From Analysis, Vol. 54 No. 2, April 1994. (c) Deirdre Golash, 1994. All rights reserved.
At the heart of retributivism is the contention that it is the wrongness of the criminal act that justifies the imposition of punishment on the offender. Yet punishment itself consists in the performance of a parallel act against the offender. Thus showing that the harmful acts that are crimes have a moral value precisely opposite to that of the harmful acts that are punishments is the central task of retributive theory. It is not enough to show that some crimes involve acts unacceptable in any context, such as rape and torture. Punishment necessarily involves acts that are ordinarily unacceptable -- specifically, acts that in another context would themselves be crimes. To highlight this tension, I shall consider the example of a simple theft that is punished by a fine. The retributivist must be able to show not only that theft is wrong while punishment by fine is right, but also that there is an intimate connection between these two judgements: that the fine is right because the theft is wrong.
Some retributivists decline to argue for this point, saying instead that it is intuitively true. (E.g., Mackie .) This position is unsatisfactory not only because some have this intuition while others don't, but because even those who do cannot escape the paradox it presents. One cannot have the retributivist intuition that `the fine is right because the theft is wrong' unless one also has the intuition `taking another's property is (usually) wrong'. This by itself should be sufficient to open the way to examination of the consistency of the two intuitions.
Other retributivists have sought to provide explanations relying on the social effects or the social meaning of crime and punishment. These attempts have encountered difficulties well documented elsewhere (Burgh ; Honderich ; Ten ). Recently, Jean Hampton  has suggested that retributivism is best understood by reference to the effect of punishment on the victim, rather than its effect on the offender or society. I shall argue that this version of retributivism founders on the rock of the essential symmetry between crime and punishment. Specifically, under this theory punishment is inconsistent with the equal value of persons and defeats the retributivist goal of restoring the status quo ante.
Hampton argues that the criminal act makes a false statement about the relative values of offender and victim, and that the purpose of punishment is to reassert the moral truth and to nullify the evidence of the offender's superiority provided by the crime. Returning like for like, however, appears antithetical to these goals. A person who curses another literally speaks an untruth about the person's moral value, perhaps reducing her to the status of an animal or a body part. How can the victim (or anyone) nullify the evidence of her value provided by the curser and show her true value? She may hold her head high and to show her true value through her conduct. Thus the curser is made to look foolish as his lie is exposed. Alternatively, she may show that she does not care about the opinion of the curser by failing to react with anger, grief, or shame. Least effective of all is an attempt by the victim to bring the curser low by responding with an equally foul vilification. Should this effort succeed, it hardly elevates the (original) victim; rather, it seems to reduce her to the same status as the (original) offender. I shall argue that punishment similarly fails to affirm the victim's value.
It is easy to see how the crime can assert the offender's superiority over the victim. If A embezzles $100 from B, A is in effect making the claim that his desire for the money is more important than B's property right to it. As rights are more important than simple desires, this amounts to a claim that A is more valuable than B. But how does the punishment refute A's claim of superiority? Hampton suggests that by depriving A of some of his own property as punishment, we assert that A is not more valuable than B: what A can do to B, we can do to A. But if A's appropriation of B's property is a claim that A is more valuable than B, then isn't our appropriation of A's property simply a claim that we are more valuable than A? This, of course, would do nothing to refute A's initial claim that he is more valuable than B. If we say that we (or the state) are acting as B's agent, so that our action in punishing A is `really' B's, the punishment would be a claim that B is more valuable than A -- a claim as inconsistent with the equal value of persons as the original crime.
This simple model does not reflect the full detail of Hampton's position. Hampton says that the retribution-seeking victim is not looking for a competitive victory over the offender; that would be the strategy of the malicious hater, who is open to moral criticism. The motives of offender and victim are crucial to the kinds of statements they are taken as making, so that even though their overt actions are the same, the statements they represent may be quite different.
The malicious hater, according to Hampton, seeks to harm his victim in order to demonstrate his superiority over her. This, we may assume, is the strategy followed by the offender. The crime demeans the victim in the sense of according her treatment that is too low for her. The offender may also seek to diminish the victim, i.e., to lower her own estimate of her value, or to degrade her, i.e., actually to change her value. The retributive hater, on the other hand, seeks only to accord the offender the treatment he deserves, and thus to give him the treatment that is in accordance with his true value. Thus, Hampton says, the retributive hater does not seek to demean the offender but only to diminish him, that is, to deflate his overblown sense of his own value by giving him the treatment that is appropriate for him. If, however, all persons are of equal value, and the offender's theft represents treatment that is too low for a person of that value, it necessarily follows that depriving the offender of his property similarly represents treatment that is too low for persons of that value. We can escape from this dilemma only by conceding that punishment makes the statement that offenders are of lower value than victims, or by saying that appropriate treatment depends on something other than a person's value.
Hampton's position is that the crime makes a statement about the relative value of offender and victim while the punishment makes a statement about the absolute value of the offender. Thus, A's theft from B says `A is more valuable than B', while the punishment says `A is less valuable than X' where X represents the value that (the crime demonstrates) A places on himself. If A and B are in fact of equal value V, then the punishment must say `A is of value V'. The punishment seeks to diminish A, that is, to reduce his subjective sense of his own value as greater than V, by according him the treatment that is appropriate for his true value. But if A's treatment of B is not appropriate for the value V, then neither is our similar treatment of A.
It may be objected that this simple model ignores the interplay of motives and actions that determines the statement made by an action. The motives of A and B are not symmetrical. A (we may suppose) is motivated simply to satisfy his desires; thus he claims (acts on his belief that) A's desires are more important than B's rights. But B's motive (as punisher, or that of B's agent) is different: he acts in order to assert moral truth; specifically, to assert that A's claim is false. His act, to be effective in achieving this goal, must amount to an assertion that B's right (to his money) is more important than A's desire (for B's money). Now the puzzle: why can B not best assert this by simply reclaiming the $100 from A? (This of course would be restitution, not punishment.)
It may be argued that such reclaiming makes no distinction between desires and rights: A's appropriation of the money is an assertion that his desire is more important than B's right; B's appropriation must be clearly understood as an assertion that B's right is more important than A's desire, rather than simply as an assertion of a contradictory desire. If we were talking about a purely private conflict between equals, this objection would have some force. But where it is the state that acts as B's agent, the statement made by punishment is a statement of (at least legal) right, not merely of desire. In the civil law context, state action is clearly seen as affirming rights. If I breach my contract to sell you my house, in effect claiming that the contract is not binding (or indeed that my desire to keep the house is more important than your right to it), you may win a civil judgement against me to the effect that the contract is binding. The law's forcing me to sell is precisely a refutation of my claim that my desires are more important than your rights. Because restitution is not punishment, and because restitution (where it is possible) does restore and vindicate the victim, it is clear that in punishing we are making some statement in addition to the simple denial of A's false claim that his desires are more important than B's rights. Thus, the interesting question is that of the content of this additional statement.
Because the act of punishment closely parallels the crime itself, the statement that it makes must also parallel the statement made by the crime, except insofar as the two acts are different. The crime of theft ranked A's desire for a specific item above B's right to that item, thus making a false statement about their relative values. Restitution correctly ranks B's right to his property above A's desire for it. Punishment, then, must make some similar statement, and if the punishment is a fine, the second term of the statement will be A's prima facie right to his own property. The question then becomes one of what right or desire of whom is being ranked above A's (prima facie) right to his (own, not B's) property. The answer to this question will determine the truth or falsity of the statement made by the punishment.
Why can't we just say that it is the punisher's right or desire to do justice or to assert moral truth that is ranked above A's right to his property? This would be narrowly circular; the issue is whether punishment is just and whether it does indeed represent moral truth. If we had a satisfactory separate justification of retribution, the idea of affirming the victim's value would be redundant. We must therefore find some other way of specifying the right or desire that is fulfilled by punishment in order to make sense of the claim that it makes a true moral statement.
Whatever right, desire or interest of the victim or punisher is ranked above A's prima facie right to his property, either it is more important than the prima facie right to one's property, or it is less important. If it is less important, the necessary implication of this is, again, that B (or the punisher) is more valuable than A. If it is something more important than one's prima facie right to one's property, then it must be a more important right of B that is at stake. (Desires or interests of others cannot outweigh rights.) It might be B's right to see A suffer, as Jeffrie Murphy suggests (, p. 89), or B's right to have A's false claim of superiority refuted. This right, however else it is specified, must at least include B's right to see A (e.g., suffer through being) deprived of his property. But of course B has no such right apart from his right (if he has one) to have A punished for wronging him.
Could we say that A has lost his right to his property through his actions? The punishment would then coherently make the true statement that A has no right (against the state) to this property, while the crime would make the false statement that B has no right (against A) to his property. But this, on analysis, is the same as the claim that something is more important than A's prima facie right to his property (which would cause him to lose his all-things-considered right to it). A cannot have lost his prima facie right to his property, though he may have lost his all- things-considered right to it. If the latter, we must still explain what it is that overrides A's prima facie right, and this must still be some right or desire of B.
Can the theory be saved at the expense of rejecting the equal value of persons? One might postulate that the crime does indeed reduce the wrongdoer's moral value. Thus, the victim's desire to see the wrongdoer suffer (proportionally to the crime) is correctly ranked above the wrongdoer's right not to suffer. The offender is punished in virtue of what he has done to B, and the state acts as B's agent in punishing him. B's (assumed) desire to see A suffer is correctly ranked above A's (prima facie) right not to suffer, because (in virtue of his act) A is lower in value than B. Thus, the punishment is justified, while the crime is not, because the crime makes a false statement that one person is lower in value than the other, and the punishment makes a true statement to the same effect.
The statements made by crime, restitution, and punishment would be as follows:
The authoritative statement that the offender is of low value, however, appears to conflict with another traditional goal of retributivism: restoration of the status quo ante. If punishment is to `annul' the crime, all persons should return to the status they had before it was committed. The assertion of the lower value of the offender in virtue of his act appears instead to ossify the status quo post. The goal of annulling the crime, it seems, is also best met by restitution.
It may seem that these problems can be solved by separating the offender from his act. If punishment makes a statement, not about the value of the offender, but instead about the value or validity of his act (the truth of his statement), then it succeeds in restoring the status quo ante and asserting the moral truth without raising the problem of assigning a lower value to wrongdoers.
What kind of statement would satisfy these conditions? Recall that punishment necessarily makes a statement parallel to the statement made by the crime. In order to say that the punishment does not make a statement about the value of the offender (as a whole person), we must also say that the crime does not make a statement about the victim (as a whole person). If it does not do this, then the entire motivation for punishment as affirming the value of the victim disappears.
What about crimes such as murder, for which restitution is not possible, or those, such as rape, where restitution is less readily seen as restoring the status quo ante? Consider murder. A kills B, let us suppose, to satisfy a passing whim, thus making the false statement that his whims are more important than B's most fundamental rights. How can we reassert that B's rights are more important than A's whims? Suppose that we respond by killing A. We have clearly said that something is more important than A's fundamental right to life. But again, this amounts to a statement that justice is more important than A's right to life only if killing A is in fact just. Have we said, perhaps, that the satisfaction of our anger is more important than A's right to life? If so, either we have made a false statement, or A is not a very valuable person. The fact that restitution is impossible may simply mean that nothing we can do to the offender will assert the relevant moral truth in this situation.
Department of Justice, Law and Society
Washington, DC 20016-8043, USA
 Richard Burgh, `Do the Guilty Deserve Punishment?' Journal of Philosophy 79 (1982) 193-210.
 Jean Hampton, `The Retributive Idea', in Jeffrie Murphy and Jean Hampton, Forgiveness and Mercy (Cambridge; New York: Cambridge University Press, 1988), ch. 4.
 Ted Honderich, Punishment: The Supposed Justifications (Harmondsworth; Baltimore; Victoria: Penguin Books, 1971).
 J.L. Mackie, `Morality and the Retributive Emotions', Criminal Justice Ethics (1982) 3- 9.
 Jeffrie Murphy, `Marxism and Retribution', Philosophy and Public Affairs 2 (Spring 1973) 217-243.
 Jeffrie Murphy, `Hatred: A Qualified Defense', in Jeffrie Murphy and Jean Hampton, Forgiveness and Mercy (Cambridge; New York: Cambridge University Press, 1988), ch. 3.
 C.L. Ten, Crime, Guilt, and Punishment (Oxford: Oxford University Press, 1987), chs. 3-4.