Você está na página 1de 23

Censure Theory and Intuitions about Punishment Author(s): Thaddeus Metz Source: Law and Philosophy, Vol.

19, No. 4 (Jul., 2000), pp. 491-512 Published by: Springer Stable URL: http://www.jstor.org/stable/3505080 . Accessed: 16/04/2014 15:55
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Springer is collaborating with JSTOR to digitize, preserve and extend access to Law and Philosophy.

http://www.jstor.org

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

THADDEUS METZ

CENSURE THEORY AND INTUITIONS ABOUT PUNISHMENT 2000) (Accepted24 February


ABSTRACT. Manyphilosophersandlaypeoplehave the following two intuitions the state has a pro tantomoralreasonto punish all those aboutlegal punishment: of a guilty breaking just law and to do so in proportionto their guilt. Accepting that there can be overridingconsiderationsnot to punish all the guilty in proportion to theirguilt, many philosophersstill considerit a strikeagainstany theoryif it does not imply that there is always a supportivemoral reason to do so. In this thatcensuretheoryaccountsfor these intuitionsmuch better paper,I demonstrate than any other theory,includingforms of retributivism such as desert theory and fairnesstheory,and explain why censuretheoryis able to do so. KEY WORDS:punishment,retributivism, expressivism,censure,proportionality

I. INTRODUCTION

Many people have the following two intuitionsabout legal punishment: the state has a pro tanto moral reason to punish all the guilty and to do so in proportionto their guilt. The pro tanto qualification is important;any reasonabletheory of legal punishmentmust not only acknowledgethat actuallypunishingall the guilty would have intolerablecosts, but also make room for mercy and forgiveness. Accepting that there are overriding considerationsnot to punish all the guilty in proportionto their guilt, many philosophers still considerit a strikeagainst any theory if it does not imply that there is always a supportivemoralreasonto do so. My thesis is that censure theory, suitably understood,accounts for these intuitions better than any other theory of punishment, retributivist or otherwise.Censuretheorymaintainsthatthe political has to community a pro tantoduty to censureinjusticein proportion the injustice done and that this must be done throughproportional punishment.It is a relatively undeveloped theory of punishment,
i

Law and Philosophy 19: 491-512, 2000. ? 2000 KluwerAcademicPublishers. Printed in the Netherlands.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

492

METZ THADDEUS

articulated andadoptedonly in the last 15 havingbeen substantively to censuretheoryhave failed to note years.' Even those sympathetic its striking advantageof best cohering with firm intuitions about legal punishment. I acknowledge that some utilitarians(and other non-retributive punishmenttheorists) may not find these intuitionsas powerful as retributivists. It was Kant,afterall, who groundedhis theory on the judgmentthat a murderershould be punishedeven if the results of not punishing would be better.2Still, utilitarianstypically do take such judgments seriously, and they should. They may not dismiss moral intuitions as conservative, cultural biases while simultaneously invokingintuitionsabouta maximizingpracticalrationality; consideredjudgments about morality are on par with those about for utilitariansto work so rationality.It has been quite appropriate hard over the last 30 years respondingto the objection that their justifies punishmentof the innocent. The theory counterintuitively following are, by the same token, relevantcriticisms:utilitarianism incorrectly provides no pro tanto moral reason to punish all the guilty or to do so in proportionto their guilt. If some utilitarians continue to find appealing to these intuitions to beg the question, then they may read this essay as workingprimarilyto develop the most promisingform of retributivism.
1 Centraldefenses of censure theory include:Joel Feinberg,"TheExpressive Function of Punishment," in repr. Doing and Deserving (Princeton:Princeton UniversityPress, 1970), pp. 95-118; AnthonyDuff, Trialand Punishments(New York: CambridgeUniversity Press, 1986), esp. ch. 9; MargaretFalls, "Retribution, Reciprocity,and Respect for Persons,"Law and Philosophy 6 (1987): as Language," 25-51; Igor Primoratz,"Punishment Philosophy 64 (1989): 187Jean "The Retributive in Jean 205; Idea," Hampton, HamptonandJeffrieMurphy, and York: Forgiveness Mercy (New CambridgeUniversity Press, 1988), ch. 4, "AnExpressiveTheory of Retribution," in Wesley Cragg, ed., Retributivism and F. Steiner Verlag, 1992), pp. 1-25, and "Correcting Harms Its Critics (Stuttgart: VersusRightingWrongs:The Goal of Retribution," UCLALawReview39 (1992): in the Philosophy of PunishAndrew von 1659-1702; Hirsch, "Proportionality ment: From 'Why Punish?' to 'How Much?'" CriminalLaw Forum 1 (1990): 259-290, and Censureand Sanctions (Oxford:Oxford UniversityPress, 1993); JohnKleinig, "Punishment andMoralSeriousness," Israel LawReview25 (1991): 401-421. 2 See ImmanuelKant, The Metaphysicsof Morals Mary Gregor,trans.(New York:Cambridge UniversityPress, 1991), pp. 142-143.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

493

In the following, I will first describe censure theory and explain what I mean when I say that it best accounts for the intuitionsthat there is always some moralreason for the state to punish the guilty to theirguilt (II). Then I will considerthe and to do so in proportion of theories, which justify legal punishability "forward-looking" ment in terms of results to be achieved, to account for these two intuitions (III). I will explore several versions of forward-looking but which promise theorywhich are less familiarthanutilitarianism to accommodatethese judgments better than it does. However, I will end up contending that no version of forward-lookingtheory accommodatesthese judgmentsvery well. Next, I will demonstrate theoriessuch as fairness thatrivalretributive or "backward-looking" theory and deserttheoryalso fail to do a good job of accountingfor these intuitions (IV). In the following section, I will demonstrate that censure theory bettersits competitorson this score, and I will provide the deep explanation of what enables it to do so (V). I will conclude by noting the major problems with censure theory which, in light of censuretheory'spromise,I believe warrant being addressedfully in anothercontext (VI).

II. AN OUTLINE OF CENSURE THEORY

Censuretheory,in the form consideredhere, holds that the government must punish in orderto fulfill its duty to denounce injustice. More specifically, the political community has a pro tanto moral to the degree of injusticebecause obligationto punishin proportion can it its discharge obligation to denounce injustice only thereby in proportionto the degree of injustice. We of course may wonder whether it is true that the state must censure injustice or that censuring injustice requires punishment.However, in this essay I grantthese claims to the censure theorist,in orderto see how well her theory fares on other grounds.In other words, I concede to the censuretheoristher accountof why legal punishmentis justified, in orderto test her theory'sabilityto accommodateintuitionsaboutfor whomand how muchlegal punishmentis justified. "Injustice"includes at least the breaking of any just law, the content of which may be left open here. Censuretheory is compatible with different notions of the content of justice (e.g., liberal,

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

494

METZ THADDEUS

or egalitarianconceptions), and also with conservative,libertarian, differentspatial and temporalscopes of justice (domestic, internaor intergenerational). tional, intragenerational, Similarly,"political community" or "government"may pick out any legal system, broadlyconstrued,e.g., the leaders of a small town, the legislature of a nation-state,or the generalassembly of the United Nations. What does it mean for a political community to "censure"or an agent for having been unjust?I believe that both "denounce"3 and detractorsof censure theory often characterizeits supporters centralidea more narrowlythanit needs to be. Specifically,censure is frequently (mis)construed as essentially a matter of communicating something about the censured to the censured. Gricean distinctions are often invoked to characterizecensure's supposed informativenature.Hence, theoristsoften deem censure inherently to involve conditions in which (a) one agent transmitsa symbol to anotheragent with the intentionthat she both understand a certain negative proposition about herself and recognize that the agent the symbol intends her to understand that proposition transmitting by means of the transmission,and (b) the other agent in fact understands that proposition by means of the transmittedsymbol and that recognizes thatthe transmitting agent intendsher to understand means of the On transmission. this model, agent X propositionby denounces agent Y (roughly)insofaras agent X intendsto educate view of Y and succeeds. agent Y aboutX's reproachful While censurecan take this form, it may plausiblybe understood to include a broaderarrayof conditions, or at least the way I will use the term"censure" is less restrictive.Firstoff, (and "denounce") I maintainthat censure of a person need not involve communication with the person censured.It is possible to denounce someone without her understanding any propositionor even without transmitting a symbol to her with the intention of her understanding a proposition.For example, one can denounce the dead. Censure has probably been identified as a way to convey a message to
Roger Wertheimerhas impressed on me the need to distinguish between which by definitioninvolves imposing a harm,and censure,which condemnation, does not. Therefore,I do not take the word "condemnation" to indicate the same idea as "censure." See his "Understanding Criminal JusticeEthics 2 Retribution," 19-38. (1983): 3

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

495

the censured because censure in the criminaljustice system often involves such a relation. Theorists imagine a judge sentencing someone to a certain punishment in the course of telling him how seriously he has flouted the community's norms. I do not, however, suppose that the concept of censuring a person requires communicatingwith the censured. At this point, one could still think that a necessary condition of censure is that someone other than the censuringagent, but not a propositionby means necessarilythe personcensured,understands of a symbol that has been sent to her with that intention.In order to denounce the dead, perhaps one must communicatesomething about her to someone alive. One might suppose, then, that censure is (roughly) a matterof an agent X intendingto educate someone else (Y or Z) aboutX's reproachful view of Y and succeeding. But I do not believe thatcensure shouldbe viewed as necessarily informative.Suppose a victim's family gatherstogetheronce a year to spit on the graveof the criminalwho killed a relative.This family can be viewed as denouncingthe offenderbut not therebycommunicating with anyone. The family membersare not communicating with the offender, at least if there is no afterlife and they do not believe in one. Nor are they communicatingwith one another.Each member of the family already knows what the others think about the criminal;it would be pointless to spit on the criminal's grave with the intention of sending a message to other family members aboutone's view of him. Instead,we can understand this instanceof censureas a matterof expressingdisapproval butnot communicating disapproval. Another reason that I do not consider censure to be inherently communicativecomes from reflection on the symbolic nature of communication. is a matterof conveyinga message Communication by means of conventionalrepresentations. Censuringby means of the silent treatment,withdrawalof privileges, and imposition of harms is not obviously to use symbols that denote disapproval. Instead, these behaviors might be non-symbolic expressions of Consideran analogy with expressionsof disrespect.To disapproval. kill a strangerfor fun and to lacerate oneself out of self-hatredare two ways to expressdisrespecttowarda personthatare not communicative. In murderinganotherperson, one expresses the judgment

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

496

THADDEUS METZ

that the otheris worthless than oneself or less than one's pleasure, but withoutsendingconventionalrepresentations to anyonewith the intention of conveying a proposition. By injuring oneself behind closed doors, one likewise shows disrespect for oneself without using symbols to transmitany message. Similarly,I propose that making someone pay a fine or serve time in jail could be ways of towardsomeone withoutusing conventional expressingdisapproval representations. In sum, against the narrowunderstandingof censure in terms of communication with the censured party, I suggest that it is possible to censure someone without communicatingwith her or even withoutcommunicating at all. I thereforeconstruecensure (or as follows: an denunciation) expressionof disapprovalof someone for a wrong perceived to have been done by her. Now, I am not sure whethera normalspeaker'ssense of "censure" allows for this possibility, or whether I am stipulatingto some degree. It makes no difference. The point is that, once one allows that the state has a duty to "censure"injustice and that it can do so in a noncommunicativeform (or once one agrees thatthe statehas a duty to of injusticewith punishment),one has a theory express disapproval which uniquely matches centralintuitionsabout legal punishment. If the readeris firmin thinkingthat"censure" by definitioninvolves then she may insteadcall the theoryI am sketching communication, here "expressivism." Some theoristshold that the concept, as opposed to the justificaincludesthe idea of censure.4Thatis, some tion, of legal punishment maintainthat"punishment" by definitioninvolves expressingdisapa for perceived wrong having been done. I do not commit proval myself to this analysis of legal punishment.For my purposes,it will the state suffice to adopt the following notion of legal punishment: punishes insofar as its officials intentionallyimpose hardtreatment upon an individualconsequent to an unjust act apparentlyhaving been done. As I understandcensure theory,it maintainsthat legal punishment,so construed,is justified because needed for the state of those who have in fact to dischargea duty to express disapproval been unjust.
4 See especially Feinberg,op. cit.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

497

Censuretheory is a retributive theory in the broadsense that the moral reason for punishing does not involve the promotionof any results. Instead of justifying punishmentby "looking forward"to the productionof virtue or happiness (or the reductionof vice and misery), censuretheory"looksbackward," justifying punishmentas itself a morallyright response to past behavior.One could also call this theory "intrinsic to highlightthe notion thatthe expressivism,"5 of is considered ethically warrantedapart expression disapproval from any good consequencesit might produce. Although censure theory is not result oriented, most of its defenders do not consider the expression of disapprovalas such to be an end-in-itself. That is, censure theory proponentstypically claim that the duty to censure follows from other duties and that these latter duties are the ultimate ground for punishing. Specifiduties which have been thoughtto necessitate cally, the fundamental censure are the following: standing up for justice, affirming the value of victims,and treatingoffendersas responsible. First, some censure theorists hold that the political community can discharge an obligation to disavow the violation of just laws only by denouncing such behavior. The idea is that the political community must distance itself from injustice and place itself on the side of justice. If the governmentdid not censureinjustice,then it would betray the value of justice. For example, after the Holocaust, it would have been wrong for nation-statesnot to express "Never again."By censuring Nazi leaders during the Nuremberg trials,manycountriesfulfilledtheirdutyto standup for international humanrights standards. A second duty thought to require censure is affirmationof the worth of the victim. The political community would degrade the victim of crime, if it did not denounce the wrong done to her. Not expressing disapprovalof the unjust action would fail to treat the victim as worth taking seriously. To returnto the Holocaust example, had the world community not censured the behavior of Nazi elites at Nuremberg,it would have been a slap in the face of Nazi victims. A third obligation of the state which has been deemed to require censure is the duty to treat offenders as responsible for
5 Primoratz, op. cit., uses this phrase.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

498

METZ THADDEUS

their behavior.Most strongly, some contend that not to denounce offenderswould be to respondto their actions as thoughthey were mere events and so to treat the offenders as animals ratherthan agents. Withoutcensure, Nazi headmenwould not have been fully treated as persons capable of moral deliberationand action who chose not to exercise these capacities. To sum up: if one thinks that the state has obligations to stand up for justice, to affirmthe worth of victims, and to treatoffenders as responsible, then one will be sympatheticto the claim that the state must censure people for breakingjust laws. Of course, one might deny that the state has such obligations. One could also deny that these obligations entail an obligation to censure (or an obligation to censure by means of punishment). However, I am not concerned to provide a justificationof the claim that the state must denounceinjustice(with punishment); thatis quite beyondthe scope of this essay. I am merely pointing out what censure theory it fromrivaltheoriesandultimately involves,in orderto differentiate to show that it accountsfor commonsensicaljudgmentsaboutlegal punishmentbetterthanany of them. When I say censuretheory "bestaccounts"for intuitions,I mean two things. First, betterthan any other theory of legal punishment, censure theory shows that the intuitions are true. Censure theory easily entails that the state has a pro tanto obligation to punish all those who have unjustifiablybroken a just law and to do so in to the seriousnessof the violation. Second,betterthanits proportion censure rivals, theoryexplainswhythese intuitionsaretrue.Censure theory provides the most attractiverationalefor the state's having some moral reason to punish all the guilty in proportionto their guilt. I will make these claims plausiblein the next three sections.

III. FORWARD-LOOKING

THEORIES

Forward-lookingtheories, which hold legal punishment to be permissibleonly if and because certainlong-termresults are forthcoming, are well knownfor being unableto show thatthereis some moral reason to punish all the guilty; for there can be situationsin which punishingthe guilty produces worse consequencesthan not punishing.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

499

First off, there are theories such as utilitarianismand moral educationtheory,which make the justificationof legal punishment contingent upon the (expected) realization of intrinsic value.6 In cases wherepunishingfails to produceat least as much happinessor virtue as not punishing,these theories fail to show that there is pro tanto moral reason for the state to punish a guilty person. It seems quite possible that the costs of punishmentcould reduce its overall benefit to the point where some other response to an offender is optimific. A natural first response is to appeal to rule-utilitarianism, in which acts of punishmentare evaluatedin terms of whether they conformto rules which maximallypromotewelfare when generally followed. However,this theory also faces the problem of counterintuitively implying that there is sometimes no moral reason to punish the guilty. This is clearest in the case of a racist society. In such a context, the following rule could well maximize happiness: punishthe guilty except when the guilty have harmeda personwho belongs to a small anddespisedminority.Althoughfollowing such a rule would greatlyupset the minority,the happinessof the majority could substantiallyoutweigh it. As a second response,the forward-looking theoristmight search for a way to discountthe happinessof majorities.For example, she might specify virtueas the intrinsicvalue to be soughtwhen considering what would happenif everyoneperformedan act. So, consider a rule version of moral educationtheory,accordingto which legal punishmentof an individual is justified only if and because it is allowed by a rule which, if generally followed, would maximize moralreform.So far as I know,no one has advocatedthis theory,but it is worthconsideringit here. Since this theorydoes not take happiness into account, it appearsto avoid the implicationthat a guilty memberof a majorityshould not be punishedin a racist society. However,one can well imagine a situationin which the racismis so entrenchedthat little or no moral reformresults from punishing
For a recentdefense of the utilitarian theoryof punishment,see J. J. C. Smart, "Utilitarianismand Punishment," Israel Law Review 25 (1991): 360-375. For good statementsof moral educationtheory, see HerbertMorris,"A Paternalistic AmericanPhilosophical Quarterly18 (1981): 263-271; Theory of Punishment," and Jean Hampton,"The Moral EducationTheory of Punishment," Philosophy and Public Affairs 13 (1984): 208-238.
6

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

500

THADDEUS METZ

membersof the majoritygroup. Suppose that, instead of reflecting on theirbehavior,majoritymemberstend to backlashwheneverone of theirgroupis punishedfor harminga minoritymember.Imagine thatmajoritymemberson balancerespondmuch betterto a judge's verbal instruction.In this case, the rule variantof moral education theorywould provideno moralreasonto punishthe guilty. Let us consider a different strategy on behalf of the forwardpromises looking theorist.Specifically, "scalarconsequentialism"7 better to accommodatethe intuitionthat there is always pro tanto reason for the state to punish a guilty individual. According to this theory,the more good consequences that an act has, the more moral reason there is to performit. The difference between scalar consequentialismand the usual variety is that moral reasons to act are not tied exclusively to a certain (maximizing or satisficing) sum of good consequences. The scalar theory says that any time an act produces some good (or reduces some bad), there is a pro tanto moral reason to do it. A scalar consequentialistaccount of legal punishmentwould be this: the more good consequences that legal punishmenthas, the more moral reason there is to impose it. This theory does not make the rightnessof punishmentcontingent on a particularaggregate of intrinsic value - indeed, it does not even invoke the concept of rightness.So long as punishmentof the guilty will produce some degree of good results such as virtue or says thatthereis pro tantomoral happiness,scalarconsequentialism reasonto punish. Scalar consequentialism,too, turns out to have counterintuitive implications about punishmentof the guilty. One can imagine a case in which punishmentof the guilty producesabsolutelyno good results (and does not reduce any bad). Suppose an innocent,friendless drifterwandersinto a communityin which everyonehates him for being different. A member of this community kills him, and the communitykeeps this fact to itself. Everyoneis happythat this individualwas killed, and so no one wantsthe killer to be punished. Here, punishingthe killer will not make anyone any happier(or, let should be us suppose, any more virtuous).But surely the murderer
7 Frances Howard-Snyder has done the most to develop this form of consequentialism. See, for example, "The Heart of Consequentialism," Philosophical Studies 76 (1994): 107-129.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

501

cannotshow thatthere punished.If so, then scalarconsequentialism is pro tantomoralreasonto punishall the guilty.Even thoughscalar consequentialismdoes not tie the permissibilityof punishmentto a specific amount of good results, there still must be some good results for punishmentto be morallyrational.This counterexample still plagues the scalartheory; shows thatinvidiousmajoritarianism sometimes punishment of the guilty has no good results, and in such a case the scalartheoryimplies thatpunishmentis not morally rational. Now, I must admitthat there is one idiosyncraticform of scalar consequentialismwhich does entail that there is always pro tanto moral reason to punish the guilty, at least if the censure theory entails this. Such a scalar consequentialismwould hold censure itself to be an intrinsic value. Specifically, the theory would say this: the more that a state censures guilty parties in proportionto their guilt, the more moral reason there is for a state to punish.8If censureof the guiltyjust censuretheoryis correctthatproportionate then is a matterof proportionate deemingproportionate punishment, censure to be an intrinsicgood that should be promotedwill entail thatthereis always pro tantomoralreasonfor the stateto punishthe guilty. Thereare two reasonswhy even this artificialform of consequentialism does not accountwell for the intuitionsat hand. First, scalar theories in general do not provide sufficient normativeguidance. A full-fledged theory of legal punishment should say not only why punishment is permissible but also when. However, scalar consequentialismdoes not even indicate a necessary condition for the permissibilityof legal punishment.I am therefore inclined to think that it does not even count as a theory of legal punishment which could rival censuretheory. Even if scalar consequentialismin general counts as a genuine competitorto censure theory, a particularscalar consequentialism that deems censure to be an intrinsicgood does worse than censure theory in explaining why punishmentof all the guilty is justified. Censure theory says that censure is right and that punishmentis
retributivism" discussed in Michael Moore, Comparethe "consequentialist Placing Blame:A GeneralTheoryof CriminalLaw (New York:OxfordUniversity Press, 1997).
8

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

502

METZ THADDEUS

therefore permissible, whereas the scalar consequentialismunder consideration says thatcensureis intrinsicallygood andso shouldbe Censure promoted. theoryis more plausiblein this respect since the scalarview oddly implies the following:it would be a good thingfor people to breakthe law, since states would then be in a position to censurethem andtherebypromoteintrinsicvalue. This scalartheory instructspeople to commit crime, since only by that means can the state be in a position to realize the intrinsicworth of censuringthe guilty. In short, even if a scalar consequentialisminstructsus to promote censure and thereby shows that there is moral reason to punishall the guilty, it providesa deficientrationalefor doing so. There are forward-lookingtheories which differ from those considered so far in that they do not justify legal punishmentin terms of results which have intrinsic value. The relevant results are insteadthose involved with the exercise of rights.For example, restitutiontheory says that punishmentis allowed only when and because it will fulfill a victim's right to be compensatedby, say, deterring crime and thereby reducing fear of attack.9 And selfdefense theory says that the innocent have the right to carry out threatsof punishmentto the guilty only when and because it will detercrime and therebyprovideprotection.10 These rights-based versions of forward-looking theory have the same problem as the intrinsic value versions. There will be circumstancesin which punishmentby the state simply will not compensate and will not protect, and in such cases these theories cannot entail thatthereis some moralreason for the state to punish someone who has unjustifiably violated a just law. The factors making it difficult for forward-lookingtheories to show thatevery guilty personis a propercandidatefor legal punishment also make it hard for them to demonstratethat the guilty shouldreceive the properdegree of punishment.In particular, these theories have difficultyshowing that all those who have committed serious crimes pro tanto warrantreceiving a severe punishment.
9 The best defense of restitutiontheory is Margaret Holmgren's"Punishment as Restitution:The Rights of the Community," CriminalJustice Ethics 2 (1983): 36-49. 10 See, e.g., Daniel Farrell,"The Justificationof DeterrentViolence,"Ethics 100 (1990): 301-317; and Phillip Montague, Punishmentas Societal Defense (Lanham,MD: Rowmanand Littlefield, 1995).

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

503

UtilitariThis should be clear, given the previous argumentation. anism cannotprescribea substantial penaltyfor someone guilty of a seriouscrime, if betterresultsare forthcomingwith a lesser penalty. Similarremarksapply to moraleducationtheory,restitutiontheory, and self-protectiontheory.
IV. BACKWARD-LOOKING THEORIES

The requirementthat legal punishment generate certain results makes it hard for forward-lookingtheories to account for the idea that the state has some moral reason to punish all the guilty in proportionto their guilt. The most promising response to make on behalf of the forward-lookingtheories is to advocate "mixed" elements. theories which combine forward-and backward-looking the moral reasons not conceive of does A backward-looking theory for punishingin terms of any conditionthat will obtainin the longrun. Rather,a backward-looking theoryholds that state punishment is justified because it is a right response to the fact of guilt. Fairness theory and desert theory are the most influentialversions of theory.One might think that either one of these backward-looking a or theories, forward-lookingtheory combined with one of them, could accountfor the relevantintuitions. However,neitherfairnesstheory,deserttheory,nor any forwardlooking theory combined with them, can well account for the judgments that the state has some moral reason to punish all the guilty and to do so in proportionto their guilt. After demonstrating the inability of fairness theory and desert theory to accommodate these intuitions,I will bring out the reason for their failure in this respect. I will note that the problem is not with backward-looking theory as such, for censuretheoryis, I will arguein the next section, the backward-lookingtheory which can meet the challenge posed here. Consider fairness theory, which holds that legal punishmentis justified only when and because it will restorean equitablebalance of burdensand benefitsamong citizens.11Fairnesstheoryconceives
1 Prominent defenders include George Sher, Desert (Princeton: Princeton University Press, 1987), ch. 5; and Michael Davis, To Make the PunishmentFit the Crime(Boulder:WestviewPress, 1992).

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

504

THADDEUSMETZ

of the maintenance of governmentas a cooperativescheme in which certainbenefitsgo to society as a whole if andonly if some subsetof the total populationundergoescertainburdens.The primarybenefit of governmentis an orderly society, but one could also include benefits such as health care, education,and so forth. Obedience to burdenthatcitizens must bear in orderto keep law is one important a governmentfunctioning;imagine the chaos that would ensue if the law. Now, if a personreceives the everyoneroutinelydisregarded benefitsof government withoutundertaking the burdenof obedience then to these she takes a "free ride"at benefits, necessary generate the expense of law-abidingcitizens. Punishment restrictsthe liberty of the guilty, so that they no longer get the benefits of government without the cost needed to produce them. The greaterthe liberty taken in breakingthe law, the more punishmentthat is needed to correctthe unfairness. It mightbe worthpausingto comparefairnesstheoryandcensure theory. While neither of these theories makes the legitimacy of statepunishmentcontingentupon resultsobtaining,they differwith regardto legal punishment'sfunction.Accordingto fairnesstheory, the point of state punishmentis to correct exploitationon the part of the offender. One who violates the law has obtained a benefit at the expense of law-abidingcitizens, and state punishmenttakes the benefit away, so that removal of the unfairnessis concomitant with the imposition of punishment.In contrast,for censure theory, the fundamentalends which are concurrentlyrealized with legal punishmentare expressionof supportfor justice and for the victim as well as treatment of offendersas responsiblefor theirbehavior. Censure theory can accept fairness theory's contention that breaking the law often involves a form of exploitation;hence, it may be that one way of expressingsupportfor justice is to remove an advantage which offenders have unfairly obtained. However, censuretheoryneed not view the ends of punishmentsolely in terms of the rectificationof exploitation.For instance,whatmakesmurder warranta stringentpenalty,accordingto fairnesstheory,is just that the offender has taken a great liberty at the expense of those who have not murdered,a liberty that the state must remove only with a severe penalty. But censure theory can maintain that stringent

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

505

punishmentis requiredin orderto affirmthe worthof the murdered party, or to treatthe offenderas a morallyresponsibleagent. These differences between fairness theory and censure theory entailthe furtherdifferentialabilityof themto accountfor the notion thatall the unjustare propercandidatesfor legal punishment.There are at least two situations in which punishing the guilty will not of benefitsand burdens;in these situations restorea fair distribution received has not the guilty party any unfairadvantagefrombreaking the law. First,imaginethata privatecitizen fromanothercountry(or, more fantastically,from anotherplanet)intentionallydrops a bomb on New YorkCity from an aircraft.Since this alien has not received the benefits of Americangovernmentand, hence, has not exploited fellow citizens in breakingthe law, fairnesstheoryfails to justify her punishment.But surely punishmentis justified for such a breachof the law againstmurder. Here is a second scenario in which it is possible to break a just law without taking unfair advantageof law-abiding citizens. Supposea personis convictedof crimeX when he is in fact innocent of having committed any crime. Imagine that he receives Y years of jail, which is whatever amount of punishmentfairness theory prescribesfor crime X. Finally, suppose that, after serving Y years, the person then commits crime X. Fairnesstheory cannotprescribe statepunishmentof this personfor actuallyhavingcommittedcrime X, since the person alreadyunfairlylost Y liberty which fits crime X. Of course, sympathyand compensationare due to the wrongly convicted person. However, we surely do not accept that the state has no moralreasonto punishhim for committingthe crime, which fairnesstheoryentails. Thejustificationof legal punishmentsolely in termsof rectifying unfairadvantagemeans thatfairnesstheorymust forbidpunishment of someone who breaksa just law when he receives no benefitat the expense of others or when he has unfairly received great burdens in the past. I conclude that fairness theory fails to show that the state has some moral reason to punish all the guilty. Let us address the othermajorversion of backward-looking theory,namely,desert theory. because For an agent to deserve somethingis for her to warrant, of some personal feature that she manifested in the past, some-

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

506

THADDEUS METZ

thing of the same kind and in the same proportionas thatfeature.12 The desert theory of punishment, then, maintains that the state may punish a person only if and because he warrantshard treatment due to, and in proportionto, his past unjust behavior. So construed,desert theory and censure theory differ in that censure on behalf theoryessentiallyprescribesan expressionof disapproval of the political community,whereas it would appearthatthe state's she deserves is not necessarilya giving someone the hardtreatment matterof expressinganything. One might wonder whether the two theories collapse if one conceives of punishment as inherentlyconstituting a formof censure (see section II); then it would the case that one deserves censorious punishment. However, there would still be an importantdifference between deserttheoryand censuretheory,when "punishment" is defined as a form of censure. According to censure theory, the fundamentalmoral reasons for expressing disapprovalthrough arethatdoing so is a matterof the politicalcommunity's punishment fulfilling requirementsto stand up for justice, affirmthe value of victims, and treat offenders as responsible. According to desert theory, in contrast, the basic rationale for punishing (which, by hypothesis,is a form of censure)is thatthe state is in the best position to give people the negativeresponseproportionate to what they have earned. Desert theoryis amenableof differentinterpretations, depending on the way one understands"negativeresponse"and "proportionality."Considerdeserttheoryl: legal punishmentis justified only if andbecausethe guilty deserveto sufferin proportion to the injustice done. This straightforward version of desert theory cannot entail that there is some moral reason for the state to punish every person who has broken a just law. For example, desert theoryl will not prescribe legal punishmentwhen a guilty person has previously suffered a great harm that was undeserved;legal punishmentin
12 Withthis statementI summarizeDon Scheid's

thoroughdiscussion of desert in his "Constructing a Theory of Punishment,Desert, and the Distributionof The CanadianJournal of Law and Jurisprudence10 (1997): esp. Punishments," 457-460.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

507

such a case would impose more suffering than is deserved.13For instance, suppose that a serial rapist gets hit by a bus, injuredbut not killed. It is quite naturalfor desert theorists to think, "He got what he deserved." Let us modify desert theory so that it corrects this defect. We need a version that will screen out sufferingwhich the guilty have experienced prior to state conviction. So, perhaps the guilty do not deserve just any kind of suffering,but ratherlegal punishment in particular.Maybe those who have broken a just law deserve suffering to be imposed by a governmentwhich intends to make them sufferbecausethey were unjust.Consider,then,deserttheory2: legal punishmentis justified only if and because the guilty deserve to theirinjustice.14 legal punishmentproportionate Desert theory2 also fails to solve the problem. Desert theory2 is vulnerableto the wrongful conviction example discussed in the context of fairnesstheory.If the state accidentallypunishedan innocent person of a crime which he then committedafterhis sentence, desert theory2 could not prescribe legal punishmentfor him. He to would have alreadyreceived the legal punishmentproportionate his injustice, and it would give him more legal punishmentthan he deserves for the state to punish him after the actual commission of the crime. Let us revise deserttheoryone last time. Considerdeserttheory3: is justifiedonly if andbecausethe guilty deserveto legal punishment receive legal punishmentconsequentto their injusticethatis proportionate to this injustice. This, finally, solves the problem. Desert theory3does entail that there is pro tanto moralreason for the state to punishall the guilty. However,to account fully for the intuition,desert theory3must not only entail it but also explain it. Why on grounds of desert should the suffering of an offender prior to his sentence not influence whether legal punishmentis deserved? And why on grounds of desert should the legal punishmentof an offender prior to his
13 This objectionis clearly developedby Gertrude Ezorskyin the introduction to her edited volume, Philosophical Perspectives on Punishment(Albany, NY: State Universityof New YorkPress, 1972), pp. xxii-xxvii. 14 For an influentialrecentexample, see Michael Moore, "TheMoralWorthof Retribution," repr.in Joel Feinberg and Hyman Gross, eds., Philosophy of Law, 5th edn. (Belmont, CA: Wadsworth,1995), pp. 632-654.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

508

THADDEUSMETZ

offense not influencewhetherlegal punishment is deservedafterthe offense? As far as I can tell, the defender of desert theory3can answer these questions only by invoking an institutionaltheory of desert accordingto which desert claims are entirelya functionof conventions. For an example, consider grades. A grade is plausibly a school-dependentpositive response deserved for writing an essay. No grade of "A" can be deserved by someone not enrolled in school, for both what is deservedandthe degreeof whatis deserved for writing an essay seem to depend utterly on the institutional context. The same might go for punishment.Perhapsboth what is deservedfor doing an evil deed and the degree of what is deserved are government-dependent negative responses. On this view, only a particularkind of harm, namely, legal punishmentconsequent to the offense, is deserved by those who are unjust, since that is simply what our conventiondictates.Any harmsufferedpriorto the commissionof a crime,or priorto sentencingfor a crimecommitted, is not germaneto what an offenderdeserves, since our institutional normsdictateotherwise. There are two serious problems with this explanationof why the state has pro tanto moral reason to punish all those who have brokenjust law. First, since conventionsdiffer among societies, the institutionaldesert theoristturnsout to be unable to say that every statehas some ethicaljustificationfor punishingall the guilty. Second, it seems wrong to thinkthatdesertclaims aboutpunishment are entirely a function of institutionalpractice. The institutional desert theory implies that if our conventionwere to impose a light penalty for murder,there would be no coherent way to hold that this offense actually deserves a greaterpenalty. But this seems quite counterintuitive. the institutionaldesert Furthermore, still that when a at theory implies large gets hit by a bus, rapist he does not thereby receive any of what he deserves. But getting hit by a bus does seem to be a way for someone who has seriously harmed others to get some of what he deserves. Likewise, getting accidentallypunished by the state before an offense does seem relevant to fixing a proportionbetween the degree of one's offense and the amount of legal punishment one deserves. Our consideredjudgmentsaboutdesertclaims suggest thatlegal punish-

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

509

ment consequent to the offense is not the only negative response which can be deserved for injustice; suffering after an offense but priorto sentencing, as well as punishmentprior to an offense, can to an offense. I conclude apparentlyfigure into desert proportional if that there is pro tantomoralreason even desert entails that, theory3 to punishall the guilty, it cannotprovidean adequateexplanationof why this is so. And an acceptableaccountof the intuitionmust both entail it and explain it well. As with the forward-looking theories, desert and fairnesstheory have the furtherdifficultyof not entailing that all the guilty should pro tanto be punishedin proportionto the gravityof their offenses. Fairness theory must enjoin a small penalty for a serious harm, if the offenderwas wrongly punishedto a large degree by the state in the past or if the benefits received from the state were not substantial. And desert theory, in its naturalformulation,recommends a light penalty for a heinous crime, if, for example, the guilty person had previously sufferedgreatly from naturalcauses or unjustlegal punishment. theoriesinitially appearto be morepromising Backward-looking thanforward-looking theories,since they do not make legal punishment's permissibilitydepend on results and since they intrinsically include accountsof proportionality. However,the problemcommon to fairnesstheoryanddeserttheoryis thatthey makethejustification of legal punishmentdepend on its being proportionalto something beyond guilt which obtained in the past. Fairness theory justifies punishmentonly if it corrects unfair advantage,where unfair advantageis a function of benefits and harms one has received in of desert the past. Similarly,on the straightforward understanding is if it matches the amount of theory, punishment justified only suffering(or punishment)one has coming when takinginto account theory past suffering(or punishment).We need a backward-looking that makes the justification of punishment depend merely on its to the degree of guilt and not on any past benefits being proportional or burdensthat an offender has had. Censure theory does exactly this.

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

510

THADDEUS METZ

V. CENSURE THEORY'S ADVANTAGE

Censuretheorybest entails that and explains why there is pro tanto moral reason for the state to punish all of the guilty in proportion to their guilt. Censuretheory says that the political communityhas a moral reason to punish all those who have broken a just law in proportionto their injustice since it has a pro tanto obligation to in proportion to the degreeof injustice.Granting expressdisapproval is a matterof proporthe assumptionthatproportionate disapproval tionate punishment(as I do in this paper),censure theory naturally accommodates the intuitions about legal punishment addressed here. Censure theory does not justify legal punishmentin terms of results to be produced,as the forward-looking theories do. Hence, it does not hold punishmentof the guilty hostage to good fortune, which may not be forthcoming.Censure theory also does not tie the permissibilityof punishmentto the degreeto which the offender in the past, as both fairnesstheoryand has been benefited/burdened desert theory (on its naturalinterpretation) do. The fact that legal punishmentwill have bad results, or that the state has wrongfully convicted the offender in the past, does not vitiate the state's pro tantodutyto expressdisapproval of those (by meansof punishment) who breakjust laws. I have indicatedin sections III and IV why the forward-looking and rival backward-looking theorieshave respectivelyfailed. I will now indicatethe fundamental reasonwhy none of these theoriescan accountfor the idea that the state has some moral reason to punish all the guilty in proportionto their guilt. The basic problem with the forward-looking theoriesand rivalbackward-looking theoriesis this: they make the moraljustificationof legal punishmentcontingent upon achieving a certain state of affairs which is metaphysically distinctfrom the mere impositionof punishment proportionate to the seriousness of the injustice. The forward-lookingtheories requirelegal punishmentto promotebenefitsin thefuture, whereas the rivalbackward-looking theoriesrequirepunishmentto realize a balance between hard treatmentand benefits/burdens the offender had in thepast. Censure theory, in contrast, does not make the justification punishmentcontingentupon the realizationof any propertymeta-

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

CENSURETHEORYAND INTUITIONSABOUT PUNISHMENT

511

punishmentof the guilty. physically distinct from a proportionate Censuretheory requireslegal punishmentmerely to express disapproval in proportionto the seriousness of the offense. Since all the guilty are propercandidatesfor proportionate censure from the must take the form of political community,which, by hypothesis, punishment,all the guilty are propercandidatesfor proportionate legal punishment.Censuretheory holds that proportionate punishment just is the right way for the political community to express disapprovalof the guilty, where this expressionjust proportionate is the right way for the state to distance itself from injustice, to stand up for victims, and to treat offenders as responsible.Again, disavowing injustice, treatingvictims as valuable, and responding to offenders as moral agents are constitutedby the state censuring which, in turn, is constitutedby proporinjustice proportionately, tionate legal punishment.These propertyidentities are what enable censuretheorybest to entail and explain the intuitionsthatthe state has a pro tanto obligation to punish all the guilty and to do so in to their guilt. proportion
VI. CONCLUSION: CENSURE THEORY'S WEAKNESS

this discussion,I have been grantingthe censuretheorist Throughout three important claims: (1) the state has duties to disavow injustice, to stand up for victims, and to treat offenders as responsible; (2) fulfilling these duties is a matter of censuring offenders proporcensure is a matter tionately to their offense; and (3) proportionate of proportionate punishment.If these claims are true, then, I have censure theory best accounts for the idea that the state has argued, some moral reason to punish all the guilty in proportionto their guilt. However,we cannot assume that these claims are true, and in fact many questiontheir truth.15 Proponentsof censuretheoryhave yet to defend these threeclaims thoroughlyor convincingly. If the argumentsmade here are sound, then censure theory can be said to answer well the questions of whom to punish and how
15 The locus classicus for doubt about censure theory's basic claims is H. L.

A. Hart,Law,Liberty,and Morality(Stanford:StanfordUniversityPress, 1963), pp. 65-66. Many others have since echoed Hart'scriticisms, to which I respond in an unpublishedmanuscript, "Whythe State Must Censurewith Punishment."

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

512

METZ THADDEUS

much to punish. However,it has yet to answeradequatelythe question of why punish. It has not been clearly demonstratedthat the political communityhas some moral reason to express disapproval of injusticeby means of hardtreatment.That one can best accommodateconsideredjudgmentsaboutwhom to punishandhow much only by supposingthereis such moralreason is some evidence that therein fact is. However,much more argumentneeds to be made to establishthatthe statehas a pro tantoobligationto censureinjustice with punishment.Censuretheory's advantageswith respect to the issues of whom is liable for legal punishmentand to what degree make it worthwhile exploring elsewhere whether there is such an
obligation.16
of Philosophy Department Universityof Missouri- St. Louis 8001 NaturalBridge RD St. Louis, MO 63121 USA

following people for useful commentson a different which included the paper germ of the thesis developedhere:Joel Anderson,Larry Davis, Lara Denis, SigurdurKristinsson,David Lyons, Don Scheid, Eleonore Stump, and Roger Wertheimer.I also want to express my gratitudeto Adila Hassim and to the editor of Law and Philosophy for helpful comments on the penultimatedraftof this paper.

16 I would like to thankthe

This content downloaded from 14.139.209.67 on Wed, 16 Apr 2014 15:55:38 PM All use subject to JSTOR Terms and Conditions

Você também pode gostar