10792 Kants theory of punishment.pdf

8/13/2019 10792 Kants theory of punishment.pdf

1/20


2/20

Ka nt's Theory of Punish m entT H O M B R O O K S

University of Sheffield

The most widespread interpretation amongst contemporary theorists ofKant 's theoryof punishment isth a tit isretributivist .On thecontrary,Iwill argue th ere areverydifferent sen ses in which Kant discusses pu nishm ent. He endorses retribution for morallaw transgressions andconseque ntialist considerations forpositive lawviolations.When these standpoints are take n into co nsideration, Kant's the ory of punishmentis more coherent andunified th an previously thought. This read ing uncovers anewproblem in Kant's theory of punish men t. By assuming apotential offender's intention aldisposition asKan t does without know ing it forcertain, wefurther exacerbatetheopportunity formisdiagnosis - although the assumption ofindividual criminalcul-pability may be all we can reasonab ly be expected to use. While this difficulty is not loston Kant,itcontinuestoremain withustoday, making K ant's theory of punishm entfarmore relevant than previously thought.

The duty ofphilosophy was,rather, toremove the deception arising frommisrepresentation, evenat the cost of destroying the most highly extolled andcherished delusion.Immanuel Kant,Critique of Pure Reason1

Here, then,we seephilosophyput infact in a precarious position, which istobe firm even though there isnothingin heavenor on earth from which itdependsor onwhichit isbased.Immanuel Kant,Groundwork of the Metaphysics of Morals 1

I. INTRODUCTIONPerhaps the most striking illustration of a retributivist position, in this caseone that appears to be entirely backward-looking and oblivious to conse-quences, is found in .. . Immanuel Kant.3This may well be the most widespread interpretation of Kant s theoryof punishment amongst contemporary theorists. Most commentatorstoday would disagree with any suggestion that Kant s theory is opento consequentialist considerations.4 On the contrary, I will demon-

1 Immanuel Kant, Critique of Pure Reason, t rans . W. S. Pluhar, Indianapolis,[1781/1787] 1996, p.9[Axiii].2 Immanuel Kant, Groundwork ofthe M etaphysics ofMorals, ed. andt rans .M. J.Gregor, Cambridge, 1991,p. 35 [4:425]. When citing allworks byKa nt excepttheCritique of Pure Reason,thefirst number will referto thepagein theedition cited.Inthe brackets,thefirst numbe r referstothe volume numb er of the P russi an Academy ofSciences edition of Kant's works w ith the second num ber after the colon referring to thepage numberinthis volume wherethequotationisfound. Unless stated otherwise,alluses of emphasisincitations a re given.3 Mark Tunick, IsKant aRetributivist?'History of Political Thought, xvii (1996),p.6 0.4 See H. B. Ashton,Kant s M oral Philosophy,London, 1970; Richard Brandt, Ethical Edinburgh University Press 2003 UtilitasV ol. 15, No. 2, July 2003


3/20

Kant sTheoryof Punishment 207strate that Kant's theory of punishment is not purely retributive.6I believe that much of the reason for interpretative confusion is dueto the very different senses in which Kant speaks of punishmentregarding moral and positive law. When these standpoints are takeninto consideration, I believe Kant's theory of punishment is morecoherent and unified tha n previously thought.

I will argue that Kant's writings often justify punishments fortransgressions of positive law on non-retributive grounds. Thisinterpretation is contrary to m ost of the scholarship in th e field. Theconfusion lies for the most part with Kant's retributive concernsregard ing violations of the moral law, inmyview. Most comm entatorsseem to understand the plentiful discussions of punishments fortransgressing moral law as either K ant's entire theory of punishmentTheory, Englewood Cliffs, 1959, pp. 496-8; Richard Dean, 'Cummiskey's KantianConsequentialism', Utilitas, xii (2000); Samuel Fleischacker, 'Kant's Theory ofPunishment ' , Essays on Kant s Political Philosophy, ed. H. Williams,Cardiff, 1992;Katrin Flikschuh, 'On Kant 's Rechtslehre , European Journal of Philosophy, v (1997);Mary J. Gregor,Laws of Freedom: A Study of Kant s Method of Applying the CategoricalImperative in the Metaphysik der Sitten, Oxford, 1963; Jea n Ha mpton, PoliticalPhilosophy,B oulder, 1997, pp. 134 f., 147,175 ; Sarah W illiams H oltman, 'Toward SocialReform: Kant's Penal Theory Reinterpreted', Utilitas, ix (1997); Conrad D. Johnson,'The Authority of the Moral Agent', Consequentialism and its Critics, ed. S. Scheffler,Oxford, 1988, p. 273; Christine M. Korsgaard, Creating the Kingdom of Ends,Cambridge, 1996; Jeffrie G. Murphy, 'Kant's Theory of Criminal Punishment',Proceedings of the Third International Kant Congress,ed. L. W. Beck, Dordre cht, 1972,p. 434; Edmund L. Pincoffs, The Rationale of Legal Punishment, Atlantic Highlands,1966,p.9; LouisP.Pojman, 'For the Dea th Penalty ,The Death Penalty: Forand Against,ed. L. P. Pojman and J. Reiman, Oxford, 1998, pp. 1, 7-11; John Rawls, PoliticalLiberalism, New York,1993;John Raw ls,A Theory of Justice, Cambridge,1971; JeffreyReima n, Why the De ath Penalty Should Be Abolished',The D eath Penalty,ed. Pojmanand Reiman, pp. 90-2; Patrick Riley,Kant s Political Philosophy, Totowa, 1983; HansSaner K ant s Political Thought: Its Origins and Development, trans. H. B. Ashton,Chicago, 1967; Nancy Sherman, Making a Nece ssity of Virtue: Aristotle a nd K ant onVirtue, Cam bridge, 1997; Avital Simhony, Wa s T. H. Green a Utilit arian ?' Utilitas,vii (1995); Michael Slote, From Morality to Virtue, Oxford, 1992; Roger J. Sullivan,Immanuel Kant s Moral Theory, Cambridge, 1989, pp. 243 f., 361n25; and HowardWilliams,Kant s Political Philosophy,Oxford, 1983.

6 See my 'Corlett on K ant, H egel, and R etribution', Philosophy, lxxvi (2001);B.Sharon Byrd, 'Kant's Theory of Punishment: Deterrence in its Threat, Retribution inits Execution', Law and Philosophy, viii (1989); J. Angelo Corlett, 'Making Sense ofRetributivism', Philosophy,lxxvi (2001); David C ummiskey,Kantian Consequentialism,Oxford, 1996; David Cummiskey, 'Ka ntia n C onsequentialism',Ethics, 1(1990); Robert P.George, Making Men Moral: Civil Liberties and Public Morality, Oxford, 1993, pp.148-53, esp. 149n39; R. M. Hare, 'Could Kant Have Been a Utilitarian?' Utilitas, v(1993); Don Scheid, 'Kant's Retributivism', Ethics, xciii (1983); and Tunick, 'Is Kanta Retributivist?' Also see A. Phillips Griffiths, 'Kant's Psychological Hedonism',Philosophy,lxvi (1991); Frederick N euho user's editorial footnote atp .245nl4 in JohannGottlieb Fichte,Founda tions of Natural Right, ed. F. Neuhouser and trans. M. Baur,Cambridge, [1795-1796] 2000; and David Weinstein, 'Between Kantianism andConseque ntialism in T. H. Green's Moral Philosophy',Political Studies, xli (1993), esp.p. 631 where he claims tha t Green reinterp reted K ant as a conseque ntialist.


4/20

208 Thorn Brooksor some ideal penal theory divorced from empirical concerns.6 Inaddition, some interpreters note the different senses in which Kantdiscusses punishment, but fail to see his theory of punishment ascoherent.7Myreading uncovers a new problem in K ant's theory of punishm ent:How are we to assess the intentional attitudes of criminals inattribu ting blame and possible sentencing?Thepotential difficulty liesin his pronouncement that we can never know for sure the criminalculpability of any potential offender, making the possibility of ourpunishing innocentpeoplea troubling reality.Byassuming a potentialoffender's intentional disposition without knowing it for certain, wefurther exacerbate the opportunity for misdiagnosis - although theassumption of individual criminal culpability may be all we canreasonably be expected to use. While this difficulty is not lost onKant, it continues to remain with us today, making Kant's theory ofpunishm ent far more relevant than previously thoug ht.

II. IS KANT'S THEORY OF PUNISHMENTRETRIBUTIVE?For Kant, '[t]he law of pun ishm ent is a categorical imperative'.8Beinga categorical imperative, the law of punishment must hold universallywith necessity for all human beings, leaving no one outside the law'sgrasp.9 Moreover, there should not be any contradictions amongstsatisfactory legal maxims. If any such contradictions arise, then themaxims leading to them are unsatisfactory. Thus, laws of punishm entare byextension unive rsal laws tha tdonot allow for any special cases.The categorical imperative is also known as 'the moral law'. Under

6 An interesting exception is Williams who claims tha t Kan tian punish me nts mus t bejustified from two standpoints: (1) the moral law's retributivism and (2) an empirico-utilitarianism. Therefore, punishments ought to both fit crimes and deter potentialoffende rs. (Williams, p. 106; see p. 101.) In add ition to Williams, Fleischacker confusesthese two standpoints (Fleischacker, p. 193).7 See Jeffrie G. Murphy, 'Does Kant Have a Theory of Punishment?' Columbia LawReview, lxxxvii (1987). At 509, Murphy says: 'I am not even sure that Kant developsanything tha t deserves to be called a theoryof punish men t a t all. I genuinely wonder ifhe has done much more tha n leave us with a random (and not entirely consistent) set ofremarks . Holtman attempts to reinvent Kant so she can 'reject the law of retributionand other disturbing de tails, but em brace more basic aspects of the a ccounts of justiceand punish men t'. Nevertheles s, she says tha t his theory is at best incomplete. (Holtman,12, 16, 21; also see Hannah Arendt, Lec tures on K ant s Political Philosophy, ed .R. Beiner, Chicago, 1982, pp. 7 f; S tu ar t Brown, Has Ka nt a Philosophy of Law? ',Philosophical Review,lxxi (1962); Fleischac ker, p.193;Saner,p.2; and Slote, pp. 31-57.)8 Kant , The Metaphysics of Morals, ed. and trans. M. J. Gregor, Cambridge, 1996,p.105 [6:331].9 Necessity arises from a unive rsal duty to respect the moral law. (Kant, Groundwork,p.13 [4:400]. See ibid., p. 14n [4:402].)


5/20

Kant sTheoryof Punishment 209the moral law, all laws are to be determined independently of ourdesires and inclinations in an effort to ensure universality.10 Thisrejection of a subjective standpoint has led most of Kant's critics tocharge his ethical theory with empty formalism, mainly for its havingthe appearance of being an ethics based purely on logical criteria.11For Kant, when we perform a particular action, it is supposed th atwe are stating to our community that all may act in the same m anneras we have.12As an example, if a thief stole property from someone,the thief s action would effectively establish a unive rsal maxim tha teveryone might do the same. Likewise, if a m urderer were to kill aninnocent victim, the mu rderer's action would by extension establish auniversal maxim sanctioning the ability of any citizen to kill innocentvictims. In addition, criminals use other persons as a means to theirown particular ends, in m aking themselves an exception to universalprinciples. This is to say th at a thief desires to steal property so th atshe might benefit from its ownership: The thief would not like herbelongings stolen from he rbya secondthief.For Kant, th e u se of otherhuman beings as a means to an individual's subjective end for anyreason is morally objectionable.13The difficulty with theft and murder- amongst many other crimes - is that they ought not to become

10 See J. Marina, 'Kant's Derivation of the Formula of the Categorical Imperative:How to Get it Right', Kant-Studien, lxxxix (1998), pp. 167 f.11 Kant says: 'All I need for morality is that freedom does not contradict itself andhence can at least be thought; I do not need to have any further insight into it ' (Kant,Critique of Pure Reason,trans. Pluhar, p. 30 [B xxix]). See Hare, 11; Fichte, pp. 245 f.[282-3 20 Remark]; Georg Wilhelm Priedrich Hegel, Elements of the Philosophyof Right, ed. A. W. Wood and tr an s. H. B. Nisbet, Cam bridge, [1820] 1991, pp. 135-40[ 105-12, with Rem arks and A dditions]; see ibid., p. 43[9 ];Kant ,Groundwork,p. 3;Daniel N. Robinson and Rom Ha rre, The Demography of the K ingdom of Ends ',Philosophy, lxix (1994); J. B. Schneewind, 'Autonomy, Obligation, and Virtue: AnOverview of Kant's Moral Philosophy , Cambridge Companion to Kant, ed. P. Guyer,Cambridge, 1992; Peter J. Steinberger, 'The Standard View of the CategoricalImperative',Kant-Studien, xc (1999), pp. 91, 99; and B ernard Williams, Ethics and theLimits of Philosophy, Cambridge, 1985. Also see Henry E. Allison, Kant s Theory ofFreedom, Cambridge, 1990, pp. 180-98; Karl Ameriks, 'The Hegelian Critique ofKantian Morality*,New Essays on Kant,ed .B.den Ouden and M . Moen, New York, 1987;Marcia Baron, Kantian Ethics Almost Without Apology, Ithaca, 1995; P.-C. Lo,'A Critical Re-evaluation of the Alleged Empty Formalism of Ka ntian E thics', Ethics,xli (1981); Onora O'Neill,Acting on Principle,N ewYork,1975;Jeffrey E dwa rds, 'Egoismand Formalism in the Development of Kant's Moral Philosophy , Kant-Studien, xci(2000); and Christian J.Onof, 'A Framew ork for the Derivation a nd Rec onstruction ofthe Categorical Imperative', Kant-Studien, lxxxix (1998).12 Often, this is solely attributed to Hegel. See Hegel, Elements of the Philosophyof Right, pp. 49 [ 15 Add ition], 113 [81 ,Remark], 115 f. [ 82, Addition], 124 f. [ 99,Remark], 126 f. [ 100, Remark], and 131 f. [ 104]. See also ibid., pp. 117 f. [ 86,Addition] and 104 [ 73].13 See Kant, Groundwork,p p. 37-40 [4:428-32]. We should not forget tha t 'm oralityis the condition under which alone a rational being can be an end in itself, since onlythrough t his is it possible to be a lawgiving member in the kingdom ofends'(ibid., p. 42[4:435]).


6/20

21 0 Thorn Brooksuniversal actions, as it would underm ine rig hts to ownership, liberty,and, most importantly, the moral law.14 As lawbreakers have wrong-fully subjected the law-abiding public to harmful actions, Kantbelieves th at it is only ju st to then subject th e former to the harm fulmaxim s they attem pted to impose on their community.15This part of Kant's moral theory is seldom disputed. The difficultyis in accurately interpre ting the justifications h e offers to sup portthe right of the state to punish these criminals for their crimes. Themost popular understanding is that Kant's theory of punishment isa retributivist theory, no t a utilitarian nor a consequentialist theory.Indeed, Kant's political writings contain several criticisms of the useof pun ish m ent as a kind of tool to realize future goals.16As an example,Kantian punishments must a lways treat human beings as ends-in-themselves an d never as a me ans to some future goal: Punishment bya court (poena forensis) ...can never be inflicted merely as a means toprom ote some other good for th e crim inal him self or for civil society. Itmust always be inflicted upon him only because he has committed acrime .11 Punishments are to be administered solely on the basis ofcriminal guilt, not on consideration of social utility. 18 In addition, thepartic ular form a punish me nt takes is to be specified by the kind andseverity of crime committed. Therefore, a punishment's form shouldnot be influenced by extraneous concerns.

One of the most oft-cited passages of Kant's in favour of the viewtha t he justifies p unish me nt p urely on retributive grounds is his 'bloodguilt ' example inThe Metaphysics of Morals:Even if a civil society were to be dissolved by the consent of all its members(e.g., if a people inhabiting an island decided to separate and dispersethroughout the world), th e last murd erer rem aining in prison would first haveto be executed,soth at each has done to him w hat his deeds deserve and bloodguiltdoesnotcling tothe people for not having insisted upon this punishmen t;

14 Kant's moral theory would not tolerate all universal maxims, such as 'destroyhuman life whenever you please', as the respect for persons must remain primary as arule. This would contradict Hare's re ading. However, Hare is quite correct to point th at'[m]oral principles do not have to be as simple and general as Kant seems to havethought, and they can still be universal all the same' (see Hare, 'Could Kant Have Beena U tilitarian? ' pp. 7 f.).15 See my 'Corlett on Kant, Hegel, and R etribution', p. 563.16 'But to look upon all punishments and rewards as mere machinery in the hands ofa higher power, serving only to put rational beings into activity toward their finalpurpose (happiness) is so patently a mechanism which does away with the freedom ofthe will that it need not detain us here' (Kant, Critique of Practical Reason, ed. andtrans.M .J. Gregor, Camb ridge, [1788] 1997, p. 35 [5:38] and K ant, Practical Philosophy,ed. and trans. M. J. Gregor, Cambridge, 1996, pp. 170 f). See Kant's essay 'The Endof All Things' in Kan t,Religion within the Boundaries of Mere Reason, ed. A. Wood andG. di Giovanni, Cambridge, 1998, p. 204 [8:338-9].17Kant ,M etaphysics of Morals,p. 105 [6:331].18 Likewise, rule-u tilitarians can justify punishm ents without being retributive.


7/20

Kant s Theory of Punishment 211for otherwise the people can be regarded as collaborators in this publicviolation of justice.19While all other debts might or might not be forgiven and all othercontracts perhaps left unfulfilled, it is only with the punishing ofcriminals where we are commanded to treat them strictly andseverely. The impo rtance K ant gives to punis hing violations of justiceis no mino r concern, for 'if just ice goes, the re is no longe r an y valu e inhuman beings living on the earth' .20 In fact, some of our value asperson s is par tly derived from our sacred defence of just ice in hu m ancommunity.

Not to be overlooked, Kant's theory of punishment is often linkedwith the lex talionis: 'whatever undeserved evil you inflict uponanother within the people, you inflict upon yourself.If you insult h im,you insultyourself; if you steal from him, you steal fromyourself; ifyou strike him, you strikeyourself;if you kill him , you killyourself. 21Indeed, with regard to executing murderers, Kant in fact arguesfor 'the strict law of retribution' in '[t]his fitting of punishment tothe crime'.22 This is justified on the grounds th at strict justice is notobjectionable in the least when the criminal's inner wickedness isparticularly noticeable.23 It must be said that it is rather difficult toreconcile Kant's motto of equity ( ' the strictest right is the greatestwrong') with this insistence that all murderers are to be killed aspunishment for killing someone else.24 Nevertheless, what is criticalfor Kant in justifying capital punish me nt for mu rder ers is themurderers ' demonstration of inner wickedness.

19 Kant ,Metaphysics of Morals,p. 106 [6:333]. It is easy to forget th at we also have aduty to forgive fellow human beings, which at times may result in abandoning thedistribution of punishments to culpable violators of moral and positive law (ibid.,p.208 [6:461]).20 Ibid., p. 105 [6:332]. Hegel think s just the opposite: 'Let justice be done' should nothave as its consequence 'even if the world should perish' (Hegel, Elements of thePhilosophy of Right,p. 157 [ 130]).21 Ibid. See Kant, Lectures on Ethics, ed. P. Heath and J. B. Schneewind, trans.P. Heath, Cambridge, 1997, pp. 310 f. [ 49 27:555]. (I am citing a passage fromJoha nn Friedrich Vigilantius's notes from Ka nt's 1793-4 lecture course entitled TheMetaphysics of Morals'.) Eugen Duhring argued that the lex talionissatisfied too low apenal threshold, as to properly negate an offending will one must cause more harm tothe criminal than he or she committed to make up the shortfall of his or her 'lack ofregard for others shown by that will in the crime' (Small, 42). At times Kant seems tobe somewhat sympathe tic to this approach: Every deed that violates a hum an being'sright deserves punishment, the function of which is to avengea crime on the one whocommitted it (not merely to make good the harm that was done)' (Kant,Metaphysics ofMorals,p. 207 [6:460]).22 Kant ,Metaphysics of Morals,p. 106 [6:332].23 Ibid., pp. 106 f. [6:333]. Othe r inte resting examples involve Ka nt's suggestedpunish me nts for rape , pederasty, and bestiality (ibid., p. 130 [6:363]).24 Ibid., p. 27 [6:235].


8/20

212 Thorn BrooksPerhap s because of its discord with modern notions of morality, oneaspect of Kantian punishment that islittle mentioned in the litera tureis the fact that, for Kant, punishment is a physical harm, properlyunderstood.25 Psychological punishments or non-painful punishmentsare not seriously considered for most crimes.26 In fact, the physicalinfliction of punishment upon a criminal is a necessary consequence

in th e allocation of justice, even if the criminal transg ressionprecipitating the punishment is not naturally connected with moralwickedness.27 In a sense, the criminal 'pays' for the crime shecommitted by experiencing a certain amount of, potentially lethal,pain.28This is not to imply tha t K ant was some kind of sadist - to beclear, he was certainly not.29 In fact, he finds public rewards offeredby the government to be 'more in harmony with morality' thanpunishments.30Idonot believe th at this evidence is enough to justify labelling K anta retributivist, for his theory seems to alsogoin a nother direction: Heis quite explicit in stating that governments can and do only imposedeterrent punishments.31 For Kant, governments are concerned pri-marily with pragmatic m atters, rathe r th an with m atters of justice,all things considered. One example that has been ignored by nearlyall commentators is the fact that in certain instances no punishmentshould be inflicted, even where the guilty parties deserve to bepunished.32 In the 'Doctrine of Right', Kant supposes that we havebefore us murderers and their accomplices. He tells us that theidea of judicial legal authority dictates th at all of these p ersons m ustbe executed, in accordance with universal laws - the categoricalimperative regarding punishment. Nevertheless, Kant asserts that

25 Kant,C ritique of Practical Re ason,p. 34 [5:37] and Kant,Lectures on Ethics, t rans .He ath , p p. 79 [27:286], 304 [ 38 27:547], 308 f. [ 43 -4 27:552 f.], 312 [ 50 27:556].(From theLectures,the first citation is from the section 'Of Rewards and Punishments'from Georg Ludwig Collins's 1784 notes. The following citations are Vigilantius's notes.)26 There is an exception for crimes where the appropria te ma nne r of punishingcriminals is to levy fines. Moreover, wealthy criminals who will not be 'hurt' by th e fineare to be publicly embarrassed, perhaps a form of psychological punishment. (Kant,Metaphysics of Morals,p. 106 [6:332].)27 Kant,C ritique of Practical R eason,pp .34 f. [5:37] and Ka nt,Metaphysics of Morals,p.104 [6:331].28See Tom Sorell,Moral Theory and Capital Punishment, Oxford, 1987, p. 134.29 'No punish men t should be coupled with cruelty Kan t,Lectures on Ethics, t rans .Heath, p. 311[49 27:556].30 Kant ,Lectures on Ethics, tra ns. L. Infield, Indianapolis, 1963, p. 57.31 Ibid., p. 55 and Kant,Lectures on Ethics, tra ns. H eath, p. 79 [27:286] in the section'Of Rewards a nd Pu nishm ents'. See th e section entitled 'Of the La wgiver' in ibid., p. 79[27:286].(Both are from the notes of G. L. Collins.) This state me nt m irrors Fichte's belieftha t the purpose of punis hme nt is solely to deter. (See Fichte, pp. 226-48 [260-8520].)32 I know of only three tre atm ent s: Byrd, 196, 196nl44; Scheid,268-71; and Tunick,63 . Corlett takes this view only with regard to his own reinterpretation of a Kantianpenal the ory (Corlett, 78, 78n7).


9/20

Kant sTheoryof Punishment 213should the number of criminals be so great a number that the statewould be left without any subjects if they were all executed, 'thesovereign mu stalsohave it in hispower,in thi s case of necessity(casusnecessitatis),to assum e the role of judge (to repre sent him) andpronounce a judgem ent th at decrees for the criminals a sentence otherthan capital punishment, such as deportation, which still preservesthe population'.33In this example, the preservation of the community outweighsretributive concerns when administering particular punishments -in violation of much of what form punishment ought to take as acategorical imperative.Kant's use of the sovereign offers another example that does notsupport a retributivist reading. First, he declares that 'the presentlyexisting legislative authority ought to be obeyed, whatever its origin',even though this authority may be completely at odds with reasonor morality.34 People would then be forced to live according to lawstha t may violate universality, for instance. In addition, citizens haveno rights to sedition or rebellion - even if the sovereign abuses hisauthority. Any attempt to do either is high treason, punishableby state execution.35 Kant says: '[l]ike a chasm that irretrievablyswallows everything, the execution of a monarch seems to be a crimefrom which the people cannot be absolved, for it is as if the statecommits suicide'.36Finally,no tonly isthe sovereign immune from legalpunishm ent as king, but even if he should be dethroned, he cannot bepunished for any actions he performed while a sovereign.37

Even when reiterating his commitment to the lex talionis, Kanttells us that punishments ought to be less severe than the criminaldeserves when the punishment would be a crime against humanity:Those convicted of rape or pederasty should be castrated, those ofbestiality expelled from the comm unity for life, etc.38His argument isthat we may forgo punishing to the letter of the law, so long as thepenalty we choose is commensurable to its spirit.3933 Kant ,Metaphysics of Morals,p. 107 [6:334].34 Ibid., p. 95 [6:319]. Should a citizen resist the rule of his sovereign, she Nvould bepunished, got rid of, or expelled in accordance with the laws of this authority, that is,with every right' (ibid.).35 Ibid., p. 96 [6:320]. He says further: 'It is the formal execution of a monarch thatstrikes horror in a soul filled with the idea of human beings' rights, a horror that onefeels repeatedly as soon as and as often as one thinks of such scenes as the fate ofCharles I or Louis XVI' (ibid., p. 97n [6:320]).36 Ibid., pp. 97n-98n [6:320] (emphasis added). As 'the people' are implicated in acrime most, if any at all, did not participate in in any way, Kant sanctions yet again theviolation of a univers al m axim.37Ibid., pp. 95, 98 [6:319, 323].38 Ibid., p. 130 [6:363].39 Ibid.


10/20

21 4 Thorn BrooksThe most well-known exam ple of Ka nt's which appe ars to justify theprimacy of consequentialist above retributive concerns regards twoshipwrecked men fighting over a raft that will only hold one of them.40Should one of them pus h the other off the raft - drowning him - inorder to save his own life, the surviving man is ' to be morally con-demned but n ot legally punished' .41 Ka nt's reason is explicit:

the punishment threatened bythe law could notbegreater than the loss of hisown life. A penal law of this so rt could not have th e effect intended, since athreat of an ill that is still uncertain (death by a judicial verdict) cannotoutweigh the fear of an ill tha t is certain(drowning). Hence the deed of savingone's life by violence is not to be judged inculpable (inculpabile) but onlyunpunishable (impunible).42The surviving person is to be relieved of legal punishment in partdue to his fulfilling a duty to preserve his own life.48 In addition, thesurviving person avoids punishment only because punishing himwould not have the desired effect that punishments ought to have,derived from ' thre at ' . Kan t argues also again st punishin g mo thers whokill their illegitimate children and soldiers who murder each other induels on the same gro unds: These persons oug ht not to be punished fordoing so would notdeter others in committing the same actions in thefuture.44

At first glance, it might appear as if Kant lacks a coherent theoryof punishment. As is well known, in some instances Kant seems tobe concerned solely with retributive justifications for punishing andfor particular punishments. For example, recall the illustration of theisland community about to dissolve with a murderer left in its prison:The community must execute the mu rder er for the sake of justice orotherwise the community shares in the m urdere r ' s guil t. Almost to thecontrary, I have shown that there are many other instances whereKa nt seem s to justify pun ishin g criminals solely upon co nsequentialistgrounds, where if persons cannot be deterred by the threat of punish-ment, they ought not to be punished at all. I would propose that theapp aren t contradiction is due mo stly to a misread ing of Kan t, for when

40 See Kant,Metaphysics of Morals,p. 28 [6:235 f.].41 I take this phra se from Tunick, 65.42 Kant , Metaphysics of Morals, p. 28 [6:235 f.] (emphasis given). See Jeffrie G.Murphy,Kant: The Philosophy of Right, London, 1970, p. 130n3.43 Kant , Groundwork, p. 11 [4:397]. Fleischacker argu es th at in this example 'Ka ntdoes not hold tha t the impossibility of dete rring someone morally justifies not punish inghim; on the contrary, he stresses the fact that deterrence concerns only the subjective(empirical), not the objective value of punis hme nt' (Fleischacker, 194). The problem w iththis view is that Fleischacker has misrepresented Kant's penology by not taking intoconsideration that punishment is itself an empirical practice - and, for Kant, govern-ments can only administer deterrent punishments.44 Kant ,Metaphysics of Morals,pp. 108 f. [6:336 ].


11/20

Kant s Theory of Punishment 215Ka n t iswrit ing infavour ofone version ofpun ishment ra the r thananother heisdoing so in very different contexts.

I I I . T W O S P H E R E S OFLAWFor Kan t, there is adichotomy of jurid ical (i.e. positive) law an d mo rallaw. Juridical laws concern themselves solely with external actionsand their conformity toagovernmen t's legislation- which may or maynotberational.46 Inordertofollowajuridical law w e m ustactsolelyin accordance with w ritten legislation governingahum an community .In sodoing,our actions maybe perfectly legalbut not always com-mensurable with s tandards set bymorality. Juridic al law s requireincentives other than moral duty,as this form ofdutyis an internalincentive.46Asan example, Kant says: It is anexternal du ty to keepa promise made in a contract;but thecommand to dothis merelybecauseitisaduty, without regard for any o ther incentive, belongstointernal lawgiving alone'.47

When weactfrom dut y alon e,we actwith respecttom orali tyandthe moral law. Acting otherwise isprimari ly a m a t t e r forjuridicallaw.48Aswehav e seen previously, forKant , themoral law is firstandforemost expressed in the categorical imperative.49 This imperat ivecomm ands us to act only in such away tha t our maxim- expressed byour action- can beauniversal law.50As ration al beings, we act w ithina 'kingdom ofends' wherebywesubject ourselve s touniversal lawsa t one and the same t ime as weperform partic ular actions.51 Mostimportantly, this law becomes the st and ard by which we may judge themoral correctness ofevery par ticu lar action. Beforeweperformanyaction wem ust first consider consciously or unconsciously the

45 See Kant,Metaphysics of Morals,pp. 14, 17 [6:220, 224]. As positive laws may beirrational, Kant s argument against the right to sedition or rebellion against the statemay be problematic (see ibid., pp. 96 f. [6:320]). In addition, human beings are only endsin themselves in morality, not in positive law (Kant,Groundwork,p. 42 [4:435]). This isnot to say that Kant s project of a kingdom of ends is complete fantasy, for one of theprimary foundations of a kingdom of ends is its possible realization through humanconduct (ibid., p. 44n [4:437]). Kant hints that there would be no need for punishment inthe ideal society (Kant,Critique of Pure Reason, trans. N. K Smith, London, 1963,A317/B373; see Fleischacker, 191 and Fichte, 244 [20 282V(f.)]).46 Kant,Metaphysics of Morals,p. 20 [6:219]. See ibid., p. 23 [6:230].47 Ibid., p. 21 [6:220].48 See ibid., p. 20 [6:219].49 Ibid., p. 15 [6:222-3]. See ibid., p. 19 [6:227] and Kant,Groundwork,pp. 5 [4:392];30-3 [4:420-4] and 39 f. [4:430-2].50 Kant,Groundwork,pp. 15 [4:402], 16 [4:403], and 31 [4:421].51 Ibid., pp. 41 [4:432] and 42 [4:434]. See ibid., p. 40 [4:432].


12/20

21 6 Thorn Brookseffect on society should all its mem bers ad opt the m axim on which weare in tend ing to act. For exam ple, if we were to consider lying, stealing,or committing a murder, we are to reject this behaviour on the groundsthat everyone might lie, steal, or murder each other, bringing about aworld without tr us t or prom ises, protection of property, or basic h um anliberty. By acting in accordance w ith morality, a hu m an being can fullyexist as an end-in-itself, 'since only through this is it possible to be alawgiving member in the kingdom of ends'.52

Were we to punish criminals from this purely moral perspective,i t would be essential tha t we justify p unish me nt prim arily as an actof justic e: 'this con stitutes w ha t is essen tial in this concept [ofpunishment] ' .6 3 In other words, punishment must be primarily com-men surable w ith justice in order for i t to be distr ibuted to a particularperson. Therefore, all punish me nts ought to be given only to the g uiltyand within certain l imits. With regard to the violation of morallaws, Kant is clearly in favour of retributive measures of punishingcriminals. On the o ther ha nd, from the standpo int of juridical , orpositive, law, Kant is concerned primarily with deterrence-based justi-fications for punishment. This crucial dichotomy is rarely appreciatedby commentators on Kant's political philosophy.54It is important to note that this matter - which he called a'quandary' - was not left unnoticed by Kant, for he says:The knot can be undone in the following way: the categorical imperative ofpenal justice remains (unlawful killing of another mustbepunishedbydeath);but the legislation itself (and consequently also the civil constitution), as longas it remains barbarous and undeveloped, is responsible for the discrepancybetween the incentives of honour in thepeople(subjectively) and the m easuresthat are (objectively) su itable for its purpose. Sothe public justice a rising fromthe state becomes an injusticefrom the perspective ofthejustice a rising fromthe people.55

52 Ibid., p. 42 [4:435]. See ibid., p. 15 [4:403] and Kant, Metaphysics of Morals,p.182 [6:429].53 Kant ,Critique of Practical Rea son,p. 34[5:37].54 See Kant, Metaphysics of Morals, pp. 17, 20 [6:219, 225] and 10, noted. Thebest treatment of Kant's theory of punishment is Tunick, 60-78, esp. 77 f. A fairlysimilar dichotomy is present in Hegel (see Hegel, Elements of the Philosophy of Right,p. 121 [ 94 Addition]). Also see Hare, 11 and Kant, Groundwork, pp. 57 [4:452] and61 [4:457].55 Kant , Metaphysics of Morals,p. 109 [6:336-7]. See Allen W. Wood, Kant s EthicalThought,Cambridge, 1999, p. 322: 'Kant holds that we have amoralobligation to limitourselves to actions th at are rig ht [R], but th at d uty is no pa rt of R itself.R grounds only juridical duties, which are distinguished from ethical duties by the factthat their concept contains no determinate incentive for complying with them'. Unfor-tunately, there is no discussion of the role(s) of moral and juridical duties regardingpunis hme nt. (Ibid., p. 407n32 and Allen W. Wood, 'Kant's Practica l Philosophy',Cambridge Companion to German Idealism, ed. Karl Am eriks, Cambridge, 2000.)


13/20

Kant s Theory of Punishment 21 7From the standpoint of the moral law, the positive law may well beinadequate. Indeed, we might best evaluate juridical law via morallaw. However, this is ju st one standp oint. The othe r stand point is th atof the state which deals with practical matters of external actions,which it must try to influence. In practical concerns, such as statelegislation, just ice is served only wh en the le tter of th e law is observed.For Kan t, the be st of all possible hum an comm unities will be one th atfully develops itself wh en it perfectly unifies th e sph eres of jurid icaland m oral law.

This noble ideal was not a reality in Kant's time nor is it truetoday. Human societies will remain in a 'barbarous and undeveloped'condition until they are able to comm ensurate juridical and moral law.Thus, our a t tem pts to determine the m ost jus t punish men t for aparticu lar crime begin in an undeveloped condition, where the conse-quential effects of a punishment take priority over the satisfaction ofpurely retributive demands. We may improve our attempts over timeby greater incorporation of retributive standards into our theory ofpunishment. Crucially, while the form a punishment takes mayeventually begin to prioritize retribution over consequences, punish-ment takes a consequentialist form at first, as it attempts to shapeexternal human behaviour. Punishing criminal conduct strictly inaccordance with w hat th e moral law dictates is the u ltim ate aim of thepractice, but not w here we begin.66

Kant states in his 1792 letter to Johann Benjamin Erhard:In a world of moral principles governed by God, punishments would becategorically necessary (insofar as transgressions occur). But in a world ofmoral principles governed by men, the necessity of punishments is onlyhypothetical, and that direct union of the concept of transgression with theidea of deserving punishmen t serves the ruler only as a prescription for whattodo.Soyou are right in saying that th epoena meremoralis[ ethicalpenalty*](which perhaps came to be calledvindicativa ['avenging punishment'] for thereason th at it preserves the divine justice), even if its goal is merely m edicinalfor the criminal and the setting of an example for others, is indeed a symbolofsomething deserving punishment, as far as th e condition ofits authorizationis concerned.57In th is passage, Kant agrees with Ernst Ferdinand Klein that'punishment is the symbol of an action's deserving punishment, by

56 For example, Kant tells us that the legal system he is describing is that of 'thestate in idea, as it ought to be in accordance with pure principles of right' (Kant,Metaphysics of Morals,p. 90 [ 45 6:313]).67 Kant, Kant: Philosophical Correspondence, 1756-99, ed. and trans. A. Zweig,Chicago, 1967, p. 199 [11:398-9] and Kant, Correspondence,ed. an d t ran s. A. Zweig,Cambridge, 1999, p. 448. See Kant, M etaphysics of Morals, pp. 24-7, 104-6 [6:231-4,331 f.]. Fleischacker cites only the first half of this letter, which may have furthercontributed to his m isinterp retation of it (Fleischacker, 204).


14/20

21 8 Thorn Brooksmeans of a m ortification of the criminal th at corresponds to the crimecommitted'.68 More importantly, Kant agrees with the interpretationof his penal theory being offered, as he admits that 'the necessity ofpunishments is only hypothetical' with regard to 'a world of moralprinciples governed by men'. Rather than justifying punishmentscompletely devoid of retributive concerns, Kant argues instead thatthese concerns ought to be considered as 'a prescription for what todo'.Thus, Kant tempers a consequentialist theory of punishment bymoulding punishment in a shape that would fulfil retributive stan-dards.69However, this m atter is one of retributive concerns helping toform what is primarily consequentialist punishment and notviceversa.Furthermore, I would like to suggest that when Kant speaks ofsovereigns60'performing injustice in the high est deg ree' when offeringpardons or reduced sentences to criminals, the injustice in qu estion ismoral, not juridical.61Such a m atter is related solely to the m oral lawexisting internally in human beings, which we can discern barelyand w ith much difficulty. From the s tandp oint of jurid ical law thesovereign who offers pardons acts justly if this pardon positivelycolours certain consequences. In fact, it would be a grave injustice if asovereign did not pardon criminals where carrying out the ir sentencewould lead to the breakdown of the community.62

58 See Kant, Kant: Philosophical Correspondence,p. 199n3. Ka nt s ays of Klein oncriminal right: 'Most of wh at he s ays is excellent and quite in accord with my own view'(ibid., p. 199).59As he says in the letter to Erhar d, the authorization of punishment is dependentupon retributive justification (esp. its concept of desert), subsumed under a primarilyconsequentialist theory of punishm ent.60 Kan t's use of 'sovereign' is a term easy to confuse. In the Groundwork, th e'sovereign' is each rational being in the kingdom ofends:'He belongs to itas sovereignwhen , as lawgiving, he is not subject to the will of any other ' (Kant, Groundwork,p. 41[4:433]). However, in Metaphysics of Morals,the sovereign is a single person who alonerules over his community (Kant,Metaphysics of Morals,pp. 95 -8 [6:319-23]). I believethat 'sovereign' in the first sense is related to individual autonomy and in the second itis related to a particular individual (i.e. a monarch).61 Kant, Metaphysics of Morals, p. 109 [6:337]. See Don Becker, 'Kant's Moral andPolitical Philosophy*, Routledge History of Philosophy, vol. 6, The Age of GermanIdealism, ed. R. C. Solomon and K. M. Higgins, London,1993,p.84 1believe that JeffrieMurphy was mistaken in declaring that Kant held no such view, as Kant says:'conformity with juridical laws is the legality of an action and conformity with ethicallaws is its morality (Kant, Metaphysics of Morals, pp. 14 [6:220] and 17 [6:225]; seeJeffrie G.Murphy, 'K an t's Con cept of A [sic] Righ t Action', Kant Studies Today,ed. LewisWBeck, La Sa lle, 1969, pp. 471-95).62 Kant ,Metaphysics of Morals,p. 107 [6:334]. To recall Kant's 'blood guilt' example,it would then seem to follow th at t he la st m urde rer in prison might not be executed uponthe dissolution of his island community if the community's preservation is no longer atissue: When its preservation may collapse, alternative penal remedies are justified tocontinue the community.


15/20

Kant sTheoryof Punishment 219IV. THE PROBLEM OF INTENTIONALITY

While they are subjecttothe p articularlawsof theirplaceof residence,Kan t believes that hum an beings ought to follow th e mo ral law at alltimes.63The moral law is not some Utopian ideal we might aspire to,yet never fulfil. Instead, the significance given to the moral law viareason is that 'the moral law is solely practical', its realization byhuman beings is a true possibility and not a product of wishfulthinking.64It is the moral law which determines the concept of rightand wrong, as a stand ard by whichwefine-tuneegislation to unite asclosely as possible juridical law with the moral law.65Rather confusingly, Kant suggests that when we violate the morallaw our transgression deserves punishment.66In this way, Ka nt speaksof punishment(poena) resulting from the m oral law (the categoricalimperative)'s violation asthe rightfuleffect .61Specifically, the 'rightfuleffect' is 'what is culpable is punishment'.68 This matter might beunderstood in one oftwoways. One interpretation is tha t persons whoviolate the moral law ought to be punished for their violation -punishment is the 'rightful effect' - but these persons may avoid beingpunished because they have not violated the juridical law. The diffi-culty with this view is tha t it m akes Kant's proposition tha t criminalsshould be punished on the basis ofmens reaappear contradictory.69

63 See Kant,Critique of Practical Reason,pp. 35 f. [5:38],70 f. [5:82-4] and 79[5:93];Kant,Metaphysics of Morals,p . 9 [6:215].64 Kant, Critique of Practical Reason, p. 44 [5:50]. See ibid., pp. 33, 74 [5:37,86-7];Kant, Metaphysics of Morals, pp. 42 f. [6:252-3]; Ka nt, 'On the Proverb: Th at M ay beTrue in Theory, But Is ofNoPractical Use',PerpetualPeaceand Other Essayson Politics,History, and Moral Practice,ed. T. Humphrey, Indiana polis, 1983, p. 71 .65 See Kant,C ritique of Practical Reason ,p . 54[5:63];Kant ,Lectures on Ethics, t rans .He ath , pp. 73 f. [27:279]( DeLittera Le gis ),80 [27:287] ('Of Rew ards and Punis hme nts'),83 [27:291] ('Of Degrees of Imputa tion'); and Ka nt,Metaphysics of Morals, p. 16 [6:223f.].(All citations from K ant'sLecturesa re from the notes of G. L. Collins.) See also Ka nt'sdiscussion of public right at K ant, Metaphysics of Morals,p. 89 [6:311]; 'On a SupposedRight to Lie because of Philanthropic Concerns', p. 165 [429]. In Kant , EthicalPhilosophy, 2nd edn., ed. and trans. J. W. Ellington, Indianapolis, 1994: in the move'from a metaphysics of right (which abstracts from all empirical determinations) to aprinciple of politics (which applies the se [metaphysical] concepts [of right] to instance sprovided by experience' we 'gain the solution of a problem of politics in accordance w iththe u niversa l principle of right'. (Brackets supplied by Ellington.)66 Kant ,Critique of Practical Reason,p . 34 [5:37].In 'Of Rewards and P unishmen ts' ,Kant says: 'Punishment in general is the physical evil visited upon a person for moralevil'. (See Ka nt,Lectures onEthics,t ra ns . Heath ,p.79 [27:286]. See pp. 308[43 27:552]('In pun ishm ents, a physical evil is coupled to moral badness'), 309 [44 27:553].)67 Kant ,Metaphysics of Morals,p . 19 [6:227] (emph asis given). See Kant,Lectures onEthics,trans. Heath, pp. 76 [27:282], 80 f. [27:288], and 306[40 27:549].68 Ibid.69As an example, see ibid., pp. 106 f. [6:333].


16/20

22 0 Thorn BrooksOn the other hand, I believe that this problem is best understoodas an argument for the importance of intentionality for Kant in theattribution of punishments. This is a new and crucial problem incoming to grips with Kant's theory of punishment. Kant insists: 'thestate of mind of the subject, whether he comm itted the deed in a stateof agitation or with cool deliberation, makes a difference in imputa-

tion, which has results'.70 This state of mind - the person's will - iswhat 'always takes first place in estimating the total worth of ouractions and constitutes the condition of all the rest'.71W hatever actionwe perform is granted some level of moral worth from the com-mensurability this action has with the maxim employed whenjustifying the particu lar action. The purpose, or consequence, pursuedis irrelevant for consideration.72When we legally punish, we ought to appraise not only the crimecommitted, but th e 'innerwickedness'of the criminal when performingan offence.73 While juridical laws bear on the external actions ofpersons independent of the reasons for acting - as moral laws areconcerned with the state of mind and decision-making process ofhuman beingsa person's intentional (i.e. culpable) disposition when

performing actions which violate the positive law is an importantfactor in attributing punishable guilt.74Individuals who deliberatelybreach these laws a re more culpable for their criminal action(s) tha nthose who do so unintentionally.75Of course, we should not always assume that a criminal's inten-tionality is united in his criminal action(s), for only in certaininstances are the two fully connected.76 On some occasions, actionsmay become criminal due to unintended consequences. This gapbetween the perceived (our physical movements) and the unperceived(the st ate of our moral disposition) marks th e difficulty of this projectof calibrating punishments to persons solely regarding their moralguilt.77Itisprecisely this gap between thetwowhich needsto bebetter

70 Kant ,Metaphysics of Morals,p . 20 [6:228].71 Kant ,Groundwork,p . 10 [4:397].72 Ibid., p. 13 [4:399 fj. See ibid., p. 14 [4:401].73 Kant ,Metaphysics of Morals,pp. 106 f. [6:333].74 See Kant,Lectures on Ethics, tra ns . Hea th, p. 74 [27:280] in the section DeLitteraLegis .75 I am referring to the common legal distinctions ofactus reusandmens reaand theiroften critical importance in attributing punishments to criminals.76 Kant , Metaphysics of Morals, p. 108 [6:335]. See Kant, Groundwork, pp. 13[4:399 ] and 19 f. [4:407].77 It should also be kept in mind th at t he calibration of juridica l law towards encom-passing m oral law is not a project to be completed overnight. This process may well takegenerations. (See Kant, 'Idea for a Universal History with a Cosmopolitan Intent', Kant,Perpetual Peaceand Other Essays,Akademie p. 19.)


17/20

Kant s Theory of Punishment 221understood if we are to impart criminal guilt and/or penalties tocriminals due, at least in part, to their inten tional disposition.78In the Critique of Pure Reason, Kant says:The real m orality of actions, their merit or guilt, even th at of ourownconduct,thus remains entirely hidden from us. O ur imputations can refer only to theempirical character.Howmuch of this characterisascribabletothe pure effectoffreedom,how much to mere nature, that is, to faults of temperament forwhich there is no responsibility, or to its happy con stitution(meritofortunae),can neverbe determined, and upon it therefore noperfectlyjust judgementscan be passed.79He states elsewhere: ' In fact, it is absolutely impossible by means ofexperience to mak e out with complete certainty a single case in whichthe maxim of an action otherwise in conformity with duty restedsimply on moral grounds an d on the rep resen tation of one's duty5.80Morality is poised at the very boundary of human knowledge:'reason would overstep all its bounds if it took upon itself to explainhow reason can be practical' . 81What Kant fails to sufficiently consideris the fact that the moral law's true content - which we have noknowledge of - might actually vary greatly with his theory of morallaw.While Kant denies that we could ever know for sure whether ornot a person was morally motivated, the moral law can only haveauthority over our wills when it is possible for us to be motivated byit:82 for 'what counts is not actions, which one sees, but those innerprinciples of actions that one does not see'. 88Simply put, the essential

78 As Kantsays,'Right [justice] must nev er be adapte d to politics; rathe r, politics m ustalways be adapted to right' (Kant, Ethical Philosophy,p. 166 [429]). See Kant,Lectureson Ethics,t ran s. Hea th, p. 306 [40 27:550]. The fact that m ost interpre tation s of Ka nt'stheory of punishm ent havenot found such a gap previously ought then to be seen as oneinstance in which many w ere misled.79 Kant ,C ritique of Pure Reason,tr an s. N. K. Smith, p. 475n [A552/B580] (emphasisadded). See Sullivan, Immanuel Kant s Moral Theory,p. 243.80 Kant , Groundwork, p. 19 [4:407]. Due to our inability of knowing dispositions withany certainty, T. H. G reen criticizes Kant for continuing to include intentiona lity in h istheory of punishment. Green does what he believes Kant ought to have done: base histheory of punishment on more consequentialist grounds. (See Thomas Hill Green,Lectures on the Principles of Political Obligation, London, [1882] 194 1, pp. 18 0-205[sections 176-206] and my 'T. H. Green's Theory of Punishment', History of PoliticalThought (forthcoming).)81 Kant, Groundwork,p. 62 [4:458-9]. Kant says: We see this as soon as we becomeconvinced that th ere is a use of pure reas on which is practical an d absolutely nece ssary(viz., its moral use). When used practically, pure rea son inevitably expan ds and reache sbeyond the bounds of sensibility* (Kant, Critique of Pure Reason, trans. Pluhar, p. 27[Bxxv];see Katrin Flikschuh,Kant and M odern Political Philosophy,Cambridge, 2000,p.194).82 I paraphrase Korsgaard, 'Introduction', p. xxi.83 Kant , Groundwork, p . 20 [4:407]. Neve rtheless, this state of affairs does not stopKan t from sta ting tha t: 'the pure thoug ht of duty and in gene ral of the m oral law, mixed


18/20

222 Thorn Brooksgood in an action 'consists in the disposition'.84 Accordingly, judgeswill have little recourse other than to 'construe the criminal's act asintending] to express a universal law even though it probably infact results from an irrational maxim'.85Ifweare to attribute penalsanctions on the basis of criminal inte nt - never sure if we are correct- then it may well be tha t K ant's theory of punishm ent would justifythe legal pun ishm ent (and possible execution) of innocent persons.

86Kant admits, 'only the moral relations of human beings to humanbeings are comprehensible to u s' as an empirical phenomenon.87Thisstatement underlies the dilemma at hand. At the very least, neitherKant nor many of ourselves would be satisfied with a system ofjuridica l law which lacked coherence and broad no tions of justice, inparticular.Yet,th e shape which coherence and no tions of justice oughtto take are q uite difficult to ascertain. My reading und erstands Kantto agree with this view and further argue tha tweare able to work outideal notions of justice via theo retical tools, such as the categoricalimperative. The problem is how best to incorporate this ideality w iththe empirical world in which we live. It is beyond the scope of thisarticle to identify an acceptable account of intentionality; suffice to say

thatwhileKantdoesnot ru le out the impossibility of such a project, heis aware of its difficulty. This has not lost its strength since Kant's

with no foreign addition of empirical inducements, has by way of reason alone aninfluence on the human heart so much more powerful than all other incentives, whichmay be summoned from the empirical field' (ibid., pp. 22 f. [4:410-11]).84 Ibid., p. 27 [4:416].85 See Fleischacker, 202. Fleischacker believes that the Kantian judge 'appears to bea rather strange human being, and indeed it is not clear that he is, in his formal role, ahum an being at a ll' (ibid., p. 203). Also see where K ant says: 'Rewards an d punish me ntsare merely subjective motivating grounds; if objective grounds no longer avail, thesubjective serve merely to replace the w ant of morality (Kant, Lectures on E thics, t rans .Heath, p. 80 [27:287]; see p. 284 [ 24 27:522]; Manfred Kuehn, Kant: A Biography,Cambridge,2001,p .41 ;and George, p. 149n39).86 This is not to say tha t Ka nt would find it desirable to punish innocent persons, for,in fact, the moral law prohibits this from happening. I would argue that the attempt toadminister punishments only to guilty persons is one example of how consequentialistpunishments are to be tempered by dictates of rational morality to the best degreepossible (and for good reason). However, the inability to always (or mostly) positivelyidentify one's moral disposition when breaching juridical laws plausibly opens th e doorwide to mistaking unintended consequences for criminal culpability. Thus, the Kantianbelieves the innocent man being punished is innocent, whereas the 'naive utilitarian'knows the person to be punished is innocent and punishes him anyway. (See my'Gilligan on Deterrence and the Death Penalty: Has Legal Punishment Failed Us?',Ethics and Justice, iii/iv (2001/2002); my Utilitarianism, Cap ital Punishm ent, a ndInnocent Persons: A Defense of Bentham', Review Journal of Philosophy and SocialScience,special issue, xxvii (2002); and F red Rosen, 'Utilitarianism and th e P unish me ntof the Innocent: The Origins of a False Doctrine', Utilitas, ix (1997). On 'naive utili-tarianism', see my 'Corlett on Kant, Hegel, and Retribution', pp. 578-80.)87 Kant ,Metaphysics of Morals,p . 232 [6:491].


19/20

Kant sTheoryof Punishment 223time, as it would appear th at jus t abou t every project seeking to unifymoral law and positive law faces jus t this dilemm a.

V. CONCLUSIONInterpretations of Kant's theory of punishment have been hamperedby the inability of commentators to distinguish properly betweenKant's distinction between juridical and moral law. Kant is notretributivist at a primary level and then open to considerations ofsocial utility at a secondary level, affixing specific punishments toparticular criminals. Instead, when Kant is arguing on behalf ofretrib utiv e or deterrent-b ased justifications he does so from differentstandpoints. The retributive Kant is concerned with transgressionsof the moral law, the consequentialist Kant with juridical law. Indeed,it is easy to confuse the two as Kant does seem to take into accountmoral culpability when determining the severity of punishment forcriminals.Some commentators, such as Kevin Thompson, are wary of anyconflation of moral and positive law: 'the very sphe res K ant wished toseparate'.88On the contrary, I believe th at Kant is 'searching for a wayto unite the moral law with empirical institutio ns of justice'.89Kant'spenal consequentialism istemperedbyits aspir ing des ire to conform tothe moral law, a project severely curtailed by our inability to knowwith certain ty the actual content of morality aside from a reasonableestimation.90This is particularly troublesome as we are supposed toimpart punishments to criminals commensurable with their criminalculpability. To compound this problem, by advocating in advance themoral measure to which potential criminals will be held to prior toactual ev ents, Kant's theory of punishm ent m ay well impugn penaltiesto undeserving persons.

Rathe r tha n th is being a weakness of Kant's theory of punishm ent,Kant demonstrates instead the weakness of all such theories wherecriminal culpability is central to the determ ination of justice. Ourinability of knowing with certainty a person's disposition whencommitting an act is a reality, but not a h indrance. Wedecide criminalguilt according to stan dard s such as 'beyond a reasonable doubt'. Thelink between intentions, actions, and criminal culpability remains in

88 Kevin Thompson, 'Kant's Transcendental Deduction of Political Authority*,Kant-Studien, xcii (2001), 78.89 Lindstedt, 133.90 I am therefore at odds with Hare's assertion that Kant wa s a sort of utilitarian,namely a rational-will utilitarian', although I am generally sympathetic with Hare'sanalysis (Hare, 4).


20/20

22 4 Thorn Brooksmodern jurisprudence despite our lack of certainty in identifyingdispositions with absolute certainty. Thus, Kant's problem remainsa contemporary dilemma and not an issue of purely [email protected]

91 Presen ted a t the Fourth Europ ean Congress for Analytic Philosophy at Lun dUniversity, the Sc ottish Postgrad uate Philosophy Association at Stirling University, andthe senior postgraduate seminar at the University of Sheffield. I am most gratefulto Gustav Arrhenius, Albert Atkin, Jes Bjarup, Meagan Brooks, Andy Clark, MartinGolding, Gerry Hough, D avid Liggins, Brian O'Connor, an d, especially, Fabian Freyen-hagen, Bob Stern, and Leif Wenar for very helpful comm ents on earlier drafts.

Documents

10792 Kants theory of punishment.pdf