PART THREE
[A]lthough the commands of the Conscience are essentially absolute, they are historically limitary. Wisdom does not seek a literal rectitude, but an useful, that is, a conditioned one, such as the faculties of man and the constitution of things will warrant. … [T]he existing world is not a dream, and cannot with impunity be treated as a dream; neither is it a disease; but it is the ground on which you stand, it is the mother of whom you were born. Reform converses with possibilities, perchance with impossibilities; but here is sacred fact. This also was true, or it could not be: it had life in it, or it could not have existed; it has life in it, or it could not continue.
—Emerson
Chapter 8
THE ARGUMENT of part 2 has proceeded entirely by way of immanent criticism. I have closely examined the liberalism of Rawls and Rorty and the traditionalism of Milbank, MacIntyre, and Hauerwas, and found both of these approaches unable to provide a satisfactory account of the role of religious traditions in modern democracy. The task of part 3 is to work out an acceptable alternative to these approaches. A successful philosophical account of democratic political culture would overcome the weaknesses I have identified in these approaches while simultaneously inheriting their strengths. It would thus put us in a position to explain both the strengths and the weaknesses of the previous approaches. Liberalism and traditionalism thrive on exposing one another’s weaknesses. Liberals have no trouble showing that traditionalism threatens to deprive us of the actual and potential benefits of exchanging reasons across the boundaries of enclaves. Traditionalists have no trouble showing that liberalism has failed to resolve a conflict between its commitment to freedom and its desire to dictate the terms of social cooperation. Both sides have gotten rather good at biting their respective bullets and at making their opponents seem despicable or foolish. But neither side has had much success at explaining the other side’s strengths. Nor have critical theorists like Benhabib done any better on the crucial issue of religion.
There are, however, some hopeful signs that suggest a path beyond the current impasse. We have seen that Rawls, in an attempt to solve problems in a social-contract theory inspired by Locke, Rousseau, and Kant has introduced concepts derived from the pragmatic expressivism of Hegel and Dewey. The conception of conversation that Rorty develops in Philosophy and the Mirror of Nature—but then forgets about when discussing the role of religion in political culture—is also a product of pragmatic expressivism. It is this pragmatic expressivism that MacIntyre has in mind when he acknowledges that Rawls and Rorty have recently recognized that liberalism must be understood as a tradition. And Benhabib’s critical theory is also indebted in various ways to the Hegelian strand of expressivism.
As for the traditionalists, Milbank openly avows a form of pragmatic expressivism influenced by Hegel and by various postmodernists. Both MacIntyre and Hauerwas stress the priority of social practices. While neither of them responds successfully to Hegelian criticisms of their positions, both have unacknowledged debts to Hegel that place them in the vicinity of pragmatic expressivism. MacIntyre wavers between two conceptions of tradition. One of these conceptions begs the question in favor of traditionalism by assuming that traditions require high levels of agreement on canonical authorities and the nature of the good. The other defines a tradition in Hegelian fashion, as a dialectical argument over goods and virtues in the context of shared social practices that endure over time. In both MacIntyre’s and Hauerwas’s writings, we can catch glimpses of how this second conception of tradition and the related notions of virtue and example can be creatively applied to contemporary society. MacIntyre sometimes makes modern ethical discourse seem hopeless, but he also explicates a notion of rationality that involves something like conversation in Rorty’s sense. In Hauerwas’s insistence that he is not a sectarian and in his praise for Dorothy Day, one can sense the possibility of a democratically engaged ethics of virtue that has been cleansed of antiliberal resentment. The same can be said for Milbank’s brief references to Christian socialism.
The way beyond the current impasse, it seems to me, is to see both sides as converging on a form of pragmatic expressivism that takes enduring democratic social practices as a tradition with which we have good reasons to identify. By working out a defensible version of pragmatism, I hope to explain the strengths of liberalism and traditionalism, as well as their weaknesses. This means:
• resolving the internal tensions in Rawls’s political liberalism by discarding his notion of a freestanding conception of justice and his loaded account of reasonableness, while retaining the idea that we owe reasons to one another when we take stands on important political questions;
• seeing our overlapping consensus on the legitimacy of constitutional democracy as involving a practical commitment to holding one another mutually responsible for our political arrangements and thus to keeping a democratic discussion going across the boundaries of ethnic, racial, and religious enclaves;
• working out an account of our discursive practices that takes seriously Rorty’s ideal of conversation and MacIntyre’s emphasis on the need to understand one’s rivals on their own terms, while jettisoning the quest for principles that no reasonable person could reasonably reject;
• developing a conception of tradition centered on enduring social practices, without begging the question in favor of traditions that are hierarchically organized and insistent on doctrinal conformity as a criterion of membership;
• and redescribing modern democracy as a tradition of this looser sort by focusing attention primarily on the relatively freewheeling discursive practices it involves.
Splitting the difference is a tried and true strategy for overcoming dialectical impasse. Hegel and Dewey were both master practitioners of the art. Indeed, I believe Hegel took himself to be splitting the difference between earlier versions of liberalism and traditionalism, and Dewey self-consciously followed in his footsteps. So we are not exactly moving into uncharted territory. Moreover, this strategy puts me in close proximity to a number of my contemporaries: Robert Brandom, Sabina Lovibond, and Cheryl Misak in philosophy, Nancy Fraser in political theory, and Rebecca Chopp and Cornel West in religious studies, to name only a few obvious examples.1
The present chapter begins this more strictly philosophical part of the book by trying to explain the sense in which democratic norms are creatures of the practices of accountability in which we exchange reasons with one another. For purposes of illustration—but also with an eye toward contemporary political relevance—I consider a familiar dilemma known to philosophers as the problem of dirty hands. This is the question of whether political officials responsible for protecting the public from attack may, under conditions of emergency, do things that are commonly thought to be not only wrong but horrible. The problem is philosophically interesting because it forces us to clarify what norms of responsibility are, to what kinds of responsibilities they pertain, and how responsibilities of different kinds can conflict. The timeliness of the problem derives from its obvious relevance to the struggle against terrorism. My aim here is not to solve the problem, but to improve our understanding of what the problem is and why it resists a solution that would be likely to command assent from the entire citizenry. In the process, I hope to clarify certain features of the version of pragmatism I favor and to shed light on the limits of moral consensus in a society as religiously divided as ours.
BEWARE OF LEADERS PLEADING NECESSITY
Political officials everywhere do bad things. One mark of tyrants is their interest in doing horrible things, such as torturing political prisoners or murdering civilians, for the purpose of making opponents fear the unscrupulous exercise of power. The point of the exercise is to impress people with their unwillingness to be constrained, in the exercise of power, by moral considerations. In a representative democracy, however, we expect political officials to answer to the citizenry for the bad things they do. They are acting in our name, as our representatives, with the consent of the governed. So we demand reasons for their actions, not least for their bad actions. The practice of asking for reasons is an exercise in scrupulosity. The reasons we demand imply or impose constraints. When our officials do bad things, they are often acting in secret, hoping to avoid public scrutiny. They are not trying to impress us with their lack of scruples; they are trying to hide from us because they do not have justifying reasons to offer. They would prefer to duck our questions. They stonewall; they deny involvement.
When their dirty hands come to light, and there is no denying the bad things they have done, officials often plead necessity. They claim that they had to do bad things if they were to serve us well in dire circumstances. There was no choice. That, they say, is the way politics works in the real world. At one level, the problem of dirty hands is the question of how democratic citizens ought to respond when political officials make this excuse for admittedly bad acts.
Often the excuse turns out to be phony—another bad act. The circumstances are not what the dirty politician says they were. Perhaps they were not dire at all. Or perhaps they were dire relative to some set of personal concerns, like the desire to remain in power, but not relative to the citizenry’s survival and well-being. In most cases where people plead necessity in ordinary life, there is more room for maneuver than they acknowledge. They say, “I had no choice,” but what they really mean is that the alternatives they did consider seemed unacceptable at the time.2 Often there are other alternatives that should have been considered, and it is not unusual for people to be mistaken about the unacceptability of the alternatives they do consider. The same is true in the political arena. The necessity excuse almost always turns out to be false.
Even when it is false, it often works. The desire to make it work therefore gives politicians an incentive to create a sense of emergency among the citizenry. Leaders who tend to act badly tend also to play upon our fears. They portray neighboring nations and domestic minorities as threats. Sometimes they even deliberately provoke their opponents to behave horribly, hoping then to be excused for responding in kind. The stronger the sense of emergency the leader can arouse, the more plausible the necessity excuse becomes. The more plausible the necessity excuse becomes, the greater the leader’s temptation will be to act badly. The more scope we give our leaders to act badly by accepting the necessity excuse, the closer we move to unconstrained rule by the holders of power. The closer we move to unconstrained rule, the less reason some citizens will have to trust officials to protect them or otherwise serve their interests. In the absence of such trust, a democratic culture tends to dissolve. The necessity excuse belongs to a vicious cycle that has been detrimental to democracy in contexts as different as the final phase of the war against Japan, nuclear deterrence during the Cold War, the Balkans in the 1990s, the Israeli-Palestinian conflict, and the global struggle against terrorism. Politicians who plead necessity are usually enemies of democracy. They are not to be trusted.
But some leaders do find themselves in real emergencies. I grant that in some such situations conscientious leaders committed to maintaining the culture of trust essential to democracy face genuinely hard choices through no fault of their own. Even if they have surveyed all of the alternatives responsibly and understood the facts accurately, they might still find themselves having to choose between doing something bad to someone and allowing something bad to happen to the people in their charge. How shall we respond to choices like this?
One familiar answer is a simple form of consequentialism: everyone should always act so as to bring about the best consequences they can; hence, leaders should do so in cases like this. If the consequences of allowing something bad to happen to the people would be worse, on the whole, than the consequences of doing something bad to someone, then one ought to go ahead and do the bad thing, knowing that the end justifies the means. The principle invoked in this answer is too simple, however, to square with the rights characteristically recognized in democratic constitutions or with the claim of some prominent democratic thinkers that cruelty is the worst thing people do.3 In practice we treat these considerations as constraints on the means leaders may select when deciding how to pursue their legitimate ends. Simple consequentialism leaves these constraints out of the picture. That is why it makes the problem of dirty hands seem easy to solve—too easy, from what I would want to call a democratic point of view. The next section offers an anatomy of the problem that puts these constraints back in the picture. The anatomy is carried out in terms borrowed from Robert Brandom’s pragmatic analysis of practical reasoning in Making It Explicit.
AN ANATOMY OF THE PROBLEM
One way to describe the problem of dirty hands is to speak of a dilemma created by conflicting desires. The first horn is the desire to protect the survival or well-being of the people. The metaphor “dirty hands” suggests that the second horn is concerned with the leader’s moral purity. Other things being equal, decent political officials would like to keep their hands clean, and we would prefer that they do so as well. But when the desire for the official’s moral purity comes into conflict with the desire for the people’s survival and well-being, how can we reasonably prefer the former to the latter? Wanting to keep one’s hands clean sounds selfish when compared with the desire to protect the people. It seems like the kind of desire one should desire not to be motivated by in such a situation. We do not want officials to concern themselves with their own purity at the people’s expense. What we want, all things considered, are political leaders who are neither unscrupulous nor squeamish. Democratic leadership is no place for tyrants, but it is also no place for purists.
While this way of posing the problem dovetails nicely with the metaphor of dirty hands, it is not clear that the second horn should be interpreted as a desire for purity. If the problem were really about the conflict between the desire for purity and the desire for such goods as social survival, then most of us would not see it as a perplexing matter at all. Aside from a few genuine purists on the fringes, those of us who want to place rigid constraints on our political representatives are more interested in the victims than in the stain left on the hands of the official who does the bad thing. The value we have in mind is not moral purity but the importance, worth, or sanctity we attribute to individual human beings. I return to this thought later in the chapter.
Whatever we make of the point about purity, I doubt that the problem of dirty hands is best described as a conflict between desires. To see why, consider Brandom’s distinction among three typical patterns of practical inference, which I will illustrate with examples modeled on his:
(a) Going to the store is my only way to get milk for my cereal, so I shall go to the store.
(b) I am a lifeguard on the job, so I shall keep close watch over the swimmers under my protection.
(c) Ridiculing a child for his limp would humiliate him needlessly, so I shall refrain from doing so.4
According to Brandom, each of the three examples is a material inference. To make any of them valid by virtue of its logical form, we would have to add a premise, but in another sense they are proper exactly as they stand. For they are materially good, in the sense that all competent participants in our evaluative practices treat such inferences as proper (and are not making a mistake by doing so). By taking these inferences as (materially) proper, we acknowledge the premises of these practical inferences as reasons for their conclusions.
Emphasizing that each of these inferences is (materially) in order as it stands is not a way of denying significance to the premises we would need to add to them to make them formally valid. It is rather a way of making clear what that significance consists in (MIE, 246–47). Suppose I added to (a) a statement expressing my desire to have milk for my cereal; to (b) the conditional that if I am a lifeguard, it is my responsibility to keep a close watch over the swimmers under my protection; or to (c) the principle that one ought not to humiliate people needlessly. In each case, according to Brandom, I would thereby make explicit the material inferential commitment implicit in the original example of practical reasoning. This has the advantage of putting the formerly implicit material inferential commitment in the explicit form of a claim, which in turn allows it to be challenged or justified inferentially in light of other considerations. This becomes especially important, of course, when conflicts arise among different material inferential commitments that we have undertaken.
The problem of dirty hands is a classic instance of such a conflict. But which types of material inferential commitment are at issue? They do not seem to be of the sort made explicit in the language of desire. The kind of inferential commitment expressed in (a) is only one typical pattern to be found in our practical reasoning. If you treat (a) as a materially good inference for me to make, this involves taking entitlement to the commitment expressed in the premise to be inferentially heritable by the commitment expressed in the conclusion. If you also treat in the same way a battery of other similarly structured inferences to the same conclusion, you are implicitly ascribing to me the desire to have milk for my cereal (MIE, 249). Desire talk is one way in which we explicitly attribute material inferential commitments in practical contexts, but there are others.
The problem of dirty hands seems to have more to do with inferential commitments of the kinds expressed in (b) and (c) than with commitments of the kind expressed in (a). When made explicit, the first horn of the dilemma clearly involves the political official’s role-specific responsibility to protect the people. The second horn appears to be a role-independent responsibility to refrain from acts of a certain sort. Brandom refers to the explicit formulation of these two types of constraint, respectively, as “institutional” and “unconditional” obligations (MIE, 252).
Instrumentalist accounts of reasons for action follow Hume in assimilating both of these to the prudential “ought” associated with inferences of type (a). Kantians, on the other hand, tend to reduce all reasons for action to the third pattern. Brandom argues, however, that both Humeans and Kantians are too quick to reduce reasons for action to a single pattern.5 A nonreductive account of practical reasoning has the advantage of allowing us to offer a deeper, more illuminating account of moral perplexity in cases of genuinely conflicting considerations. In the present case, it allows us to work through our intuition that the problem of dirty hands is genuinely problematic, rather than brusquely explaining it away.
Let us begin by thinking in more detail about institutional responsibilities. Lifeguards have an institutional role. By virtue of that role they have a responsibility to keep a close eye on the swimmers under their protection. We implicitly ascribe this responsibility to them whenever we treat (b) as a commitment-preserving inference for a lifeguard to make. Notice that the propriety of my inference in (b) does not depend on my desire to keep a close eye on the swimmers under my protection. Nor does it depend on my desire to be, or remain, a lifeguard—to occupy the role that entails the responsibility. To be a lifeguard is to have responsibilities of this kind. When I acknowledge the responsibilities, I treat inferences of type (b) as providing me, qua lifeguard, with reasons for action that are independent of my desires.
We can make the inferential commitment in (b) explicit in the form of a claim about my responsibilities as a lifeguard. The language of responsibility allows us to challenge or justify such a claim as the need arises. This need might arise, for example, in the event that my responsibility to keep a close eye on the swimmers conflicts with my desire to leave the beach (say, to buy ice cream) while I am on duty. It is also possible, however, for two of the responsibilities a given lifeguard has, qua lifeguard, to come into conflict. A given lifeguard’s responsibilities, qua lifeguard, can also conflict with his or her responsibilities as a promise-keeper or as a spouse.
Heads of state hold an office that comes with unique responsibilities. If they were not responsible, in our eyes, for protecting the people from serious harm, then the problem of dirty hands they sometimes face would not capture our attention. If simple consequentialism were true, there would be no special problem of dirty hands pertaining to holders of high office as such. If the end of maximizing the balance of good over evil in the consequences of our acts entitled everyone, regardless of role, to do anything necessary in the pursuit of that end, we would not need to consider the political leader’s predicament as a morally distinctive kind of case. This problem merits its own chapter in the ethics of democracy, if it does, only because holders of high office are in some respects unique. They are unlike the rest of us in being officially responsible for the exercise of coercive power on behalf of the people’s survival and well-being. By virtue of their office they have powers and responsibilities the rest of us do not have. Their responsibilities can give rise to hard choices. These choices invite them to exercise power in ways that leave their hands dirty.
The most plausible appeals to necessity in such cases involve officials who are entitled to claim that they must do something bad if they are to fulfill the role-specific responsibility to protect the people’s survival. If the necessity excuse isn’t phony, and there really are no alternative means available, the question is whether officials thus situated are entitled to go ahead and fulfill the relevant responsibility, willing the bad thing as a necessary means toward that end. The problem is that by refraining from doing the bad thing, they know that this will leave the relevant responsibility unfulfilled.
Lifeguards are responsible for protecting the swimmers under their care from drowning. But suppose I am on duty in this capacity and notice a swimmer going down in violent surf for the third time. Suppose further that an old man in a wheelchair, through no fault of his own, is blocking my path on the jetty. I know that my only way to save the swimmer is to push the wheelchair aside with sufficient force to cause some harm to the old man. I have no doubt that this would be a bad thing to do to the old man. Yet if I refrain from doing it, I will be responsible for a drowning that it is my responsibility to prevent. What shall I do?
The case is artificial and abstract. Lifeguards do not confront cases of this kind very often. So while it does illustrate how a role-specific responsibility can come into conflict with another reason for action, our conception of a lifeguard’s responsibilities does not make special provisions for it. Quandaries of this sort, when they do arise, are normally left to a lifeguard’s discretion. A lifeguard caught in such a situation will be expected to assess the gravity of the harm likely to be inflicted on the old man if he is shoved out of the way and to weigh this against the life of the drowning swimmer. Especially if the harm to the old man is likely to be slight, the necessity excuse will be accepted. Some quandaries faced by political officials are like this. They involve the prospect of knowingly causing relatively slight harm to someone in unforeseeable, unusual circumstances where a judgment call needs to be made on a life-or-death matter. We are prepared to see the harm in such cases as regrettable but proportionate. We would not want to have either lifeguards or heads of state who were too squeamish to do this sort of thing when necessary.
Now consider another analogy. A Catholic priest has a role-specific responsibility to maintain the secrecy of the confessional. He is also responsible as a citizen or as a resident to provide information to the police that might lead to the apprehension of violent criminals. Suppose he has such information but refuses to provide it because he acquired it when hearing a confession. Catholics do not treat this kind of case as something priests have the discretion to resolve according to their assessment of the circumstances. Instead they have found it necessary to codify their conception of the priest’s role with maximal explicitness in canon law and moral theology. The Roman Catholic code takes the priest’s responsibility to hold confessions in confidence to be absolutely overriding with respect to other responsibilities. A priest who fails to fulfill this responsibility necessarily incurs guilt, according to the code. A priest who, in fulfilling this responsibility, withholds information that the police need if they are to prevent numerous murders, is not held responsible for those murders if and when they occur. He incurs no guilt for that consequence of his action, even if he has every reason to expect it to occur.6
It makes sense to give officials some discretionary authority to do bad things if the thing being done is bad only in the sense of wronging someone in a minor way and if it is a necessary means toward an end essential to the fulfillment of an important official responsibility. When, however, the wrong or harm would involve severe cruelty we are much more reluctant to give officials—or lifeguards—discretionary authority. When the bad things being done are horrible, it becomes more plausible to suppose that the second horn of the dilemma derives from an obligation that is not only independent of particular roles, but also absolutely binding. Political officials, it is said, are citizens and human beings before they assume their responsibilities as leaders. If there are indeed some unconditional obligations, they apply by definition to political leaders no less than they apply to the rest of us. And if some of these obligations override any role-specific responsibility with which they might conflict, then there are some things political leaders should never do, regardless of the circumstances. Nobody, even a politician, has license to do horrible things.
The Catholic code imposes an absolute constraint on priests to preserve the confidentiality of the confessional. It thus allows a priestly responsibility to override the responsibility all citizens have to give information to the police that might lead to the prevention of capital crimes. An analogous solution to the problem of dirty hands would treat the political official’s responsibility to protect the people as capable in some circumstances of overriding the responsibility to refrain from acts recognized as absolutely wrong when committed by ordinary people. The Catholic rationale for the code of priestly confidentiality is the sacramental status of the confessional in Catholic theology. No one guilty of capital crimes would have reason to confess such sins to a priest not bound by such a code. The most compelling democratic rationale for the parallel solution to the problem of dirty hands is the claim that in situations of supreme emergency, the worst thing a political official can do is to allow the people to perish at the hands of its enemies.7
It seems foolish to insist that a lie or a bribe would not be justified if that is what it takes for a leader to extricate her people from such grim circumstances. But what about acts often cited by prominent democratic thinkers as the worst sorts of thing human beings do to one another—acts involving torture or murder of innocent civilians? Can we not agree that we are all obliged to refrain from such acts, regardless of our roles and however severe the emergency of the moment may seem? If our answer to the latter question is yes, we are thereby treating the responsibility to avoid committing moral horrors as both unconditional and capable of overriding a political official’s other responsibilities. This is the possibility I wish to pursue in the remainder of this chapter. I will deal first with the issue of unconditional obligations and then with what it means for one “ought” to override another.
UNCONDITIONAL OBLIGATIONS AS EXPRESSIONS OF DEMOCRATIC CULTURE
An unconditional obligation, as I am using the term here, is universal in the sense that it is not role-specific or otherwise restricted to a specific group in its application. Some philosophers doubt that there are any such obligations. They see the belief in such obligations as an unwelcome by-product of what Hegel, in The Phenomenology of Spirit, denigrates as Kantian Moralität—the thinned-out, abstract point of view he contrasts with the “ethical life” or “substance” of prereflective custom, or Sittlichkeit. If I may transpose Hegel’s categories into Brandom’s terms, the prereflective ethical community treats material inferences of types (a) and (b) as proper but lacks the expressive resources to make these types of inferences explicit in the form of claims. Such a community also lacks inferential commitments of type (c), which is to say that it does not even implicitly acknowledge unconditional responsibilities. In Sittlichkeit no responsibilities are made explicit in the form of contestable claims, and whatever implicit responsibilities there are, are duties of one’s station, not the obligations of a human being per se.
The ethical life of such a community is constituted in part by the material-practical inferential commitments it treats as proper—that is, by something it does. What it cannot yet do in the stage of Sittlichkeit is reflect on the adequacy of these commitments, because it lacks the concepts—the expressive resources—required to make them explicit in the form of claims. Students of ethics have sometimes observed that before a certain date, one finds no expressions in ancient languages that have the sense of our expression “morally ought” and little or no mention of desires as reasons for action. This is what Brandom and Hegel would lead one to expect—namely, that practical reasoning begins in implicit acknowledgment of material inferential commitments and evolves in time toward their explicit articulation in a normative vocabulary suitable for critical reflection. Hegel argues that implicitly acknowledged conflict involving role-specific responsibilities drives practical reasoning toward explicitness. Catholic teaching on priestly confidentiality is but one instance of what can result from this drive. Other communities have responded to similar conflicts in other ways.
A crucial turning point in the Phenomenology is Hegel’s description of ancient Greek tragedy as a dramatic means of representing conflicts involving role-specific responsibilities. The norms of Sittlichkeit are implicit in the practical inferences of a people. They are ethically immediate in the sense that the practical reasoning transpiring under these circumstances is not mediated by explicit formulation of norms. Yet conflicts inevitably arise, as Sophoclean tragedies like Antigone make clear. Creon’s role as a political leader carries with it responsibility for the proper administration of the state. Antigone’s role as the sister of Polynices carries with it responsibility for burying his body after he dies. The two responsibilities come into tragic conflict when Creon, for reasons of state, orders that Polynices’ corpse be consigned to the dogs and the vultures. Sophocles puts the spectator in a position to see what generates the conflict, thus inviting reflection of a sort in which neither of the main characters engages.8 Reflection on such a spectacle allows an audience to infer that role-specific responsibilities can come into conflict not only for two people occupying different roles but for a single person who occupies roles in both the state and the family. Once such reflection gets going, however, it rapidly develops the expressive resources to allow agents to take responsibility, not only for their actions, but also for explicit claims about what their responsibilities are and claims about how those responsibilities should be conceived. These claims are well suited to play the role of reasons for action, but they are also inherently vulnerable to potential challenge insofar as reasons can be demanded for them. The norms that Creon and Antigone see as given are thus brought within the sphere of responsibility. For individuals can now be held responsible for treating some explicit formulation of these norms as a reason for action. To hold someone responsible in this way is to demand reasons for his or her explicitly acknowledged norms.9 In a modern democratic culture, this sort of demand is made every day.
Tragedy, law, rhetoric, logic, dialectic, and democracy all signal the creation of the expressive resources required for reflective discourse about norms. The critical distance from a culture’s normative inheritance that is made possible by these resources can also be experienced, however, as a form of estrangement from the roles and responsibilities in which one finds oneself. In its most extreme form this estrangement represents Moralität, the alienated condition in which moral reflection, once triggered by conflicts relating to socially recognized role-specific responsibilities, finds itself trying to make itself completely independent of the ethical life of a people. In Brandom’s terms, this would involve trying to think normatively without relying in any way on the material inferential commitments in which the responsibilities now being critically scrutinized were initially implicit. Hegel argues persuasively that no such attempt can succeed, because only the ethical life of a people—what a people does discursively—can give substantive content to the norms in question.10
Right-wing Hegelianism identifies the ideals of modern democracy with the vacuity of Moralität, a move emulated by recent communitarians and traditionalists interested in criticizing modern democracy in the name of premodern Sittlichkeit. This move is intended to make the concepts of justice, rights, and decency that are central to modern democratic cultures seem inherently ill-suited for the discursive roles they are meant to fulfill. Democracy is pictured here in Burkean fashion as an essentially destructive force, directed against the ethical life of traditional cultures. It sets the unconditional obligations of pure practical reason against the role-specific responsibilities of traditional culture. Democratic ideals are diagnosed as symptoms of a self-defeating attempt to rise above sittlich communities or traditions to the perspective of universal reason.
But there is a dangerous confusion at work in this way of interpreting the moral of Hegel’s story. An obligation can be universal in the sense of applying (as we see it) to everyone, without requiring a supposedly universal point of view (wholly independent of the ethical life of a people) for its justification. These two senses of universality are in fact distinct. An obligation that we attribute to everyone has what Rawls and Rorty call “universal reach.” But an obligation that is universal in this sense need not possess what Rorty calls “universal validity.”11 The most promising defense of an unconditional obligation to avoid acts of moral horror would frankly acknowledge its expressive function in the ethical life of a democratic people. That is, it would purport to be making explicit a norm that is already implicit in the practical inferential commitments of a community that exchanges reasons democratically.12 Once the norm is made explicit, it automatically becomes a candidate for critical inspection, because it now takes the form of a claim for which reasons may be requested. The background of material inferential proprieties, the expressive resources for making norms explicit, and the practice of exchanging reasons and requests for reasons with fellow citizens are, taken together, the discursive core of democratic culture. Democracy, far from being a freestanding set of institutional arrangements and abstract norms essentially opposed to culture, is a culture in its own right. Democratic norms are its expressive fruition. These norms allow us to ask such questions as whether there should be a role of master, whether women ought to have the opportunity to be political leaders, and whether some obligations should be attributed to the holders of any role whatsoever. We are fortunate to be able to ask these questions.
The fact that not everyone embraces our norms is no reason to think that such norms cannot include unconditional obligations. There is a distinction between those to whom democrats attribute the obligations and those—namely, the democrats themselves—who acknowledge the obligations by avowing them or attributing them to others. An unconditional obligation is one attributed to everyone, regardless of role. But this leaves open whether everyone acknowledges the obligation thus attributed. Sensible democrats understand that not everyone acknowledges the norms that they themselves avow. There is no paradox in supposing that some of these norms apply even to people who do not acknowledge them.
If I say that no one should torture another human being, I am avowing an obligation and attributing it to everyone else. This need not involve claiming that everyone else acknowledges it. Torturers who acknowledge the norm will be prone to secrecy, bad consciences, and hypocrisy. Other torturers, however, might do their dirty work openly, with clear consciences, and in ignorance of the norm I avow and the reasons I would offer for it. The ignorance of the latter group might be culpable or not, depending on the circumstances. Imagine a monk who turned the rack during the Spanish Inquisition. Suppose he was completely unfamiliar, through no fault of his own, with many of the considerations that lead me to condemn torture. (This is at least conceivable. I am not doing history at the moment.) The reason for concluding that the monk’s ignorance is truly nonculpable would be that he made proper use of the evidence, concepts, and norms available to him—all factors that vary from one cultural setting to another. I am less inclined to excuse a contemporary right-wing dictator who tortures political prisoners some of the time, gives speeches on supreme emergency on other occasions, and burns books by his democratic opponents whenever it suits his purposes. If he is ignorant of the reasons for condemning torture, the chances are good that he is culpably so.
Endorsing an unconditional obligation to avoid torturing people has the significance of avowing that obligation oneself and attributing it to everyone else. The avowal involves treating the practical inferential schema “Torturing X would be cruel, so I shall refrain from torturing X” as materially proper. Attributing the obligation to everyone else involves treating the cruelty of torturing someone as a reason that would entitle anyone to refrain from torturing people. In the example of the imaginary monk, however, differences between my collateral commitments and his complicate matters. The unconditional norm I endorse implies that the monk should not have tortured human beings. But in saying so I do not claim that he had access to the reason I recognize as a decisive count against torture. When I endorse cruelty as a reason for refraining from torture, I am offering the reason to everybody and I implicitly express confidence in my reasons for acknowledging this reason. But in saying all this I am keeping track of ethical responsibilities in light of my collateral commitments (MIE, chap. 3). I am talking about what everybody should refrain from doing. I am saying of everybody that they should refrain from doing something. But I am speaking from my own social perspective, making use of reasons and collateral commitments native to my setting.13 I need not suppose that the monk (given his social perspective) would in fact be able to acknowledge the force of the reason I acknowledge. There can be such a reason, and it can be a decisive reason when placed in the context of my collateral commitments, even if the monk and various others have no reason to treat it as a reason.14
The point might become clearer when put in terms of an image. A norm is a license for making inferences of a certain kind—in this case, inferences conforming to the schema mentioned in the previous paragraph. To endorse a norm implying an unconditional obligation is to issue a license to everyone to make inferences of that kind. But in issuing the license—that is, in offering the reason whose authority I myself acknowledge—I need not suppose that people outside my own discursive community can get their hands on it. Even if they came across it by accident, they might not be in a position to recognize it as a license, to acknowledge its authority. This particular license is on offer to all comers, even though I recognize that people who lack something like my conception of cruelty are unlikely to make use of it. Its intelligibility and authority derive from a background of material inferential commitments native to my environment, not from pure practical reason. It is universal only in one sense.
There is a simple way of finessing the issue by redefining unconditional obligations as a variety of role-specific responsibilities. Instead of saying that unconditional obligations apply to everyone (period), we could say that they apply to all holders of the (democratically basic) role of norm user. A norm user is anyone who possesses the expressive resources to exchange reasons for and against explicit normative claims. Then democrats can be seen as committed to conscripting everyone they can into this role: the young by educating them, the foreign by offering them reasons and asking them for reasons, the dead by imagining ourselves in conversation with them. But the role comes with responsibilities; we attribute the role and the responsibilities together. Some of these responsibilities are discursive; they pertain to the proprieties of reasoning about norms. Others impose restrictions on what may be done.
The idea that we are free to attribute this role and its attendant responsibilities to people outside our own culture is no more puzzling than the Roman designation of someone as a barbarian or the Jewish designation of someone as a gentile. It is simply our way of beginning the process of holding them responsible to us and ourselves responsible to them. Ancient, sittlich cultures had their ways of coping conceptually with people who were not friends, neighbors, or members of the family. They called them strangers or enemies—roles differentiated from each other by the presence or absence of hostile behavior, each with its distinctive role-specific rights and responsibilities. A stranger who spat on the sandals of a native official would be held responsible for his act. If the offense seemed grave enough, he might be reclassified as an enemy and treated accordingly. The spitting in this context might give license for what we would call a cruel response. We have our own ways of holding strangers and enemies responsible. One of these ways is by attributing to them the obligation to avoid torturing people. Strangers and enemies, like the rest of us, employ norms. We therefore press them for reasons and in various other ways hold them accountable for indecencies and cruelties.
The foregoing discussion suggests, then, that one can endorse an unconditional obligation while also standing with Hegel in rejecting pure practical reason. For the notion of pure practical reason, in the pejorative Hegelian sense I am using here, is that of practical reason operating independently of the ethical life of a people. In attributing to such people as the monk and the dictator the obligation to avoid torture, I have not adopted the standpoint of pure practical reason in this sense. But at least in the dictator’s case, I have begun a process intended to expand my own discursive community to include him as a fellow norm user, as someone I might hold responsible for his actions and as someone who might hold me responsible for mine. The unconditional obligation to avoid torture that I invoke against him has two things going for it. It is expressive of commitments implicit in our own practices, and it has thus far withstood criticism in the give-and-take of democratic discourse. Anyone who has mastered the relevant normative vocabulary can of course challenge the norm by questioning our entitlement to it or offering reasons against it.
My examples raise another complication worth mentioning. There is more than one way of becoming committed to a norm. The most obvious ways are by acknowledging it, explicitly through avowals or implicitly in action. But I can also be committed to a norm that follows from other commitments I have made. When I acknowledge a normative commitment, it directly implies other commitments, which I implicitly undertake whether I am aware of the implications or not. Together with my previous beliefs and practical commitments, the new commitment implies many more, which I also implicitly undertake. In this way the inferential significance of my acknowledged normative commitments can far exceed my awareness.15 And the same holds for the monk and the dictator.
As a result, it is possible that even the monk is committed unawares to considerations that deprive him of what should count, from his point of view, as entitlement to the policy of torturing people. The rudiments of my conception of cruelty might already be present among the unacknowledged commitments he has undertaken, waiting, so to speak, to be articulated and put together. If that is so, there may be a fairly strong sense in which he did have access to my reason for opposing torture after all. The relativity of entitlement to context does not necessarily get any torturer who is unfamiliar with the explicit articulation of democratic norms and arguments off the hook (as inculpable for his or her moral ignorance). Whether it does or not in a given case is something that can be determined, if at all, only through meticulous historical inquiry and immanent criticism. But even if our monk turns out to be on the hook, in the sense of being culpable through negligence for his moral ignorance, it would not take an appeal to pure practical reason to put him there. For in concluding that torturing people is in tension with some of his own unacknowledged commitments, we would still be appealing to the ethical substance of a culture, namely his. We need not suppose that everyone who engages in practical reasoning—or who engages, as Habermas would have it, in communicative acts—has implicitly undertaken commitment to the normative conclusions a democrat would like to see explicitly endorsed.
OVERRIDING OBLIGATIONS AND QUESTIONS OF IMPORTANCE
The previous section attempts to make sense of the idea that there is an unconditional obligation to avoid some sorts of especially bad acts—horrible acts—such as the act of torturing people. That analysis does not resolve the problem of dirty hands, however. Its first purpose is to bring the problem into focus as a conflict between a responsibility we attribute to those holding a certain role and an obligation we attribute to everybody. The second purpose is to show that even the latter can be interpreted as an expression of commitments implicit in democratic culture, rather than as a theorem in transcendental philosophy. The question that remains is which of these two norms overrides the other in cases where a political official cannot fulfill both.16 This is the crux of the problem of dirty hands.
The first point to grasp is that whether one norm overrides another is distinct from the question of their status as role-specific or unconditional. In the Catholic code, as we have seen, the priestly responsibility to maintain the confidentiality of the confessional overrides the obligation we all have to provide information to the police in certain circumstances. It is perfectly coherent to suppose that a role-specific responsibility can trump an unconditional obligation, or vice versa. The reason for this is that the distinction between role-specific and unconditional constraints has nothing to do with the question of how important these constraints are—with how and why we care about having them fulfilled. Judgments about whether one norm overrides another express commitments about the relative importance of the norms being compared. An obligation can, in theory, be both unconditional and trivial. The problem of dirty hands is as problematic as it is, however, because both of the concerns that constitute the problem are obviously very important to most democratic citizens and because we ascribe importance to those concerns from somewhat different points of view.
The Catholic stand on priestly confidentiality reflects an underlying commitment to the importance of confession in the sacramental system for saving souls from damnation.17 In the grand scheme of things, an obligation Catholics attribute to everyone need not be more important than a responsibility they attribute to the occupant of a particular role. But Catholicism has a view of the grand scheme of things. This view authorizes a definitive stand on the problem of priestly confidentiality—and also, for that matter, on the problem of dirty hands. Modern democratic culture, in contrast, does not have a single view of the grand scheme of things; it opens up space in which many such views can be held and acted upon. It is not vacuous in the sense that Hegel thought Moralität was, for the ethical life of democratic peoples clearly commits them to the importance of such matters as the people’s survival and the decent treatment of others.18 It does not, however, entail a single ranking of the most important things, for most of its citizens have agreed to disagree on questions of ultimate importance. They have not been able to settle these questions by rational argument so far, and do not expect to do so in the foreseeable future. They are therefore prepared to grant one another considerable latitude in answering them.19 They cannot decide which of their highly important moral concerns is the most important to them as a society. (All of them agree that bombing a civilian population is very bad, but not all of them think that such a horror would in all circumstances be literally inexcusable.) It should come as no surprise, then, that they are of several minds on the problem of dirty hands, that the problem remains problematic for them.
Nor should it come as a surprise that this is one of the issues on which some citizens feel compelled, in conscience, to express their religious commitments in public. Questions about the relative importance of highly important values are for most people so intimately connected with religious commitments that they would be hard-pressed to defend their answers to them without employing such commitments as premises.20 This fact counts heavily against any theory that proposes to exclude the expression of such premises from the public sphere. These are things we need to talk about with one another, even at the risk of conflict, awkwardness, and a good deal of stuttering.
I, for one, aspire to belong to a society that would treat the commission of horrible acts like the intentional bombing of civilian populations as literally inexcusable, even in situations where such a tactic is thought on plausible evidence to be necessary in fending off terrorist destruction. I fear that anything less than an absolutely overriding prohibition of such acts will tempt our leaders to commit moral horrors in situations that only seem, at the time, to be emergencies. So I worry that endorsing the supreme-emergency exception would weaken democratic culture by undermining its ability to sustain a genuinely democratic politics, here and now. I am also prepared to bite the bullet in the event of an actual emergency. A society resolutely committed to avoiding the infliction of moral horrors would rather go down in flames, while treating this commitment as a matter of integrity, than survive by instructing its leaders in advance to perform such acts on its behalf if an emergency arises. I would defend this commitment as an apt expression of the importance we ought to attribute to human beings as unique, irreplaceable individuals with the capacity to love, exchange reasons, repent of wrongdoing, and suffer. As I see it, the issue is what kind of people we are going to be—a matter of self-definition and integrity. It is about what we care about most, of what we deem sacred or supremely valuable or inviolable, not the desire to have clean hands.21
I don’t expect to change the minds of those who feel that in an actual supreme emergency they would attach more importance to the people’s survival than to the people’s identity and integrity as a community dedicated to removing moral horrors from the list of permissible political means.22 Nor do I think philosophy has the means to discredit as irrational either absolute intolerance of horrible acts or a willingness to grant rare exceptions. While I aspire to live in a society that shares my convictions about the moral limits of war, I understand that the utopia I am projecting asks more of my actual fellow citizens than many of them are prepared to give when they are threatened by potentially massive terrorist attacks. Moreover, one of the things that is highly important to me is living in a society where I am free to decide what is most important to me and in which others are similarly free. So there are limits on what I am willing to do to bring others into perfect agreement with me on the morally most important things.
But I still abominate the bombing of civilian targets, the torture of prisoners, and various other things my government is either ordering or contemplating. And because of this, I am necessarily left in severe tension with those of my fellow citizens who today authorize our leaders, albeit conditionally, to do things which horrify me. Living with such tension is not an easy task, because it borders on complicity in horrible things. That is why Thoreau, a profound modern democrat, so often felt like resigning symbolically from his country. One needn’t be a sectarian to have such feelings.
It is because modern democratic discourse does not proceed from antecedent agreement on how the most important values should be ranked, but nonetheless invites citizens to express their own reasons for their commitments on important public questions, that democratic culture tends to evoke ambivalence even from its defenders. The sectarian temptation is to allow the ambivalence, which is appropriate, to turn into alienation, which only makes things worse. Sectarians propose to withdraw from the broader community of discourse, convinced that it would be a violation of their commitments on matters of high importance even to exchange reasons and requests for reasons with those who do not share those commitments.
I have just been portraying the debate over the problem of dirty hands in the age of terrorism as an argument over the relative importance one ought to assign to two conflicting prima facie responsibilities—one role-specific, one not. But it would be misleading to leave the impression that reasoning on this issue within religious traditions like Judaism, Christianity, and Islam has generally proceeded from uncertainty about how much the relevant prima facie responsibilities weigh. That is not in fact how the reasoning always—or even very often—has worked in those traditions historically. Within such traditions, the overriding importance of the responsibility to avoid committing certain injustices has often been taken for granted in such a way that the problem of dirty hands does not even arise. For those committed to the resulting forms of absolutism, practical reasoning typically begins by classifying actions as either intrinsically prohibited or not, so that “weighing” the relative importance of conflicting responsibilities comes into play explicitly only for actions falling into the latter class.23 One of the deepest worries about modern democratic society, from the perspective of those committed to such absolutism, is that this society does not take absolutist commitments for granted as a premise from which all practical reasoning proceeds. In this context, absolute dedication to justice, or even to the avoidance of moral horrors, is a commitment for which reasons are constantly being requested, not a premise on which all implicitly agree. This means that our ethical discourse does not take the shape that many of us would like to see. So long as this remains the case, the problem of dirty hands will remain with us, and some of us will remain ambivalent, at best, about participating in our common life at all.
Without pretending to solve the problem of dirty hands per se, I want to conclude by offering two contextually specific arguments for placing especially rigorous constraints on the political officials who are leading us in the struggle against terrorism. The first argument is simple and appeals to consistency. What we condemn in terrorism is precisely the moral horror it involves—its intentional targeting of civilian populations. We cannot maintain consistency without holding our leaders to the same standards of conduct we apply to the leaders of Afghanistan and Iraq. If we are not prepared to make exceptions for our enemies, we should not make them for ourselves.
The second argument is prudential. The struggle against terrorism is not only military, but also ideological. We are unlikely to win it on the ideological front if we cannot persuade people who are tempted to side with the terrorists that we are not essentially hypocritical in condemning the terrorism that threatens us. If we show callous disregard for the lives of innocent civilians—or, for that matter, intentionally frustrate the legitimate democratic aspirations of other peoples—in order to protect our own country from terrorism, then our country will not be seen as a champion of justice and democracy. The outcome of this struggle depends largely on the perceived sincerity and rigor of our ideals and principles—on what people all around the world take our character to be. We will win only if we gradually earn the trust of those people. We can do that only by proving ourselves true to our principles even at those moments when we are sorely tempted to forgive our leaders for violating them. If we cannot manage to attain the moral high ground and stay there over the long haul, we are going to lose the ideological battle. And if we lose that, the flow of terrorist recruits will never cease.