Hatred and anger are the greatest poison to the happiness of a good mind.

-- Adam Smith, The Theory of Moral Sentiments

1. Character and the Minimal Functions of the State

In the course of defending the American Revolution, Thomas Paine made a statement that neatly distinguishes between the functions performed by various institutional arrangements:

Some writers have so confounded society with government, as to leave little or no distinction between them; whereas, they are not only different, but have different origins. Society is produced by our wants, and government by our wickedness; the former promotes our happiness positively by uniting our affections, the latter negatively by restraining our vices. ... The first is a patron, the last is a punisher.(1)

Part of what Paine meant to express here was a conception of government that for a long time had considerable influence in this country. This was the idea that, as Paine says elsewhere in the same pamphlet, "security" is "the true design and end of government."(2) That is, government should only carry out what might be called "the minimal functions of the state": those in which it either discourages the use of force and fraud by exacting retribution from those who use them or compels such offenders to make amends for what they have done by compensating their victims. What government does is negative: it corrects human misbehavior. For the rest, Paine meant to say that the positive benefits we can expect from our institutions must come from "society," which is simply the non-governmental part the social world. These benefits include positive effects on our character, or what he called "the affections." One clear implication of this is that the state can have no positive effect on character. Indeed, this idea seems to be logically connected, for Paine, with his minimalist conception of government: the proper business of government is to carry out its minimal functions and we therefore cannot expect it have a positive effect on what sorts of people we are.

At times one of these ideas, or one of their presuppositions, has been used to defend a conception of government that in one respect is the opposite of the one Paine held. George Will has argued that the minimalist conception of the state is wrong on the grounds that government ought to engage in "soulcraft," in the activity of making better people of us. Though his conclusion is the opposite of Paine's, there is an assumption lurking behind it that is identical to one that Paine relies on: if the state does not deliberately perform any functions but the minimal ones, it will have no reliable positive effect on the character of its subjects.(3)

In this chapter I will argue that this assumption, shared by the radical individualist Paine and the conservative communitarian Will, is not true. I maintain that, as states perform their minimal functions - in particular, as they enforce the criminal law - they do tend to have a certain beneficial effect on the character of the people to whom their laws apply. This effect is a result of certain morally important differences between the way in which states exact retribution and the way in which it would be exacted by the victims and private individuals acting in their behalf: by displacing private retribution, the state helps to drain revenge from the community it governs. It does so by means that also serve to ease the tension that we earlier saw exists between forgiveness and self-respect (IV.5). It has these effects, not as the result of artful intentions on the part of legislators, but inevitably and spontaneously.

In this chapter I will give reasons for thinking that my thesis is true, and will try to make them more vivid by asking the reader to imagine what would happen if the minimal functions of the state were performed by private protection agencies, business firms that, for a fee, would exact retribution from individuals who victimize their clients.

2. Punishment

I will begin by pointing out some linguistic facts, certain differences between the ways we speak in different circumstances. In a later section of this chapter I will provide evidence that these differences in speech reflect real and important differences between the facts we are speaking about in those circumstances, but I will not be concerned with that just yet.

Suppose that a friend of mine does something to offend me. Overcome by anger, I slap him for it. There are a number of things that might be said about such behavior, but no one, I think, would say that I have punished him. No one would say "You punished him too severely" or "It's barbaric to punish your friends with violence - use more civilized methods." This is what we call "getting back at" someone, or "getting even." Whatever I might do to my friends and however harshly or gently I might do it, I cannot be said to punish them. On the other hand, when parents slap their children that is the sort of thing that can be called punishment. We observe this distinction even if we think in both cases the slapped person "had it coming."

Why is it that one of these acts is not a punishment while the other is? One might think that the reason why the first case is not called punishment is that it is a "selfish" act: I hit him just because I don't like what he did to me. That is what "getting back at" someone is. But it would still not be called punishment if I had slapped my friend for offending my brother. I can get back at someone in someone else's behalf - altruistically, one might say.

It will be easy to give a more adequate account of the differences between these cases, though still only a partial one, if we consider a few more examples. The charges that my bank levies against me when I carelessly over-draw my account resemble the fines a government might exact from me, except that while what the government thereby does to me can be called a punishment, what the bank does cannot. If the Israeli government assassinates a Palestinian terrorist in Lebanon, that is not called punishment, even if the killing is ordered because the terrorist killed fifty Israelis the week before, and even if it is carried out because the terrorist is thought to deserve it. On the other hand, if the same government executes an Israeli subject for killing someone, we do call that punishment (even if we think that the subject is innocent or the penalty is excessive).

From these examples we can extract certain general principles which, I think, are all true. People are not said to punish their friends, businesses are not said to punish their clients, governments are not said to punish their external enemies; governments are said to punish their subjects, parents are said to punish their children. There is a pattern here that is perhaps obvious by now: in the case where one is said to punish, one is thought to have authority over the person one punishes; in the other cases this is not so.

"Punishment," as we ordinarily use the word, is not identical to what Nietzsche called "the drama" of punishment: "a certain strict set of procedures" such as, to use Nietzsche's grim examples, drawing and quartering, boiling in oil, and dropping a millstone on the malefactor's head.(4) We have also seen that punishment is not identical to retribution, either; we still do not have the concept of punishment if we add to the idea of "the drama" the notion that such things are done to give the victim what he or she deserves. One must add - at least - that the one who does it is in a position of authority over the one to whom it is done. This does not mean that only those in such a position have a right to punish; only they may do something that counts as punishing.

Punishment is generally spoken of as a perfectly proper human activity. The alternative sorts of retribution, most often called "revenge" and "vengeance," are treated quite differently, as something ignoble. Their names seem to function in our language as the names both of passions and of vices, and even when treated as passions they are generally referred to with disapproval.

Here, as elsewhere, though, one needn't follow the way people generally speak. One can take at least two different points of view on the conventional relations between punishment, authority, and revenge. On the one hand, we might think that punishment is the same human activity as revenge, we simply call it punishment when it is done to us by our superiors. The enormous emotive difference between "punishment" and "revenge" merely signifies that we accept from our superiors a sort of treatment that we would not stand for from our equals. What seems to be a great moral difference is simply a symptom of human servility. On the other hand, we might take the view that these are two genuinely different forms of conduct and that there is something in the nature of authority, or of some kinds of authority, which makes this difference possible. People in authority, or some of them, are in a position to do things that are proper and not ignoble.

While the first of these two points of view is surely not entirely false, I think the second is entirely true. Before I try to show this, I should say something about what authority is.

3. Authority

We can divide authorities into two kinds. On the one hand, some people are said to be authorities on certain subjects, such as numismatics or economics. On the other hand, some people have authority over other people, such as their children or their employees. One may be an authority on a subject or over people. I will call the former sort "theoretical" authorities and the latter "practical" authorities.

One exercises theoretical authority by making statements. If someone with this sort of authority makes a statement within the sphere of his or her competence, that is a good reason for believing what that statement says. If Joe DiMaggio says "This is the way to hold a baseball bat," his saying so is evidence that this is the way to hold a baseball bat. Note that this means that one may be a theoretical authority on practical matters - on what is to be done. Theoretical authority is not distinguished from the practical kind by its subject matter. Practical authority is exercised by issuing orders. If a mother tells her five-year old son to go to bed, that is a good reason for the son to go to bed. When a practical authority issues an order, it is a good reason to do what the order tells one to do.

It is not difficult to see why both sorts of authority are called by the same name: in both cases, an "authority" is a creator of reasons. But there is an important difference between them that is relevant here. Theoretical authorities are not said to punish people. Accordingly, when I say "authority" in what follows, I will mean practical authority.

What I have said about authority is admittedly brief and sketchy, but it enables us to ask with some clarity whether and in what way it is necessary for states to possess either the reality or the appearance of authority. It seems easy to imagine a state that carries out its minimal functions without claiming to have any authority over its subjects. Such an organization, it seems, could pursue and catch people who violate the rights of others and then do such things as taking their property, confining them in prisons, or killing them. Its representatives can execute murderers without claiming that the reason they should have refrained from murder was that the state had laid down laws telling them to refrain. They may simply hold that murderers should not act as they do because it violates the rights of others. They need not tell murderers that they are putting them to death because they disobeyed the state; they may just claim that they are killing them because they are murderers. They might even claim that, on grounds having to do with risk and procedural justice, they are the only ones who have the right to exact retribution for such behavior. None of this would amount to claiming authority, because in none of it do they claim that their say-so by itself constitutes a reason why their subjects should do what they say.(5) But what I have said earlier clearly suggests that a state that claims no authority cannot make one particular claim that all states do in fact make, one that seems essential to the nature of the state: it suggests that they cannot claim that, in doing such things as putting people to death, they are punishing them. In this way, at least, the concept of authority does seem to be essential to understanding the state.

The short account of the nature of authority that I have given suggests one possible way to explain why only practical authorities can punish. Only practical authorities can issue orders which are reasons for doing what the order enjoins. Thus, only they can enforce such orders; only they can harm people for disobeying authoritative orders that they have issued.

Unless this account is amended somehow, it will surely do my case no good, because it does not show how punishment could be superior to non-authoritative retribution. In fact, unless it is augmented, it rather tends to show the reverse. If this is all there is to it, authorities, when punishing people, would seem simply to be trying to get back at them for ignoring their orders and thus insulting them by implicitly denying their authority. This would mean that punishment is simply a sort of revenge that is peculiar to people who claim to possess authority.

Actually, there is another reason for thinking that this account is in fact not complete. If this were a full explanation of the conceptual connections between authority and punishment, all practical authorities would be treated as competent to punish, but in fact they are not. This can be seen by considering a few examples. Employers are treated as having authority over their employees: they give orders and, normally, they are obeyed. Yet, although they can enforce their orders by threatening their subordinates with unpleasant things, such penalties are not generally referred to as punishments. Doctors are typically viewed as practical as well as theoretical authorities - we take "doctor's orders" to be reasons for doing whatever it is the doctor orders. Teachers are thought to exercise a sort authority over their students, but the low grades they give students who do not do their work are not called punishments.

What reason might there be for regarding these authorities as not competent to punish? They have a number of relevant characteristics in common. In each case, the authority of the person who has it is the result of a contract into which that person and his or her subordinates have entered freely. They have come together, not in order to pursue a whole way of life together, but only to satisfy certain specific desires with relatively limited objectives. In the examples I have cited, these desires include the desire for remunerated work, the desire for medical treatment, and the desire for knowledge. Punishment requires an area of concern which is in some sense broader than the sort involved here. For instance, authorities, in punishing, may thereby express a concern for their subordinates' true welfare (as when parents punish their children for playing with fire and paternalistic governments punish their subjects for using heroin) or they may express a concern for whether their subordinates' behavior is morally right or wrong (for instance, whether it constitutes a wrongful use of force). Both of these concerns are typical of the two practical authorities I have mentioned that are regarded as competent to punish: parents and states.

4. The Problem of Revenge

This is no doubt an incomplete explanation of why some authorities can punish, but it already has implications regarding the question that most immediately concerns me here: how can the punishments that at least some authorities perform plausibly be held to be morally superior to revenge? I have said that the fact that some authorities are not regarded as competent to punish has something to do with the way in which the actions of those authorities are prompted by the desires of certain people. It can be argued, on the other hand, that certain authorities that are said to punish people are in a position to plausibly claim that, when they punish, their own desires are not involved in the same way and to the same extent they would be if their punishments were acts of vengeance. It can also be argued that when this claim is true, what they do is morally superior to revenge.

To show this, some definitions are needed. "Resentment" is an emotion that is inspired by the thought that one has been wronged. If I experience resentment, it is because I interpret the thing that was done as wrong and because I realize that this thing was done to me. It is a quite different sort of passion from what I will call "indignation," which is provoked simply by thinking that a wrong has been done. When I am indignant either I am upset about a wrong done to someone else, or I am upset about a wrong done to me just as I would be if it were done to someone else. The fact that it was done to me is not part of what makes me feel this way. It does seem possible to be merely indignant about a wrong done to oneself. I say "merely" because resentment has a peculiar, poisonous intensity that indignation generally lacks.

In the sense in which it is the name of an emotion, "revenge" (or "vengeance") is the sort of resentment in which the crucial interpretation - that someone has wronged oneself - gives rise to a felt desire to harm the one who has done the wrong. It is a desire for retribution.

What is wrong with revenge, regarded as a passion? First, it is easy to find objections to the sort of behavior that typically flows out of revenge. Such conduct has a number of undesirable consequences. Partly because it is such an intense passion, it tends to make us overestimate the wrong we have suffered and thereby tends to make us exact more retribution than strict justice would allow. In the same way, it goads us into harming the innocent. Both these excesses provoke vengeful reactions in others, which begins the vicious spiral of feuding.(6)

What is much more important for my present purpose is the fact that, even apart from these consequences, the passion of revenge tends to have bad effects on people who experience it. Because of these effects they are liable to become worse human beings and live worse lives that before. The effects involved follow not so much from vengeful behavior itself as from the passions and standards of value that lie behind it. When wrongs that have been done inspire resentment, the individuals who experience it think about those wrongs more than they would have otherwise. They brood over them. The more they think about their grievances, the more resentful they feel. If left to itself, this cycle can spiral into a monomania that pushes more rational interests aside and clouds the principled concern for right and wrong that inspires resentment in the first place. Even if they do not do anything, they become willing to do things that they would have considered beneath themselves before. Through resentment, their principled rejection of wrongdoing undermines itself.

Revenge unsettles one's reason. The cycle it begins is difficult to break because, although, as Adam Smith said, there is "something harsh, jarring, and convulsive"(7) about resentment, there is also something curiously pleasant in the cruel fantasies it brings with it. It leads to ways of thinking, irrational and out of control, that seem bad in themselves, but at least as important is the content this thinking and of the resulting behaviour. It tends to distance us from other people, resulting in a certain all-too-familiar objectification of the other and a resulting indifference and hostility to their well-being. In other words, it undermines the values that forgiveness promotes: our concern for the good of others and our connection with them.

All these objections to revenge, regarded as an emotion, are compounded by the fact that it is closely related to a vice that sometimes goes by the same name, though it is more often called "vengefulness" or perhaps "vindictiveness." The most troublesome aspect of this relationship is, simply, that experiencing the passion can easily cause one to also have the vice.

The emotion of revenge, being a species of resentment, is based on a standard of value. In this way it is like any other emotion (see V.8). In the case of resentment the standard is one that implies that something that has been done to oneself is wrong. This standard might be shared by many other people, including ones who are incapable of revenge. Revenge includes something in addition to this: it also contains a felt desire to get back at the one who did the harm, a desire that arises from the belief that the wrong was done to oneself. As a passion, however, revenge does not include a belief that doing such harm is justified. When one is in the grips of vengeful feelings, harming the enemy seems like the thing to do, but this is a matter of emotional interpretation. It is partly for this reason that the passion of revenge can be a mere episode, unrelated to one's enduring character. As a passion, revenge can be experienced and acted on akratically: one can act on it despite the fact that one does not believe that vindictive behavior is really justified. The akrasia begins to disappear when one begins to accept a certain principle which, unlike the one that lies behind resentment, is not shared by all sorts of people. This is the notion that wrongs to oneself do justify getting even with the perpetrator. The justification need not be the sort of thing that a contemporary moral philosopher would think of as a moral justification. It might be based on a medieval conception of "honor" or a mere belief in the absolute necessity of maintaining an appearance of invulnerable power. The transformation of revenge into a trait of character is complete if this principle, whatever it might be based on, becomes a settled part of one's own behavior.

As I have said, one of the hazards of the passion of revenge is the fact that it can lead to the formation of the vice of revenge. The brief description I have just given of the vice indicates part of the reason why the passion is liable to lead to it. Ordinary life, unfortunately, brings us many occasions, both petty and grand, for the passion of revenge. People who experience an emotion persistently are under some pressure to see it as justified. This is particularly true if they act on it. One feels called upon to rationalize the emotion by developing views that would justify it, and this lends some appearance of rightness to the actions that flow from it. But this change in how our actions appear is the sort of change that I have just said is essential to hardening the passion of revenge into an entrenched trait of character. Once we have undergone this transformation, we accept a principle to the effect that wrongs to oneself justify getting even with the perpetrator. Further, if what I have said about vengeful behavior and the passion of revenge is true, this trait of character is a vice. All the objections raised against them applies to it as well, but with greater force. To have this trait of character means to increase the frequency of these intense feelings and confused thoughts, with a corresponding increase in the undesirable practical consequences. But the difference the trait of character makes is not just a matter of quantity, of more of the same bad things happening. As long as my vengefulness is a matter of having the episodic thoughts and feelings that come to most of us from time to time, they come to me unbidden and perhaps unwelcome. However much I am to blame for my reactions to them or what they lead me to do, I can still say that they do not represent who I am or what I stand for. If my vengefulness becomes a trait of character, however, this saving truth is no longer available to me. I am no longer the weak victim of my thoughts and feelings: I have gone over to the enemy's side. Not only am I somehow deluded into thinking that these things that I am doing are right, but this error of mine is the reason I am doing them.

Clearly, it would be desireable to do something about the passions of revenge and resentment, if that is possible. But what should we do about them? There is certainly no question of eliminating both of them from our lives altogether. Even if that were possible I submit that, paradoxical as this might sound, it would not be desireable. Part of the reason for this lies in some psychological facts that I will assume are familiar ones. It seems to be inevitable that most of us are upset when people victimize others. It seems almost as inevitable that most of us believe in the moral appropriateness of retribution of some sort. Of course, not all of us live by the principle that "everything is dischargeable, everything must be discharged" (see IV.2), but most of do think that retribution of some sort is at least a fitting response to wrongdoing and is often the right thing to do. There are probably good reasons why we should think this, but for the present the point is that most of us do think that way and will probably continue to do so.

Given this, eliminating resentment would amount to replacing it with indignation: if we somehow eliminated the highly personal negative emotional responses to wrongs done to us, we would experience the impersonal one instead. That is, we would be upset about wrongs done to ourselves just as we would be if they were done to someone else. This would mean being utterly indifferent to one's own dignity in the sense in which I have been using it, in which dignity is the fact that one appears, in one's own eyes, to be a worthy human being. Wrongs committed against oneself are always assaults against one's dignity in this sense (see II.5).

This is part of the reason we are so upset when we think someone has wronged us. This psychological fact seems, once again, to be more or less inevitable. The only way to avoid being upset for this sort of reason, other than having a sense of one's own worth that is so invulnerable that it cannot be touched by the world, would be to be unconcerned about one's worth at all. The former way seems quite impossible for most of us and the latter is highly undesireable. But this means that one component of one's upset about wrongs done to oneself will be that one realizes that it was done to oneself, and this means that it constitutes resentment and not mere indignation.

This concern with our own dignity, necessary and desireable though it is, stands at the beginning of a process which, if it is not hindered, causes serious harm to the community and to the character and well-being of the individual. The common belief that some sort of retributive response is appropriate to wrongdoing provides those who are its victims with a ready antidote to the hurts their dignity has suffered. If someone has mistreated me, that is palpable evidence that I am unable to prevent such things from being done to me. As such, it is evidence (maybe smartingly good evidence) that I am impotent. If I manage to exact some sort of retribution, even if it is retribution of a mild sort, that would tend to show that I lack this negative characteristic and have instead a certain positive one, namely, efficacy. Depending on the context, my efficacy will indicate to me various other positive traits, including perhaps courage, resourcefulness, and various sorts of skill. But if one wants to take this solution one is not merely resentful, one is experiencing the passion of revenge.

The line between self-respect and revenge is very thin. Concern with one's dignity, together with ideas and traits that are perfectly reasonable and normal, to seem to lead very naturally to an emotion that has an explosive power to do harm. The problem is how interrupt the process by which this happens.

Most likely, this interruption must occur after the beginning of the process: suppressing all concern for one's dignity is not an option, and as long as there is any such concern the potential for revenge is there. In the scenario I have presented, revenge does not come from excessive concern for dignity, but from reaching for the wrong antidotes to a sense of degradation that may, for all that, be perfectly legitimate.

A more modest sort of solution to our problem would be to introduce the interruption fairly late in the process: instead of trying to obliterate all concern for one's own dignity, or all forms of resentment, or indeed even revenge itself, we might try to contain the influence that the passion of revenge has over the workings of our minds. Revenge, after all, seems to be the proximate cause of the trouble I have described. As we have already seen, we can to some extent contain the passion of revenge if we can prevent it from growing into a settled trait of character. Both the power of revenge and its badness increase greatly if one accepts a principle that justifies it. Essentially, it is this acceptance that transforms it into a trait of character. Without this transformation, one's resentment either remains an isolated episode or, at worst, a temperamental disposition against which one can struggle akratically. To the extent that this transformation can be prevented, a particularly virulent form of resentment will have been contained, and without simply repressing it from consciousness.

5. A Solution

Even this relatively modest approach to the problem might seem dauntingly difficult. It does require of me that I exert control over the influence my emotions have on the rest of my life, and it is not obvious how this can be done. Actually, it is not as difficult as I have perhaps made it seem. As we have seen in the last three chapters, my efforts to alter or preserve my character do not take place entirely in the solitude of my own mind. Around me are institutions that act as a sort of moral furniture, supporting me in this enterprise. In particular, the state performs this sort of service as it carrying out its minimal functions. To the extent that its authority and the conventions that surround it are accepted by its subjects, it helps to drain malicious resentment from the community it governs. It does so by discouraging thoughts that justify revenge and extend its power over us.

Parents sometimes claim that in punishing their children they are doing it for the child's own good. We often find this claim a plausible one because we believe - with some reason - that concern for the child's welfare is natural in parents and can be stronger than anger at the child's misbehavior. Thus, at least on occasion, we can reasonably believe that parental punishment is not a vengeful act.

Governments, by means of various symbols and ceremonies, and also through direct statement, make the same sort of claim. The beliefs thus fostered, and the trust that arises from them, are probably indispensable conditions for the stable existence of states. In the case of governments such professions of disinterestedness are much more plausible than they are in the case of parents. The misdeeds for which parents punish their children are often things that harm or irritate the parents themselves: the child has ruined an expensive machine or been distractingly noisy, and so forth. This means that occasions for parental punishment often present temptations for parental vengeance. Except in a small number of cases (such as treason) this sort of thing is not true of the misdeeds for which states punish their subjects. One can imagine the officials of a state being irritated with a burglar for showing so little respect for their laws against theft, but they are not victims of the theft, and the provocation they have suffered is a far less serious insult than what the victim has suffered. The state generally does not have that personal stake in the matter which is required by revenge: it is what Locke appropriately calls "an indifferent judge."(8) For the most part, when it punishes someone, it is not getting revenge for harm done to it.

Nonetheless, what the state does would still be vengeful if it were seeking retribution for the victim's sake - this is the sort of thing I would be doing if I slapped my friend for offending my brother. But in various ways governments claim, more or less convincingly, that insofar as they enforce the criminal law for the sake of someone's interests, they do it for the sake of the community in general: more precisely, they do it in order to deter the sort of injustice the criminal has perpetrated and, thus, to protect the community from danger. This, officially, is part of what distinguishes the laws thereby enforced as "criminal," in contrast to the "civil" law which affords remedies to private individuals.

Moreover, most states disseminate the belief that what motivates their officials is to a large extent not a goal at all, not even the goal one seeks out of honest indignation: rather, they are moved to act by the law and by their understanding that the law requires that this be done. Indeed, if this were not true of a given state, it is doubtful that it could be said to have a legal system at all, since guiding the conduct of those who administer it is one of the indispensable functions of the law.(9)

What is more important, for our purposes at least, is the fact that the subjects of the state tend to share these beliefs: they generally think that the state exacts retribution coldly, impersonally, as a matter of policy. If I, a subject, believe a certain further presumption that all political authorities make - namely, that the seeking of retribution is solely the prerogative of the state - I will think that this cold way of proceeding is the proper way, and I will also think that seeking retribution in my own behalf is literally none of my business: it is a matter for the police and the courts. Thus the state and the beliefs that conventionally come with it discourage me from having vengeful thoughts. That is, it discourages me from thinking that wrongs against me justify my exacting retribution for them. In addition, by accomplishing what would have been accomplished by personal revenge (namely, by hurting the person who has hurt me) it removes a large part of the point of having such thoughts. Thus it tends to prevent episodes of temperamental revenge from growing and hardening into a trait of character and a vice.

If the state, by accomplishing the palpable results of vengeful action, were also to afford its subjects the emotional gratification that is characteristic of such action, it would probably inflame vengeful passions rather than discourage them. However, it offers very little in the way of such gratification. To a vengeful person, it is most gratifying to achieve such results oneself. It is less gratifying, but typically still acceptable, to hire someone to do it (e.g., to hire a thug to break the malefactor's legs); in that case, retribution is exacted through something one does and for one's own sake, although it is not done directly by oneself. It would also be satisfying, though still less so, if an ally were to step in and express sympathy with the vengeful person by "evening the score" without being asked to do so; at least the one for whom the score is evened can feel that he or she personally had something to do with it.

Adam Smith identified two things we want when we act vengefully: we want those who have mistreated us to "be made to repent and be sorry for" what they have done, and we want this to come about "by our means."(10) If my enemy is accidentally crushed to death by a falling rock, that may be an occasion for some gloating, but it will not gratify my revenge because it fails to achieve either of these purposes. Both of them, and especially the latter one - that the lesson comes about "by our means" - are necessary if our feeling of lost dignity is to be displaced by a clear demonstration of our efficacy. It is precisely the latter purpose, however, which is not served by state punishment. I am aware that the government's legal aparatus can teach my enemy a lesson - insofar as that is possible - but I know that this will not be done by me, through me, or even for me. Without this, the retribution gained does not promise to be the intense source of pleasure it would otherwise be.

So far, then, such punishment counteracts revenge by counteracting the belief that hardens it into a trait of character and, in addition, by depriving it of emotional gratification. But it acts against revenge in ways that go further and deeper than this. One of the beliefs that tend to be widely shared within a political system, held by both the sovereign and its subjects, can prevent resentful feelings from arising in the first place, at least to the extent that victims of aggression accept it. This is the familiar idea that the wrongdoer is to be seen simply as one who breaks the law: "lawlessness" is what is wrong with what the criminal does. This has a very important result, one which was identified long ago by Nietzsche: "from now on the eye is trained to an ever more impersonal evaluation of the deed, and this applies even to the injured party himself."(11) To the extent that injured parties accept this idea, their attitudes toward retribution tend to some extent lose the reference to oneself without which resentment is impossible.

The benefits of a state that performs its minimal functions are difficult to perceive because we live with them, but they should show up more clearly if we consider what sort of community we would have if private firms did the same work. Suppose that I live in such a community. I come home one day to realize, with horror, that my apartment has been robbed and maliciously vandalized. The protection agency to which I subscribe discovers who the culprit is, catches him or her, and exacts retribution.

Does their retribution constitute punishment? If we follow ordinary usage the answer is clearly "no." As private businesses, they cannot be said to punish their clients or their employees, much less could they be said to punish someone who is a stranger to them, as the culprit in this case may well be. They cannot be said to punish at all. According to some remarks I have made earlier on, this is because they derive what authority they have form a contract that all parties enter into in order to satisfy certain desires with relatively limited objectives. For the same reason, what they deal in is in fact revenge. More specifically, this is due to the nature of the desires that the contract satisfies. If I subscribe to their retribution-exacting service, it will be, as far as this issue is concerned, just as though I had paid a thug to beat up people who have harmed me: they would be peddlers of broken bones. If they are successful, they afford me the satisfaction of revenge: not the intense satisfaction I would have by doing it myself, but at least the sort that I can secure by having hired agents of mine - competent professionals, presumably - to do it for me.(12)

In such a system, no cultural barrier interposes itself between the goals of self-respect and the means employed by revenge. The benefits of a system in which there is such a barrier cannot be percieved directly, but we can identify them conceptually by contrasting such a system with one in which retribution is private. In the private system, offenses against oneself are directly relevant to one's self-respect and the vengeful solution is made all the more attractive by the fact that it is visibly and effectively carried out. On the other hand, to the extent that a state performs its minimal functions, and to the extent that one accepts the ideology that comes with the relevant governmental institutions, neither of these facts is the case. One's revenge is not being gratified and, more importantly, the offenses are not relevant to self-respect at all. Insofar as I adopt the impersonal evaluation of the offense, it loses its character as an affront to my dignity. What the offender offends against is not my dignity an impersonal system.

In such an environment, the conduct in which we act on and seek to protect our sense of our own worth is somewhat less likely to take the form that is familiar to us in tales of Rennaissance Italy or the Wild West, in which fight with others over offences against them. It is at least a little more likely to take the form of conduct that is that is more productive of good and less adverse to forgiveness and related traits. It can show itself in our efforts to achieve our highest goals, and to conduct ourselves in ways that suit our dignity as human beings.

Chapter IX


1. Thomas Paine, Common Sense, in Tracts of the American Revolution, ed. by Merrill Jensen (Indianapolis: Bobbs-Merrill, 1977), pp. 402-403.

2. Common Sense, p. 403.

3. Will does not actually say this, but his book on this subject, which is otherwise very intelligently written, would be a massive non sequitur without it. See George F. Will, Statecraft as Soulcraft: What Government Does (New York: Simon and Schuster, 1983).

4. Friedrich Nietzsche, On the Genealogy of Morals in Basic Writings of Nietzsche, trans. and ed. by Walter Kaufmann (New York: The Modern Library, 1968), pp. 498 and 515.

5. Non-minimal states must claim authority because, unlike minimal states, they require us to do things which we had no reason to do before the state began to require them. Everyone has reasons for being generally helpful to people in need, but a government which requires us to help the needy through our taxes requires us to give a certain amount of money to a certain organization to help the needy, and that is something we had no reason to do before the relevant laws were passed.

6. These are among the "inconveniences" of the "state of nature" - that is, of a world without government - named by John Locke and Robert Nozick. See Nozick's Anarchy, State, and Utopia (New York: Basic Books, 1974), pp. 10-11.

7. Adam Smith, The Theory of Moral Sentiments (Indianapolis: Liberty Classics, 1976), p. 92.

8. John Locke, The Second Treatise of Government, section 125.

9. Lon Fuller, The Morality of Law (New Haven: Yale University Press, 1969), pp. 81 ff.

10. The Theory of Moral Sentiments, pp. 138 and 181.

11. On the Genealogy of Morals, p. 512.

12. These remarks may be read as a criticism of the "individualist" sort of anarchism defended, among others, by David Friedman in Part III of The Machinery of Freedom (New Rochelle, 1978). Some individualist anarchists sidestep the sort of problem I am pointing out here by saying that in their system punishment would not exist. Its place would be taken by compensation of the sort found in our current system of tort law. Once that move is made, I suppose the issue becomes why they are so sure that retribution will not exist in their system, but that lies outside the constellation of topics I have limited myself to here.