Without Guilt & Justice – The Death of Retributive Justice

« Decidophobia | An Attack on Distributive Justice »


THE ROAD to autonomy is blocked by a two-headed dragon. One head is Guilt, the other Justice. Justice roars: “You have no right to decide for yourself; you have been told what is good, right, and just. There is one righteous road, and there are many unrighteous ones. Turn back and seek justice!”

Frightened, man stops and marvels at his own presumption, when Guilt cries: “Those who succeed in getting past Justice are devoured by Guilt. Seek the road to which Justice directs you and dare not to strike out on paths of your own. Guilt has a thousand eyes to swallow you, and the lids above and below each are lined with poison fangs. Turn back: autonomy IS sacrilege.” Whoever wants to reach autonomy must first slay this dragon and do battle with Justice and Guilt. But Justice has many wiles and is not always as fierce as her roar. She can change herself into a beautiful woman – no longer young, to be sure – and say: “In my youth the Hebrew prophets loved me, and Plato sang my praises. Christianity taught generations to think of me as divine and linked me to God’s righteous judgment of all men. When faith in God declined, philosophers of widely different views tried to dissociate me from religion and linked me with reason.” At that point the voice of Justice becomes less wistful; she continues firmly, even imperiously: “Legions now no longer appeal to God as moral arbiter; they invoke me.”

Asked how they know what is just, large numbers of people would deny that they were relying on their upbringing or on their personal intuition; they would insist that their claims were rational and that any reasonable person who was not blinded by prejudice could see what was just. People who think that way are decidophobes who have fallen for the seventh strategy: moral rationalism.

They believe that there is one righteous way and that justice demands one particular punishment or one specific distribution. There is no need to weigh alternatives and then make difficult decisions; there is no room for excruciating choices: reason – they think – tells us plainly what we ought to do. Formerly it was religion that was thought to tell us what was just, and God was the ultimate authority. With the death of God, the prestige of justice rose, if only temporarily, and now she receives some of the reverence hitherto reserved for God.

Justice is widely held to be objective if not absolute, precise and not subject to emotion, timeless and above mere preferences. These attributes are crucial features of what people mean by justice. When justice demands something, it is no longer up to mere human beings to try to decide what to do; the individual is supposed to submit and do the bidding of justice.

In fact, justice is not at all timeless. Yesterday’s just punishment or distribution may be considered blatantly unjust today. Soon I shall give examples to show how what was once demanded in the name of simple justice is now felt to be outrageous. To the skeptic, any claim that “justice has been done” looks arrogant and foolish right away. A generation or two later, it will also look absurd to those who are not skeptics and who use the same rhetoric themselves. I shall argue that the demands of simple justice are simple indeed but not just.

There is more than one way in which justice is not timeless. It will be helpful to distinguish several stages in the development of the idea of justice.

First, justice was conformity with custom, and injustice meant a violation of tradition. Even at this stage, justice blocked the road to autonomy; it was not up to the individual to make fateful decisions for himself. An extraordinary passage in the Iliad seems to illustrate this stage. Menelaus is about to take a Trojan prisoner of war in order to collect ransom, when Agamemnon reproaches him: “No let us not leave even one of them alive, down to the babies in their mothers’ wombs – not even they should live. The whole people must be wiped out of existence, and none survive to think of them or weep.” And Homer continues: “He turned his brother’s heart, for he urged justice.” This seems to mean that he reminded his brother of the hallowed custom of genocide. In time, the tragic poets and the Sophists questioned the authority of custom and convention, and Euripides attacked this particular custom in The Trojan Women.

At the second stage, justice becomes the sum of the virtues. The classical formulation is found in another Greek poet, Theognis, in the second half of the sixth century B.C.:

Justice contains the sum of all virtue,
and every just man, Kyrnos, is good.

This is a development of the first stage and originally meant that those who conformed to custom were good. But when convention was felt to be problematic, a higher law was postulated both in Athens and in Jerusalem – what later came to be known as “natural law” – and whoever lived in conformity with that was considered just. This second stage was consummated by Plato and the Hebrew prophets.

The third state is reached when justice becomes a particular virtue. Aristotle, in the fourth century, expressly distinguished the justice that is the sum of the virtues from the justice that is one of the virtues. Further, he distinguished distributive and rectificatory justice, associating the latter with restitution – not with retribution.

When justice is no longer primarily a virtue but rather a quality of punishments and distributions, the fourth stage has been reached. On the rare occasions when a person is still called just at this stage, this is either an archaism that harks back to the second stage, and the meaning is that he is, in Hebrew, a tsaddik – a man who is just in the sense that he has all the virtues – or what is meant is that the distributions he makes or the punishments he imposes are just. In the modern age, justice is primarily a predicate of punishments and distributions, and the ascription of justice as a virtue to individuals is derivative.

The conception of justice that underlies retributive and distributive justice is the same: distributions and punishments are considered just when each gets what he deserves, and unjust when this is not the case. In other words, justice consists of meting out to men what they deserve. When they are punished because they are held to deserve evil and suffering, one speaks of retributive justice. When what is distributed is good one speaks of distributive justice. I propose to criticize this kind of justice – retributive justice in the remainder of this chapter, distributive justice in the next. The notion that distributive justice is better understood as fairness will also be taken up in the next chapter.

The four stages in the development of the idea of justice are not so distinct that one could say when each began and ended. Even when justice was considered the sum of the virtues, giving each what he deserved was often held to be an especially important part of it; and when justice became a particular virtue it came to consist more and more of punishing and rewarding every man in accordance with his deserts. God embodied this kind of justice, and in the Last Judgment it was made manifest to all. But now the fifth stage is upon us: the death of retributive justice.

Retributive justice has been subjected to so much criticism that one might suppose the time had come to say that one should not speak ill of the dead. But the fifth stage, in which we are living, is a time of moral confusion. The aversion to retributive justice is rather sentimental, and even philosophers who find her utterly repugnant still cling to distributive justice as if that were an entirely different matter. They go to great pains to dissociate distributive justice from her ugly sister, who is dying; they speak of distributive justice as if she alone had a right to the name; and they outdo each other in their reverence for her.

Consider two of the most respected moral philosophers in the United States. William Frankena calls retributive justice “quite incredible,” but considers distributive justice one of “the two ‘cardinal’ moral virtues,” along with benevolence. John Rawls writes A Theory of Justice and devotes a single page out of six hundred to retributive justice – merely to reject out of hand the notion that it is “the opposite” of distributive justice. This page is not marked by the subtlety that distinguishes much of the rest of the book: “A propensity to commit such acts [i.e., ‘acts proscribed by penal statutes’] is a mark of bad character, and in a just society legal punishments will only fall upon those who display these faults.” I shall argue that punishment should be dissociated completely from any judgment of character. Rawls is not arguing at all at this point: he is merely trying to dismiss retributive justice. But as long as retributive justice is ignored, the nature of justice can hardly be understood.

I want to do my best to usher in stage six: the death of distributive justice. To accomplish that, one must begin at the stage in which we live now, the fifth, and consider the death of retributive justice first. For the case against distributive justice closely parallels that against retributive justice. The faith in retributive justice is going fast; and distributive justice cannot long survive the death of her Siamese twin.


In the Old Testament and the New, in Judaism and Christianity as well as Hinduism, retributive justice has always been of the essence of justice, and it has actually tended to overshadow distributive justice. To this day, the claim that “justice has been done” brings to mind punishment first of all, and the phrase “justice demands . . .” is as often completed with a specific penalty as it is with something good. God’s justice was always held to consist in large measure of his punishment of the wicked. That so many evildoers flourished raised doubts about God’s justice, but eventually the justice of God would be made manifest to all as each received what he deserved. Human justice was held to consist of an emulation of the divine judgment, and it was often very cruel indeed.

Rawls insists on his own Kantianism but quite overlooks that even in the Philosophy of Right (Rechtslehre, 1797), which has been translated into English under the title The Metaphysical Elements of Justice, Kant hardly ever speaks of justice (Gerechtigkeit), except in the section on punishment For him justice meant pre-eminently retributive justice. He did not even speak of justice when in his postulate of God’s existence (1788) he implicitly demanded distributive justice in the hereafter.

Hegel also published a Philosophy of Right without ever discussing justice at any length; and he, too, found justice above all in punishment. Distributive justice has never held the place in German moral philosophy that it obtained in British and American ethics. Karl Marx stood squarely in the German tradition and pleaded not for distributive justice (Robert Tucker has shown this conclusively) but for self-realization and, in a sense, autonomy. (I shall return to Marx in the chapter on alienation; for what he fought against was not distributive injustice but alienation.) The modem liberal champions of distributive justice tend to ignore retributive justice, but before our own time almost everybody except David Hume recognized that retributive justice was of the essence of justice.

This recognition was by no means confined to Catholics, Calvinists, and Kant’s heirs. Take Thomas Jefferson, the very model of an “enlightened” opponent of Calvinism and Catholicism. When Napoleon was in St. Helena, Jefferson said of him, in a letter:

The penance he is now doing for all his atrocities must be soothing to every virtuous heart. It proves that we have a god in heaven. That he is just, and not careless of what passes in this world. And We cannot but wish to this inhuman wretch, a long, long life, that time as well as intensity may fill up his sufferings to the measure of his enormities. But indeed what sufferings can atone for his crimes . . . !

The final exclamation suggests the limits, if not the absurdity, of the dream of proportionality. Yet the notion that justice is done only when every crime is punished proportionately is extremely widespread. Jefferson shows this, too. In his First Inaugural Address he proposed “Equal and exact [!] justice to all men.” To his mind this did not entail the abolition of slavery. But he had spelled out some of his relevant ideas in 1779 in “A Bill for Proportioning Crimes and Punishments,” which ends: “Slaves guilty of any offence punishable in others by labor in the public works, shall be transported to such parts in the West Indies, South America, or Africa, as the Governor shall direct, there to be continued in slavery.”

In Jefferson’s Bill, “petty treason” and murder are to be punished by death, and “Whosoever committeth murder by poisoning shall suffer death by poison.” “Whosoever shall be guilty of rape, polygamy, or sodomy with man or woman, shall be punished, if a man, by castration, if a woman, by cutting through the cartilage of her nose a hole of one half inch in diameter at the least.” And

Whosoever on purpose, and of malice forethought, shall maim another, or shall disfigure him, by cutting out or disabling the tongue, slitting or cutting of a nose, lip, or ear, branding, or otherwise, shall be maimed, or disfigured in like sort: or if that cannot be, for want of the same part, then as nearly as may be, in some other part of at least equal value and estimation, in the opinion of the jury, and moreover shall forfeit one half of his lands and goods to the sufferer.

Many people nowadays associate “an eye for an eye” with the Old Testament, and many liberals believe that this conception of justice was transcended in the New Testament. Often one goes on to associate the former with justice and the latter with love, and one contrasts the laws of Moses not with the laws of the Christian Middle Ages, some two thousand years later, or with the laws of Christian countries in the eighteenth and nineteenth centuries, but with the most edifying dicta about personal conduct in the Sermon on the Mount. Yet this Sermon is studded with promises of rewards and punishments, and the Gospels are punctuated by threats of judgment, damnation, and hell. It is also in the Sermon on the Mount that Jesus offers the classical formulation of a notion of justice that embraces retribution as well as distribution: “the measure you give, shall be the measure you get.” But the distinctive conception of justice in the New Testament is that on the day of judgment few will be saved from eternal torment. The idea that any man – not even to speak of most of mankind – should be punished with eternal torture is so repugnant to liberals that millions refuse to acknowledge its presence where it stares them in the face.

A typical liberal reaction to Jefferson’s “Bill for Proportioning Crimes and Punishments” would be to consider it a relapse into Mosaic cruelties. It would be less out of touch with historical fact and Jefferson’s spirit to see it as an advance over the cruelty of his own time and of the Gospels.


In eighteenth-century England the punishment for treason began with hanging; then the offender was taken down while still alive and his entrails were cut out and burned before his eyes; and then he was beheaded and quartered. In 1694 an attainder for treason was reversed by the King’s Bench after a man had been duly drawn, hanged, and quartered because the judge, after saying that the traitor’s entrails were to be cut out, had omitted the words “and burned while he is still alive.” To rule that this phrase could be left out because a man would scarcely survive after his entrails were cut from his belly, said the King’s Bench, would make “judgments in high treason . . . discretionary, which indeed is only a softer word for arbitrary.”

This might invite all kinds of decisions, including “a Jewish judgment, ‘that the offender should be stoned to death’; or a Turkish judgment, ‘that he should be strangled’; . . .or a French judgment, ‘that he should be broken on the wheel. . .”’

In France two men were broken on the wheel for petty theft as late as 1770; one had stolen a piece of cheese. In England, a nine-year-old child was sentenced to death in 1832 for smashing a window and stealing two-pence worth of paint. It was only in 1837 that two hundred offences hitherto punishable by death ceased to be capital crimes in England.

Paul Reiwald is right when he says:

Men were not even able to confine themselves to the law of talio, the law of eye for eye, tooth for tooth, which seems so primitive and barbarous to us. In truth, the principle of an eye for an eye, a tooth for a tooth, with which the Jews are occasionally reproached to this day because it is held to be typical of their God of vengeance, belongs with the great and decisive advances of humanity, as is now generally recognized in the world of scholarship.

Reiwald is also right when he says in his discussion of medieval punishments: “The barrier of the lex talionis is torn down. Compared to what was actually done, its application would have signified gentle mercy.”

As for the Gospels, consider how Jesus comforted his disciples:

If anyone will not receive you or listen to your words, shake off the dust from your feet as you leave that house or town. Truly, I say to you, it shall be more tolerable on the day of judgment for the land of Sodom and Gomorrah than for that town. Behold I send you 01.&, as sheep in the midst of wolves . . .

But these lambs pack some clout! Anyone who does not care to listen to their preaching will be punished worse than the most notorious evildoers of all time. It is of the essence of liberal Christianity that it feels sure without any need for evidence that Jesus did not really say this and that Matthew and Luke must have misquoted him. That there are many similar passages, also in the other two Gospels, one does not recall any more than one remembered these sentences. Whatever the evidence, one simply knows that Jesus did not say any such thing.

There is a Manichaean streak in the Gospels: the enemy is beyond the protection of proportionality. If he will not listen, the worst punishment is still too good for him. Liberals are appalled by such cruelty and favor a sense of proportion, certainly as far as rewards are concerned; but when it comes to punishment, they are confused. It certainly should not be disproportionate. But liberals prefer to think of justice as having nothing to do with anything as unpleasant as punishment. In one context they uphold the superiority of love and speak of justice as transcended by Christianity. In another, they are all for justice and associate it with causes they believe in. In short, many are against retributive justice but for distributive justice.

It would make no sense to saddle either the New Testament or the King’s Bench in England or the lawmakers of the Middle Ages with moral rationalism. Here we are still in the realm of religion with God’s awesome justice as the model. But some of the men of the Enlightenment sought to counter the inhumanity of their Christian predecessors with appeals to reason. They thought that retributive justice had a mathematical quality and that murder called for capital punishment in much the same way in which two plus two equals four.

Not only Jefferson tried to find proportionate punishments for other crimes. In a similar spirit, Kant tried to prove that reason requires thieves to be sentenced to forced labor in a penitentiary:

Whoever steals makes everybody else’s property insecure; he thus robs himself (in accordance with the law of retribution) of the security of all possible property; he has nothing nor can acquire anything but still wants to live, which is not possible unless others feed him. But since the state will not do this for nothing, he has to place his powers at the disposal of the state for whatever labor it deems fit . . .

Such sophistry is the direct consequence of the attempt to approximate morality to mathematics: “What kind and what degree of punishment does public justice adopt as a principle and standard? None other than the principle of equality (the little tongue of the scales of justice) . . .” Against those who even in Kant’s time were arguing against capital punishment he insists:

Even if civil society were to dissolve with the full agreement of all its members (e.g., a people on an island resolved to scatter over all the world), the last murderer still confined in prison would first have to be executed in order that everybody received what his deeds deserved, lest a blood guilt should stick to the people that had not insisted on this penalty . . .

Most modern readers will find this as uncongenial as Jefferson’s “Bill for Proportioning Crimes and Punishments.” Those who revere Kant or Jefferson will find much of this downright embarrassing. Why? Because the faith in retributive justice is all but dead. What are the causes of its death? The answer to this question will also show why today, for the first time in human history, autonomy has become a live option for millions.


The first phase of the movement that is leading to the death of justice might be called moral skepticism. This could be traced back to the Greeks; Plato could be seen as a reaction to it, and Christianity as a great countermovement. But for our purposes it will suffice to consider the familiar resurgence of moral skepticism in the modern era. This development is so well known that we need only recall very briefly a few of its major elements. First, religion lost its authority in moral matters for most of mankind. Then, the habit of considering alternatives and weighing pros and cons spread with the rise of modern science; and when this approach is applied to moral claims, the result is moral skepticism. The development of comparative sociology and anthropology has done its share to make this explicit. So has the study of comparative religion – this, too, is a way of considering alternatives – in spite of the last-ditch “holy lie” of some decidophobes that all the great religious teachers taught the same morality.

On another plane, it has become more and more unusual for all the children in a family to stay in the town where they were born. People are exposed to different environments and mores. Tens of millions have been uprooted and moved since World War II, and vast numbers have left farms and villages and small towns for big cities. Travel has also proliferated; hitherto isolated people who are far too poor to travel are suddenly brought face to face with foreigners; and magazines, films, radio, and television have done their share to expose men to different value systems.

Our moral philosophers have on the whole been more conservative than many of their students. It makes a difference if one has grown up in a relatively stable environment, under the tutelage of parents and teachers who were still closer to absolutism. Those who grew up after Auschwitz and Nagasaki cannot recall a “normal,” stable world. That many students became Manichaeans in the 1960s and reverted to a form of moral absolutism – in some cases, moral rationalism – was due to the fact that they had traveled further down the road of skepticism and had reached nihilism and despair. The wars in Vietnam and Algeria and the slaughter in India and Pakistan, in the Congo, Indonesia, and Nigeria, and the world’s reaction, made a mockery of the morality to which Western societies as well Asian and African governments paid lip service. These vast atrocities and the numbing anonymity of metropolitan life had contributed to a desperate sense of futility. Many had gone beyond moral skepticism into moral nihilism: from the reasonable position that whatever we do is not likely to make any difference a thousand years hence, they inferred fallaciously that it therefore made no difference – not even now. It was from this nihilistic despair that some students sought salvation in a new absolutism.

Skepticism about natural law is implicit in moral skepticism. The very concept of natural law is not widely familiar – philosophers, theologians, and lawyers know it; few others do. Not only is the term mildly esoteric, but the idea that a single moral law is binding for all men, regardless of time and place, lost its plausibility as moral skepticism spread. Few except Catholics still cling to this notion, and fewer and fewer Catholics do. Even many Catholic theologians now defend the Inquisition by saying that it was justified in its time but would not be justifiable today.

The man who did most to promote skepticism about positive law – the law actually in force in a state – was Hitler. The war crimes trials from Nuremberg to Jerusalem convinced millions that obedience to the laws of the state in which one lives is by no means always one’s duty. What a few had learned earlier from Sophocles’ Antigone, Tolstoy, Thoreau, or Gandhi, millions learned from these trials. Some learned it directly, as it were; but there was no lack of mentors.

One of the most influential of these was Sartre. During the Algerian War he kept exhorting intellectuals to speak out and defy their government. He reached an international constituency. At the same time Martin Luther King’s civil disobedience campaign in the American South did its share to shake the faith in positive law. King had taken his doctorate in philosophy, but again it would be misleading to understand the change in attitude in purely intellectual terms. In the United States, for example, the Draconic laws against possession of marijuana and other drugs carried masses of young people beyond skepticism about law into downright contempt for law. The prohibition of alcohol in the 1920s had had a similar effect. But the constellation of incomparably more severe penalties with the civil rights struggle and opposition to the war in Vietnam made this new contempt far more impassioned.

In time, moral skepticism will be seen to entail doubts about justice, but so far skepticism about distributive justice is not yet widespread at all. Why, then, is retributive justice dying? In addition to the historical developments summarized here, three major points are worth stressing.


First, attitudes towards criminals have changed to the point where the demand not to hate them but to remain mindful of their humanity no longer sounds utopian. This change is due in no small measure to some nineteenth-century novelists. Charles Dickens and Victor Hugo come to mind along with Dostoevsky and Tolstoy. Their depiction of suffering was nothing new, and the image of prison conditions in the nineteenth-century novel is no more cruel than much that can be found in earlier literature, including Dante’s Inferno. What is distinctive is the novelists’ attitude toward these conditions and the sympathy for the criminals that is evoked in the reader. The culmination of this movement is reached at the turn of the century in Tolstoy’s Resurrection, the novel for which he was excommunicated by the Orthodox Church in 1901.

Second, we have developed a kind of second sight. To say that we have become more perceptive in psychological matters would be an understatement, not because our age is so perceptive, which it is not, but rather because the psychological obtuseness that prevailed until quite recently is almost unbelievable. Again, Dostoevsky and Tolstoy deserve much of the credit for this change, along with Nietzsche and, above all, Freud.

To tear down the wall that respectable people had built up between themselves and those who were “abnormal,” these writers approached it from two sides. Unlike Dostoevsky and Tolstoy, Freud did not think much of the dictum that one ought to love one’s enemies, but far more than any Christian saint or theologian, he showed that our enemies, and criminals for that matter, were not essentially different from ourselves. One did not have to accept his theories in detail to be strongly affected by this implication of his work.

The other approach to the wall is much less obvious. In Paul W. Tappan’s massive standard text on Crime, Justice and Correction, for example, all ten references to Freud (in seven hundred fifty pages) concern the light he shed on criminals. But Freud – like Nietzsche, whom Tappan does not mention at all – also turned a searchlight on respectable society, illuminating the unedifying motives that come to the fore in punishment. Not only is the criminal a human being like you, but you, alas, are like the criminal.

It is not surprising that this insight is much less popular than the first approach, which goes so well with the liberal faith in humanity. But even where this second approach is not accepted explicitly, it has come to color our way of thinking. When one reads a typical defense of retributive justice by a nineteenth-century philosopher who claims that “Indignation against wrong done to another has nothing in common with a desire to revenge a wrong done to ourself,” one is struck by its psychological naïveté. Philosophers, to be sure, always have a tendency to stick to words and thrive on neat conceptual distinctions, but after Freud even a philosopher may be pardoned for asking how “indignation against wrong done to another” looks in practice.

Consider the institution of the pillory. The following report from the British Morning Herald of January 28, 1804, is typical:

The enormity of Thomas Scott’s offence, in endeavouring to accuse Captain Kennah, a respectable officer, together with his servant, of robbery, having attracted much public notice, his conviction, that followed the attempt, could not be but gratifying to all lovers of justice. Yesterday, the culprit underwent a part of his punishment: he was placed in the pillory, at Charing Cross, for one hour. On his first appearance, he was greeted by a large mob with a discharge of small shot, such as rotten eggs, filth, and dirt from the streets, which was followed up by dead cats, rats, etc., which had been collected in the vicinity of the Metropolis by the boys in the morning. . . .

Here we see “indignation against wrong done to another” in action. Of course, such indignation does not always look like this, and soon I shall analyze the various functions of punishment. But what I have quoted is not an odd item about the behavior of a crowd that has got out of hand; it is rather a representative description of a form of punishment that was quite popular for a long time. Why are we unconvinced by all attempts to prove that what was meted out to Thomas Scott was exact justice? He had endeavored to undermine respect for Captain Kennah; now he was subjected to loss of respect. Similar arguments could be offered for other Curious Punishments of Bygone Days, to cite the title of a book by Alice Morse Earle (1896) which is set in America. Her chapter headings give some idea of the contents: the ducking stool, the stocks, the pillory, the whipping post, the scarlet letter, branks and gags, branding and maiming. Branks were an iron frame placed over a woman’s head, with a sharp metal bit entering the mouth, and were used to punish scolds. One would not need the subtle ingenuity of Kant to show that this punishment was in a way appropriate and not by any means completely disproportionate; and yet few readers nowadays would concede that simple justice demanded it. Why not?

Or consider one of the oddest passages in Curious Punishments: “Truly long hair and wigs had their ulterior uses in colonial days when ear-cropping was thus rife. . . . Life was dull and cramped in those days, but there were diversions; when the breeze might lift the locks from your friend’s or lover’s cheek and give a glimpse of ghastly hole instead of an ear. . .” (sic). Or this: “One woman at the whipping post ‘created much amusement by her resistance.’ ” We do not even ask for what crimes ear-cropping or whipping might be proportionate, though it stands to reason that a Jefferson or Kant might have come up with a thoughtful answer. Why?

The answer to these questions has been given above: skepticism about positive law, sympathy for and even identification with the criminal, and a horror of the unedifying motives that find expression in punishment. But one final point may help to illuminate the other developments considered here. The death of retributive justice is linked to the death of God.
As long as men believed in the Last Judgment and in hell, they could hardly question retributive justice. Pope Pius XII made this point plainly and emphatically when he addressed the Sixth International Congress of Penal Law, October 3, 1953: Against “modern theories” that “fail to consider expiation of the crime committed. . . as the most important function of punishment” he cited Matthew 16:27 and Romans 2:6 and 13:4, concluding:

The function of protection disappears completely in the after-life. The Omnipotent and All-Knowing Creator can always prevent the repetition of a crime by the interior moral conversion of the delinquent. But the supreme Judge, in His Last Judgment, applies uniquely the principle of retribution. This, then, must be of great importance.

As long as traditional Christianity flourished, retributive justice did, too. When the faith in hell and the Last Judgment lost its grip, Jefferson and Kant, as well as other writers, still tried to save the faith in retributive justice by providing a new, rationalist foundation for it. While men still had the old religious faith in their bones, such efforts seemed to have some plausibility; but no more. Millions realize that neither God nor reason has determined once and for all what each person deserves and that it is up to us to weigh alternatives and to make difficult decisions.


To decide whether and when punishment is needed, one must first of all be clear about what precisely punishment is and what its functions are. It is such a familiar institution that most people never realize how subtle it is.

Punishment involves at least two persons (call them A and B) and two acts. A holds a position of authority in relation to B, claims that B has done some wrong, and by virtue of his authority causes something unpleasant to happen to B in return for (as a punishment for) this claimed wrong. This is what is meant by punishment. If A does not claim that B has done some wrong, one speaks of maltreatment or torture, not of punishment; and if A does not hold a position of authority one speaks of revenge.

B could be an animal, but only if A treats B more or less as a person. Thus B could well be a dog or a cat; but we do not call it punishment when we kill a mosquito that has just bitten us. If A and B are one and the same person and we say, “Why do you keep punishing yourself?” we are using the term figuratively but still in a manner that is wholly consistent with our explication: B assumes the role of A and punishes himself. Finally, A could be a deity who in that case would act more or less like a person – specifically, like a father, a judge, or possibly a teacher.

What is the purpose of this institution of punishment? It is encountered in many, if not all, societies and is used not only by political authorities but also by parents and teachers and even in games. Its ubiquity makes a mockery of any search for “the purpose,” as if there were always one purpose only, the same everywhere. In different societies, contexts, and ages, punishment served various functions. Its entertainment value was more important in some places than in others. But the desire to see justice done, to do to the offender what he deserved, was never the primary reason for instituting punishments. The primary purpose of proclaiming a penal code is to prevent some evil. But this does not mean that the penalties are intended solely for deterrence. In Deuteronomy 19, “eye for eye” is actually introduced: “The rest shall hear and fear, and shall never again commit any such evil in your midst.” Deterrence is very important indeed, but often understood far too narrowly.

A penal code deters people from committing crimes not only (1) by engendering fear but also (2) by inculcating a moral sense. A trivial penalty (say, a five cent fine) suggests that an offense is trivial, while a severe penalty conveys the sense that the crime for which it is decreed is grave. The code may also deter people simply (3) by informing them of what is forbidden. At first glance, it may seem to be overly subtle to distinguish this function from the first two. In fact, in many cases one is neither frightened nor led to feel that anything is immoral, and it is quite common for people to know that certain acts are forbidden without having any idea what penalties have been decreed for offenders. In such cases the third function is in evidence, but not the first two. But crimes occur in spite of all this, and the penalties are intended to undo, or at least to minimize, the damage. How?

  1. By preventing private vengeance, lynchings, and a general breakdown of order. Often the offense injured others who, in the absence of a penal code, might have taken the law into their own hands.

  2. By seeing to it that the breaking of a law does not become an invitation to other men to emulate the lawbreaker. The punishment is meant to deter others and thus to re-enforce the code. The offender has weakened the law and come close to annulling its deterrent effect; now the punishment is meant to undo this negative consequence and thus to restore the deterrent effect.

  3. By providing a safety valve for the unlawful desires that smolder below the surface and are fanned to the danger point by the commission of a crime. Many people have wanted to do what the criminal did but were kept from doing it by the law or by their conscience. Now he makes them look silly; they were timid, he was bold; they were weak, and he was strong – if he gets away with it. And he seems to have gotten away with it. Hence many people are burning to de what he did. The penal code provides an outlet for this criminal desire. He has killed someone, ‘and now you – many of you – also want to kill? All right; kill him! He has maimed someone, and now many of you also want to maim someone? All right; maim him! Thus the desire for talion – for doing to the criminal what he has done to someone else – does not evidence any profound sense of justice or a primordial conviction that this is clearly what the criminal deserves.

These last three functions (4-6) interpenetrate. But the desire to proportion punishments to crimes is not born of the feeling that anything less than this would not be justice; it represents an attempt – as in Jefferson’s case – to keep cruelty in bounds. For as soon as people are invited to vent their criminal desires on the criminal, the same dangers reappear that we have just considered (under 4 and 5): as long as he is to be killed in any case, why merely kill him? Why not. hang him first, then take him down alive, cut out his entrails. . . Why not have an orgy? Historically, the call for talion has generally signified a great advance over wanton cruelty (see page 44 above).

The fourth and fifth functions still come under the heading of deterrence. The sixth might be called cathartic, to use an ugly word for an ugly fact. Punishment purges the society – not, as often claimed, by removing some mythical pollution, but in a more palpable psychological sense. The purge, of course, affords only temporary relief, and unfortunately there is evidence that it is addictive. But this function of punishment has often been mistaken for a demand for retributive justice.

The traditional distinction between three functions of punishment – deterrence, reform, and retribution – is not subtle enough. One should distinguish ten functions – four more in addition to the six considered so far.

  1. Punishment is often justified as a means of reforming the offender. Thus a child is punished to teach him a lesson and to make him a better person. Lawbreakers have been pilloried, whipped, sent to prison, branded, maimed, and fined to re-educate them. Hardly solely for that purpose, but we need not doubt that this was often held to be one aim of punishment – and more rarely also one function of punishment.

  2. Recompense or restitution is scarcely a punishment as long as it is merely a matter of returning stolen goods or money. But suppose one has insulted another person and is required to make a public apology, or one has to make up to someone else some other form of humiliation, inconvenience, or suffering. When the offender is humiliated, inconvenienced, or made to suffer in turn because this is held to be some recompense for the offended party, we enter the realm of punishment. Similarly, when it is claimed that the lawbreaker has harmed society and must now pay his debt to society, recompense is invoked as the purpose of punishment. The point is not that the offender deserves to suffer; it is rather that the offended party desires compensation. Again, the various functions often interpenetrate.

  3. Expiation is also a form of recompense, but here the underlying idea is that some god has been offended and must be appeased. The notion of expiation depends on religious beliefs and makes no sense apart from them. Here I am sticking closely to the traditional meaning of “expiation.” If it were objected that the notion also makes sense in relation to a sovereign, a parent, or anyone at all who sees himself as standing in God’s place, I should say that such cases are best included under number 8.

  4. Finally, there is the claim that justice requires retribution, and that justice is done when, and only when, the offender is punished: he deserves to be punished, and until he actually is punished he fails to get what he deserves. This claim, which figures prominently in the rhetoric about punishment, is open to several criticisms:

a. The notion of desert is questionable and will be criticized at length in the next two chapters.

b. The first seven functions are clearly future-oriented. The eighth (recompense) is at least partly future-oriented, but it also hinges on the notion of desert. The ninth (expiation) is a variant of the eighth that introduces the supernatural. But retribution is past-oriented. This contrast of two orientations and my objections to any such fixation on the past will be developed in the chapter on guilt. Specifically, this claim (10) is frequently based on the conviction that a past event needs to be – and can be – undone. This is a superstition. The past is not a blackboard, punishments are not erasers, and the slate can never be wiped clean: what is done is done and cannot be undone.

c. The intuitive certainty that nevertheless often accompanies the belief that an offender fails to get what he deserves until he is punished will be explained in the chapter on the birth of guilt and justice.


The decidophobe loves retributive justice because she tells him precisely what is to be done: wrongdoing must be punished, and there is one penalty that is just and therefore mandatory. But I say:

  1. Punishments can never be just.

  2. Even if a punishment could be proportionate, it would not follow that it ought to be imposed.

  3. The preoccupation with retributive justice is inhumane.

The first thesis means that a punishment can never be deserved or who11y proportionate. If the nine-year-old child sentenced to death in 1832 for smashing a window and stealing two-pence worth of paint had actually done these things, and if the penalty conformed with precedent and custom, that would not entail that the punishment was deserved and just. The same goes for a man broken on the wheel for stealing a piece of cheese.

Jefferson plainly felt that at least some of the punishments provided in existing penal codes were unjust; but he believed in the possibility of “proportioning crimes and punishments.” To be just, a punishment must satisfy three conditions: the accusation must be proved; the punishment must accord with precedent and custom; and the punishment must be proportioned to the crime. Our clear sense that some punishments are outrageous even though they satisfy the first two conditions results from the feeling that they seem out of all proportion to the crime. (One could introduce further complications by stipulating how the accusation must be proved; for example, in accordance with established procedure, without recourse to torture, and so forth. But this would take us too far afield.)

The first two conditions concern particular instances of punishment. The third condition, proportionality, concerns the penal law and is far more interesting. The crucial point is that the admission that some punishments are cruel and unusual does not commit one to the view that for every crime – or even any crime – there is a proportionate and hence deserved and just penalty. Indeed, it seems very plain that for some crimes there is not, and I shall try to show in the next chapter that there is no just punishment for any crime.

To begin with crimes for which there is clearly no proportionate punishment: how could one possibly establish what a man deserves for seducing a child, for raping a child, or for arson or treason? The question of how one should deal with such crimes calls for excruciating decisions. The moral rationalist avoids the frightening task of weighing alternatives; he claims that reason demands such and such a penalty, backs up his claim with a proof a la Kant, and shuts his eyes to objections and alternatives. The moral irrationalist relies on authority, most likely on God’s revelation or the law, and then engages at most in exegetical thinking. The autonomous human being uses his reason to eliminate various alternatives, but finds that after this he is still left with several tenable positions between which he must make a choice. He may have little doubt that his choice is better than many that are clearly inferior, but he will not have the arrogance to claim that the penalty he chooses is the one that is proportionate, deserved, and just.

This question about desert is as difficult as it is important. It is as relevant to distributive justice as it is to retributive justice, and I shall deal with it more fully in the next chapter.


When Adolf Eichmann was kidnapped to stand trial, a truthful verdict was possible, a just punishment was not. Still, a punishment can be more or less inappropriate. Thinking in terms of degrees like this is anathema to the Manichaean, who likes to insist that a punishment is either just or unjust. He dreads being confused by multiple choices. Putting a child to death for stealing two-pence worth of paint may be crueler than cutting off its right arm, and perhaps giving it two hundred lashes is not quite as outrageous as maiming it, but it is hardly just. Fining the child a shilling and then getting it a job at which it can earn that much might make more sense. But is that what the child deserves, or might one find a preferable penalty?

Thinking in degrees of just and unjust is actually still far too Manichaean. Such one-dimensional thinking assumes that all possibilities can be arranged in a single sequence, on a linear scale. In fact, there are countless variables and endless possibilities.

The child was accused of a petty crime. Now consider Eichmann. Visiting on Hitler’s leading henchmen at least some of the tortures to which they had subjected millions of people, and all but putting to death these mass murderers again and again would have been more proportionate to their crimes than hanging them. But the punishment that is more proportionate and more nearly deserved is not necessarily preferable even on purely moral grounds. That is the point of my second thesis.

This thesis cannot be proved. The best way to back it up is to consider concrete cases, like those of Eichmann or, better yet, Hitler and Himmler, Stalin and Beria, and to ask whether the more nearly proportionate punishment for their systematic mass tortures and mass murders would necessarily be preferable. Those whose moral sense was formed by the doctrine of hell may say yes. Nevertheless, some intuitive grasp of my thesis is almost as old as criminal justice itself: justice is sometimes tempered by mercy, and there is the sovereign’s right to pardon.

This traditional way of taking my point into account is, however, utterly inadequate. It gives expression to a deep confusion. Selective mercy and selective pardon raise grave doubts about the cases in which it is claimed that justice has been done. They call into question the claim that in cases where mercy does not come into play justice was done.

Consider St. Augustine’s claim that all men deserve damnation; that God elects a few for salvation although they do not deserve it; and that the damned cannot complain that God is unjust. After all, says the saint, nobody is punished worse than he deserves, and the fact that a few fare better than they deserve merely shows the infinite mercy of God.

Such reasoning is specious. First, such arbitrary inequality of treatment is what philosophers call a “paradigm case” of injustice. For it is a necessary, though not a sufficient, condition of “just” treatment that like cases are treated alike. Second, Augustine’s God exemplifies anything but infinite mercy. In connection with this last point, consider Dante, whose concern with proportioning punishments to crimes was second to no man’s. He gave the most beautiful and eloquent expression to the traditional Christian view of justice. In his sublime inscription over the gate to the inferno he stressed the eternity of suffering – the word “eternal” recurs three times in the nine lines – before concluding:

Abandon, as you enter, every hope.

But it is the central triplet about hell that requires comment here:

Justice moved my Architect above,
What made me was divine Omnipotence,
The highest Wisdom and the Primal Love.

The power of Dante’s poetry in the original Italian evokes admiration, and almost twenty centuries of Christian teaching have helped to keep most readers from being struck by the enormity of this incredible perversion of the meaning of justice and love. The only parallel that comes to mind is bound to sound like blasphemy, but it requires some shock to awaken those who are not shocked by Dante’s lines and by the Christian view. Over the gate of Auschwitz those who entered saw the words: Arbeit macht frei – “work liberates.”

One can still wander about this camp for hours, walk through barracks, stare at mountains of shoes and hair, at ovens, and then see those words when leaving. Those who take language lightly and have no love for words may feel that this inscription adds nothing to the horror. Yet it is the ultimate in brazen cynicism and dishonesty – a final, almost unbelievable, affront.

The whole Third Reich lasted barely more than twelve years, Auschwitz only about three – a drop in the bucket compared to the eternal torments of hell. But what on earth could one liken to the Christian hell if not a concentration camp? And what to the Auschwitz inscription if not the infinitely more fateful claim that eternal tortures are compatible with, and were actually devised by, the greatest love that ever was – and by justice?

Augustine’s and Dante’s God does not really treat the mass of men in accordance with their deserts. But as long as men believed that he did and that this meant that eternal torments awaited most men after death, it made good sense to torture men for a few days or weeks if need be to save them from hell and to silence all who might endanger the faith and salvation of their fellow men.


These reflections on Dante lead to my third thesis, that the preoccupation with retributive justice is inhumane. But my analysis of the functions of punishment shows that this thesis does not entail any demand for the abolition of punishment. Punishments are needed, invocations of justice are not.

In deciding what to punish and how to punish, we should banish from our minds the chimaera of justice. The suggestion made by Rawls that “in a just society legal punishments will only fall upon those” who have a “bad character” is ill considered. Having a “bad character” is neither a necessary nor a sufficient condition of being punished legally even in a morally admirable society. It makes sense to punish people for parking violations, but it does not make sense to insist that those who have violated various parking regulations have thus shown that they are wicked. Parking laws, if sensible, are enacted to make for a better society: they should eliminate or reduce traffic congestion, or insure some turnover of cars to make it possible for many people to visit a certain area. The reason for instituting penalties is that a prohibition that is not backed up by any penalties is generally useless if there is any great temptation to disregard it.

When a person has been duly convicted of a violation of the law and punished in accordance with precedent, it does not follow that he deserved the punishment and that justice was done. He may be a very decent person who has more than enough troubles and ailments as it is, while many people who cause much more suffering to their fellow men go free and flourish. It is bad enough that we cannot dispense with punishments. We do not have to add insult to injury by claiming that the poor man who gets caught receives his just desert. Desert is out of the picture.

Of course, we can and should ask whether the prohibition is reasonable, and whether the penalties provided by the law are reasonable or excessive. The critical evaluation of a law is centered in three questions: What purposes does it serve? Are these purposes good? And does it serve them efficiently?

It is important to be clear about the purposes because the law must be judged in relation to them; and if it serves no purpose, it ought to be abolished. Whether the purposes are good will sometimes be a matter for debate, but debate is futile if it does not come to grips with purposes.

None of these points depends on the relative triviality of the parking illustration. If the purpose of a law were to prevent aggressive wars or the killing of unarmed civilians by armed forces, these aims might well win wide assent, although the definition of aggressive war and the application to specific cases might pose very serious problems. That would still leave the question of efficiency. If there were no penalties, the law would almost certainly be ineffective, but even if it provided stringent penalties, a la Nuremberg, it might still prove ineffective. If the purpose can be agreed on, reasonable discussion should center on the question of how the law can be made effective; how the crimes that we want to prevent can be prevented. Asking about the price one has to pay for probable gains is part of the question of efficiency. But one does not have to fret about what those breaking the law deserve.

It may be objected that if desert is out of the picture no good reason remains for not punishing the innocent. But that is not so, as I shall show in the chapter on guilt.

Jefferson’s and Kant’s quest for the just punishment for various crimes was ill advised, illusory, and inhumane. A man who steals a piece of cheese does not deserve to be broken on the wheel; neither does he deserve forced labor in a penitentiary, as Kant argued. It would be more sensible, fruitful, and humane to ask altogether different questions about penitentiaries; for example, whether the following claims made by two penologists are true:

Prison as organized today is a real sewer that continually pours into society a flood of pus and contagious germs of a physiological as well as moral kind. It poisons, dulls, depresses, and corrupts. It is a factory that simultaneously produces the tubercular, madmen, and criminals.

A few semesters of prison or penitentiary do more to perfect the professional criminal than years of practical work in freedom.

I cannot here determine to what extent these statements still apply to prisons in various parts of the world. What I have tried to show is this: we cannot dispense with punishments, but we should realize that punishments cannot be just; that a less disproportionate punishment is not always morally preferable; and that the preoccupation with retributive justice is inhumane.


It is widely assumed that the sense of retributive justice – the sense that certain crimes clearly call for certain punishments – is primordial, instinctive, and universal. It still remains to be shown that this is false. The moral sense of different ages and communities differs very widely, and there could hardly be a better illustration of the fact that conduct viewed with utter horror by one society is frequently enjoined by another than the way in which we ourselves react to the punishments imposed by Europeans and Americans in the not distant past.

On reflection, murder is probably the only crime of which large numbers of people still believe that it is somehow self-evident that it calls for a particular penalty: capital punishment. It is assumed that the feeling that murderers deserve death is inscribed in the hearts of men, and that only modern reformers have forgotten this ancient truth. I shall confine myself to this example and show how wrong this assumption is.

In his study of Primitive Law, A. S. Diamond has shown that all early and what he calls “Early Middle Codes” punished homicide with fines, and in the many more or less primitive tribes he studied, pecuniary fines for homicide outnumbered capital punishment by a ratio of better than five to one: 73 percent versus 14 percent. In the remaining 13 percent the punishment was also a fine; the slayer had to turn over to the family of the slain a number of persons – women, children, or slaves. It is only in “Late Middle and Late Codes (including England, 1150 and onwards)” that intentional homicide is taken to require capital punishment.

In his discussion of the old Icelandic saga, Burnt Njal, Diamond quotes the narrator as saying admiringly of one of the heroes: “He was a strong man well skilled in arms, and has slain many men, and made no atonement in money for one of them.” The same kind of admiration is not uncommon to this day; but the point here is that homicide was considered “a purely civil wrong, a matter for the individuals or families affected to avenge or compromise as they think fit. In fact, they always or almost always compromised by the giving and acceptance of an agreed sum of money.”

We need not even go that far afield. In the Iliad, Ajax explains how unreasonable he finds Achilles’ refusal to accept amends for the beautiful slave girl whom Agamemnon has taken away:

Men have accepted a fine from their brother’s
slayer and even accepted it after a son had been slain, and
having paid a great deal, the slayer remains in the country,
while the injured man’s heart and pride are appeased when
he has received the fine.

Our term “punishment,” like the French punition, comes from the Latin poena, which originally designated the fine that the accused paid to the plaintiff; and poena is a loan word from the Greek – the Greek word that Homer uses twice in the passage cited above. The liberal mind was fond of seeing all of human history as a steady progress from primitive cruelty to modem humanity. Seen in this mythical perspective, Hitler’s atrocities looked like a scarcely credible throwback into barbarism. In fact, many scholars have come to the conclusion that neither primitive tribes nor antiquity match the cruelty that gradually developed in the penal codes of Christian Europe. Ancient Rome went the same way, though not quite so far, and was far crueler in the end than in early days. In Mexico none of the earlier civilizations matched the cruelty of the last one, that of the Aztecs. And the five books of Moses have no inkling of the Gospels’ eternal torment or the tortures of the Inquisition.

The last point that still needs to be made about retributive justice can be put into three words: desert is incalculable. Not only is it impossible to measure desert with the sort of precision on which many believers in retributive justice staked their case, but the whole concept of a man’s desert is confused and untenable. This claim is as fatal for distributive justice as it is for retributive justice, and I shall deal with it at length in the next chapter.

« Decidophobia | An Attack on Distributive Justice »