principle—justifiability to others—supplies the answers to both questions.

For example, when I contemplate divulging my friend’s secret despite my promise not to, I contemplate whether I could justify to her my doing so. I consider the strenuous objections she would make, such as from the harm done to her and from my misleading her with a false promise. Can I respond to these objections by appeal to any principles that she could not reasonably reject, such as, perhaps, my friendship with her husband, or the importance of some new, unforeseen circumstances, or my need to maintain my own integrity in answering questions about her from third parties? Although these reasons are rooted in our competing values and interests, which are not particularly moral in themselves, they are as it were moralized by my acceptance of the constraint that I act only on principles that no one could reasonably reject. Indeed, this constraint is very strong, effectively giving everyone in the world veto power over my actions! The veto power is not arbitrary, however: it can only be exercised by a person who can reasonably reject the principles on which I act, and we have seen that that implies an ideal of reasonableness grounded in what are taken to be objective reasons. Still, the scope of these reasons is very broad and includes all of any person’s legitimate interests, which I may not violate unless they are overruled by competing interests of sufficient weight. Finally, with regard to motivation, what makes me ask whether the people affected by my actions could reasonably reject the principles of those actions is the reason I have to live on justifiable terms with others (specifically, on terms which they cannot reasonably reject). And that is the same reason that motivates me to actually comply with those terms.

Scanlon’s account obviously places enormous weight on the supposed reason we have “to live with others on terms that they could not reasonably reject.” I wonder whether it is true that (i) we have such a reason, (ii) it has the weight Scanlon assigns it, and (iii) this is really the proper focus for moral theory. Part of what this strategy does is to replace what otherwise might be a focus on human well-being or on preference satisfaction or on personal autonomy or on some sort of collective or community interest with a focus on what general principles human beings as individuals have reason to accept. The Kantian flavor of this strategy is obvious, although the way moral reasoning on Scanlon’s view incorporates the particular interests and preferences of individuals is a distinctly non-Kantian element. As Scanlon says, his account of right and wrong, is “in Kantian terms, avowedly heteronomous” (6).

However, rather than directly question “justifiability to others” as opposed to other proposed foundational principles of morality, I want to focus on the problem of circularity in Scanlon’s theory. The basic problem is this: if justifiability to others is the fundamental standard of morality, then what are the grounds of justifiability? Reasons, of course. What sort of reasons? Moral reasons, apparently. But if so, then how have we explained anything about morality? The theory seems to say no more than that an action is morally right if it can be justified to others as right on moral grounds. That is not illuminating!

Colin McGinn, in his highly critical review of the book, puts the criticism this way:

The first and most pressing problem with Scanlon’s theory is that it is circular. It presupposes what it is meant to explain, in the most obvious way. An action is said to be wrong if it may be criticized by others. But on what grounds is it criticizable? On moral grounds, surely. But then aren’t we back where we began? An action is wrong if it can correctly be said to be wrong. But weren’t we being told what it means to say that an action is wrong? Circle! (35)

Initially, this objection may seem unfair. For, it may seem that a certain moral structure has been imposed on the process of reflecting about right and wrong by the contractualist framework already, so that the reasons being assessed within that framework need not be themselves moral reasons. As an example, consider how a contractualist might conclude that murder is wrong. Thus, suppose I decide to murder you because you annoy me. The principle under consideration would be something like, “if A is annoying to B, then B may kill A to remove the annoyance.” You would naturally have the utmost strenuous objection to this principle, on the ground that you have a maximal interest in not dying. And I would have to accept that your rejection of the principle is reasonable, since, although my annoyance may be considerable, it is not nearly as bad as dying, and after all there are other remedies for annoyances. Therefore, murder would be wrong in this case and also in a very wide range of similar cases to which the argument could be straightforwardly extended. So, in general, murder is wrong according to contractualism. Now, it is hard to see how this argument has helped itself to a pre-existing moral rule against murder. Rather, the argument weighs the agents’ competing claims based on their largely material interests (namely, not being annoyed versus not being killed). Thus, the moral character of the argument enters not through a pre-existing moral status of these interests, but through the contractualist framework that insists that I cannot perform any action if the persons affected (as individuals) could reasonably reject the principle of that action on grounds of how they would be affected. The contractualist framework proposes that people in a community may act only on principles justifiable to 100% of the individual members of that community. Each member of the community must therefore take the interests of the other members very seriously indeed, and this is what puts the morality in contractualist moral reflection.

The trouble is that many judgments about right and wrong are not so straightforward as the case of murder over annoyance. Consider, for example, the question of helping people who are in need. Despite the vast reduction in the number of people living in “extreme poverty” (defined as living on less than $1.90 per day) worldwide during the past three decades, there are still well over half a billion people still living in that condition. And of course there is tremendous suffering and want in the world even where it does not reach that level. It would seem that contractualism clearly leads to some sort of moral rule dictating that people who are better off must provide aid to those who are in need. For, a person living in extreme poverty would reasonably reject any principle that allows a person who is well off to spend his money on nice clothes and an iPhone and $50 dinners and spend nothing on the person who is living on <$1.90 per day. And that is only the beginning. Every person has a compelling interest in at least the necessary components of their own well-being, so that they would reasonably reject any principle denying them aid from those whose well-being is secure in its necessary components. Thus, the SEP article on contractualism says, “It certainly seems possible that contractualism will generate very demanding principles, as it seems reasonable for those who are starving to reject any principle permitting me to retain inessential resources rather than meeting their most basic needs” (33). Scanlon himself attempts to defeat such demanding principles through a proviso that “the threshold of sacrifice is understood to take account of previous contributions (so that the principle does not demand unlimited sacrifice…).” He writes,

It would… be reasonable to reject a principle that required us, in every decision we make, to give no more weight to our own interests than to the similar interests of others. From an agent’s standpoint such a principle would be intolerably intrusive. (224)

This does seem reasonable. And yet we seem to hear the poor saying, “I’m sorry if my starving ‘intolerably intrudes’ on your other plans!”

And so we are caught in a familiar moral controversy between those who say that “if it is in our power to prevent something bad from happening [sc., someone starving], without thereby sacrificing something of comparable moral importance [sc., oneself not starving], we ought, morally, to do it” (Singer, “Famine, Affluence, and Morality,” 1972) and those who say that it is morally acceptable to give “more weight to our own interests than to the similar interests of others.” The point is that it doesn’t seem that contractualism has contributed much to this familiar debate. The contractualists themselves are divided on the question (see the SEP article, section 9). The contractualist method of adjudicating the matter, as in all cases, is to consider what reasons the various parties have for accepting or rejecting competing principles of behavior from their various standpoints. Do people in fact have reasons to prefer themselves and their families, say, to others in deciding where to devote their resources? Surely they do to some extent; but how much? Does proximity matter, so that a person in need on your doorstep is different from one who is on the other side of the planet? Is there a minimum of “essential well-being” beyond which further aid is no longer morally obligatory? How is that level to be determined? Should the level be relative to community standards (e.g., higher in developed than in underdeveloped countries)? Which of the many human goods should be considered “necessary”? Does putting the question purely in terms of giving aid commit the fallacy of zero-sum thinking? Or does putting the question purely in terms of giving aid distract from the more fundamental structural injustice embedded in the very institutions of private property and global capitalism?

Does contractualism have distinctive answers to any of these questions? Or any way of settling the larger controversy over the limits of a duty of benevolence? As noted, contractualism respects individuals (since a single person’s reasonable rejection vetoes any proposed principle of behavior) and recognizes values other than well-being (such as respect); and contractualism looks for general principles of behavior that apply in the same way to everyone; but this is all the structure contractualism imposes on morals. So, it is not utilitarianism. That’s not nothing. But it’s not much, either. Beyond that, contractualism looks for the “reasons” that individuals have for accepting or rejecting principles of behavior. But it provides no assistance in determining what those reasons are or how competing reasons are to be adjudicated.

Probably, this is no accident. By making justifiability to others primary and independent of any prior standard, contractualism crowds out the possibility of any such standard. If there were any master value in terms of which to assess reasons and adjudicate competing principles, such as human flourishing, justifiability to others would be knocked off its pedestal. To be sure, moral principles would still be “justifiable to others.” But what this would mean is that they could be shown to be appropriate by the standard of human flourishing (or whatever). Justifiability to others would become a byproduct of meeting the standard, not the standard itself. But then, contractualism would no longer be contractualism.

Formalistic (Kantian) approaches to moral theory in general seem to have a problem supplying content for moral rules. And the reason seems clear: it is that they attempt to get by without a specific conception of human nature. That is what makes them formalistic. Here we see this problem biting contractualism. If this is right, it would be a reason to move in a more broadly Aristotelian direction in moral theory.