German Idealism and the Concept of Punishment Modern European Philosophy

http://www.cambridge.org/9780521886840

This page intentionally left blank

GERMAN IDEALISM AND THECONCEPT OF PUNISHMENT

Against the background of early modernism – a period that justifiedpunishment by general deterrence – Kant is usually thought torepresent a radical turn toward retributivism. For Kant, and later forFichte and Hegel, a just punishment respects the humanity inherentin the criminal, and serves no external ends: it is instituted only becausethe criminal deserves it. In this original study, Jean-Christophe Merleuses close analysis of texts to show that these philosophers did not infact hold a retributivist position, or even a mixed position; insteadhe traces in their work the gradual emergence of views in favor ofdeterrence and resocialization. He also examines Nietzsche’s view thatmorality rests on the rejection of retribution. His final chapter offers achallenge to the retributivist position, and a defense of resocialization,in the context of current legal theory and practice concerning thepunishment of crimes against humanity.

jean-christophe merle is a senior researcher in philosophy at theUniversity of Tours, an Honorary Professor at the University of Saarland,and a lecturer at the University of Tubingen.

MODERN EUROPEAN PHILOSOPHY

General Editor

ROBERT B. PIPPIN, University of Chicago

Advisory Board

GARY GUTTING, University of Notre DameROLF-PETER HORSTMANN, Humboldt University, Berlin

Some recent titles

Daniel W. Conway: Nietzsche’s Dangerous GameJohn P. McCormick: Carl Schmitt’s Critique of LiberalismFrederick A. Olafson: Heidegger and the Ground of Ethics

Gunter Zoller: Fichte’s Transcendental PhilosophyWarren Breckman: Marx, the Young Hegelians, and the

Origins of Radical Social Theory

William Blattner: Heidegger’s Temporal Idealism

Charles Griswold: Adam Smith and the Virtues of Enlightenment

Gary Gutting: Pragmatic Liberalism and the Critique of Modernity

Allen Wood: Kant’s Ethical ThoughtKarl Ameriks: Kant and the Fate of Autonomy

Alfredo Ferrarin: Hegel and Aristotle

Cristina Lafont: Heidegger, Language, and World-Disclosure

Nicholas Wolsterstorff: Thomas Reid and the Story of Epistemology

Daniel Dahlstrom: Heidegger’s Concept of Truth

Michelle Grier: Kant’s Doctrine of Transcendental IllusionHenry Allison: Kant’s Theory of Taste

Allen Speight: Hegel, Literature, and the Problem of Agency

J. M. Bernstein: AdornoWill Dudley: Hegel, Nietzsche, and Philosophy

Taylor Carman: Heidegger’s Analytic

Douglas Moggach: The Philosophy and Politics of Bruno Bauer

Rudiger Bubner: The Innovations of IdealismJon Stewart: Kierkegaard’s Relations to Hegel Reconsidered

Michael Quante: Hegel’s Concept of Action

Wolfgang Detel: Foucault and Classical Antiquity

Robert M. Wallace: Hegel’s Philosophy of Reality, Freedom, and God

Johanna Oksala: Foucault on Freedom

Beatrice Longuenesse: Kant on the Human Standpoint

Wayne Martin: Theories of JudgmentHeinrich Meier: Leo Strauss and the Theologico-Political Problem

Otfried Hoffe: Kant’s Cosmopolitan Theory of the Law and Peace

Beatrice Longuenesse: Hegel’s Critique of Metaphysics

Rachel Zuckert: Kant on Beauty and Biology

Andrew Bowie: Music, Philosophy and Modernity

Paul Redding: Analytic Philosophy and the Return of Hegelian Thought

Kristin Gjesdal: Gadamer and the Legacy of German Idealism

GERMAN IDEALISM AND

THE CONCEPT OF

PUNISHMENT

JEAN-CHRISTOPHE MERLE

translated from the German by

JOSEPH J. KOMINKIEWICZ

with

JEAN-CHRISTOPHE MERLE

and

FRANCES BROWN

CAMBRIDGE UNIVERSITY PRESS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore,

São Paulo, Delhi, Dubai, Tokyo

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

First published in print format

ISBN-13 978-0-521-88684-0

© Jean-Christophe Merle 2009

2009

Information on this title: www.cambridge.org/9780521886840

This publication is in copyright. Subject to statutory exception and to the

provision of relevant collective licensing agreements, no reproduction of any part

may take place without the written permission of Cambridge University Press.

Cambridge University Press has no responsibility for the persistence or accuracy

of urls for external or third-party internet websites referred to in this publication,

and does not guarantee that any content on such websites is, or will remain,

accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

Hardback

http://www.cambridge.org/9780521886840

http://www.cambridge.org

. . . we all know today’s executioners are humanists.

Albert Camus

CONTENTS

Preface page xi

List of abbreviations xiii

Introduction 1

part i desert as the sole justificationfor punishment

1 The two Kantian concepts of right 17

2 Kant’s legal justification of punishment 44

3 Kant’s moral justification of punishment 72

part ii punishment as a means ofrehabilitation

4 Fichte’s “expiation contract” 87

5 Hegel’s “negation of crime” 107

part iii retributivist inhumanity

6 Nietzsche and punishment without remorse 149

7 What is the purpose of punishing crimesagainst humanity? 171

Conclusion 187

Bibliography 197

Index 204

ix

PREFACE

“Nemo prudens punit quia peccatum est sed ne peccetur,” says Senecain De ira, and many philosophers who have come after him recommendsuch a justification of punishment by deterrence. Since ImmanuelKant, a completely different concept has spread among philosophers,considerably more so than among legal scholars and lawyers.According to Kant, the question of justification of punishment shouldnot read: For what purpose punish? Rather, according to Kant’sabsolutist or categorical imperative regarding punishment, punish-ment can only be carried out because the malefactor is deserving ofthe punishment. Everything else is allegedly unjust, and is detrimentalto the malefactor’s human dignity as a moral subject. Such a theory ofretributive justice, which draws not only from Kant but also fromG. W. F. Hegel, inspires a great deal of fascination in many philoso-phers, but that notwithstanding it still stands on shaky ground.A precise analysis of Kant’s and Hegel’s philosophy of law and moral-ity leads rather to a special form of deterrence theory.

I will attempt to conduct this analysis within the confines of thisbook. The analysis begins with Kant, continues with J. G. Fichte andHegel, leads to Friedrich Nietzsche, and then concludes with a discus-sion of the justification of punishment for crimes against humanity.This closing discussion should be seen as the touchstone. Should myposition be able to explain this difficult case, then it should be evenmore able to explain cases of lesser difficulty.

I would like to thank Manfred Frank and Anton Schindling fortheir comments on the manuscript as well as two anonymous refereesfrom Cambridge University Press. Special thanks go to Sharon Byrd,PhilippeCoppens, RomanEisele, George Fletcher, ThomasGrundmann,Jan C. Joerden, Matthias Kaufmann, John Kleinig, Eugenio Pacelli de

xi

Oliveira, Herve Pourtois, Alexandre Travessoni Gomes, Luiz Moreiraand Jean-Claude Wolf. I have also benefited from questions andremarks from: European Network “Applied Global Justice,” Kant-Gesellschaft, Fichte-Gesellschaft, CAPPE (ANU Canberra), ChaireHoover and Centre de Philosophie du Droit (Louvain), University ofGrenoble, PUCRS Porto Alegre, UFMG Belo Horizonte, UFSC Flor-ianopolis, Ankara Bar Association, University of Fribourg, Graduier-tenkolleg Globale Herausforderung (Tubingen), Charles SturtUniversity at Wagga Wagga, and a conference of the European Net-work “Global Justice” in Otzenhausen, as well as from my seminars onthese topics in Aachen, Saarbrucken, Tours and Tubingen. Last, butnot least, I am very grateful to Hilary Gaskin of Cambridge UniversityPress, as well as to Frances Brown.

xii preface

ABBREVIATIONS

Kant

GMS Groundwork of the metaphysics of morals(Grundlegung zur Metaphysik der Sitten)(1785, Ak iv:385–464)Immanuel Kant, Practical philosophy, ed. Mary Gregor(Cambridge: Cambridge University Press, 1996), pp. 37–108

Idee Idea for a universal history with a cosmopolitan purpose(Idee zu einer allgemeinen Geschichte in weltburgerlicher Absicht)(1784, Ak viii:15–32)Immanuel Kant, Political writings, ed. Hans Reiss, trans.H. B. Nisbet, second edition, (Cambridge: CambridgeUniversity Press), pp. 41–53

KpV Critique of practical reason(Kritik der praktischen Vernunft)(1788, Ak v:1–164)Immanuel Kant, Practical philosophy, ed. Mary Gregor(Cambridge: Cambridge University Press, 1996), pp. 133–272

KrV Critique of pure reason(Kritik der reinen Vernunft)(1st edn 1781, 2nd edn 1787, Ak iii:1–552)Page numbers are from the second editionImmanuel Kant, Critique of pure reason, ed. and trans.Paul Guyer and Alan W. Wood (Cambridge: CambridgeUniversity Press, 1997)

Pad Lecture On pedagogy(Padagogik)(1803, Ak ix:437–99) (no translation)

Rel Religion within the boundaries of mere reason

xiii

(Religion innerhalb der Grenzen der bloben Vernunft)(1793, Ak vi:1–202)Immanuel Kant, Religion within the boundaries of mere reason,in Kant, Religion within the boundaries of mere reason andother writings, ed. Allen Wood and George di Giovanni(Cambridge: Cambridge University Press, 1998), pp. 31–192

RL The doctrine of right (Part 1 of The metaphysics of morals)(1st edn 1797, 2nd edn 1798, Ak vi:203–372)Immanuel Kant, Practical philosophy, ed. Mary Gregor(Cambridge:CambridgeUniversityPress,1996), pp.363–506

TL The doctrine of virtue (Part 2 of The metaphysics of morals)(1st edn 1797, 2nd edn 1798, Ak vi:373–493)Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cam-bridge: Cambridge University Press, 1996), pp. 507–615

VE Immanuel Kant, Lectures on ethics, ed. Peter Heath and J. B.Schneewind, trans. Peter Heath (Cambridge: CambridgeUniversity Press, 1997)(Eine Vorlesung Kants uber Ethik)(c. 1875–80, Ak xxvii:286)

ZeF Toward perpetual peace(Zum ewigen Frieden)(1795, Ak vii:341–86)Immanuel Kant, Practical philosophy, ed. Mary Gregor(Cambridge: Cambridge University Press, 1996), pp. 311–52

Fichte

GNR Foundations of natural right, according to the principles of theWissenschaftslehre(Grundlage des Naturrechts nach Principien der Wissenschaftslehre)Johann Gottlieb Fichte, Foundations of natural right: Grundlagedes Naturrechts nach Principien der Wissenschaftslehre, ed.Frederick Neuhouser, trans. Michael Baur (Cambridge:Cambridge University Press, 2000)

Hegel

GPhR Elements of the philosophy of right(Grundlinien der Philosophie des Rechts)G. W. F. Hegel, Elements of the philosophy of right, trans.H. B. Nisbet (Cambridge: Cambridge University Press, 1991)

NRSW Lecture on Natural law and the science of state(Vorlesung uberNaturrecht und Staatswissenschaft)(1818–19)(No translation)

xiv list of abbreviations

PhR Lecture on The philosophy of right(Vorlesung uber Philosophie des Rechts (1824–5))(No translation)

Nietzsche

GdM On the genealogy of morality(Zur Genealogie der Moral)Friedrich Nietzsche, On the genealogy of morality, ed. KeithAnsell-Pearson, trans. Carol Diethe (Cambridge: CambridgeUniversity Press, 2007), pp. 1–128

WuL On truth and lies in a nonmoral sense(Uber Wahrheit und Luge)FriedrichNietzsche,Writings from the early notebooks, ed. LadislausLob, Raymond Geuss and Alexander Nehamas (Cambridge:Cambridge University Press, forthcoming)

Note on translations of primary and secondary literature

Every effort has been made to find published English translations ofall foreign-language texts. Where there is no published translation,German passages have been translated for the purposes of this book.The reader should assume that if a quotation is from a German workfor which no English-language citation is given, then the text has beennewly translated. In the interests of simplifying the footnote citations,this will not always be noted unless there is a specific need forclarification.

list of abbreviations xv

INTRODUCTION

A nearly total unanimity prevails with regard to the fundamentalnecessity of a public penal system.1 Even among those few who advo-cate the abolition of all punishments, a large majority advocatesinstituting alternatives to the usual prison sentence, rather thancalling for the abolition of punishment without anything to replaceit. When seen in this light, the existence of public penal law can beregarded as being completely justified. The manner in which punish-ment might actually be justified, however, remains just as controversiala subject as determining the appropriate amount of punishment. Thisis because these issues are closely related to one another.

Every theory of punishment currently advocated shares the rejec-tion of the system of punishment which was prevalent in the earlymodern age. This rejected system, illustrated by such penal provisionsas the Constitutio criminalis Carolina, enacted in 1532, was placed inopposition to the modern system of punishment by Michel Foucaultin Discipline and punish. The early modern system differs from themodern system in the sense that the latter prefers either prisonsentences or (if any) the most painless and most decent death sen-tences possible.2 It is worth noting that well into the eighteenthcentury more than one hundred crimes were capital offenses. Torture,

1. For an example of the few exceptions, see Herman Bianchi, “Abolition: assensus andsanctuary,” in Alexander R. Duff and David Garland (eds.), A reader on punishment(Oxford: Oxford University Press, 1994), pp. 336–51.

2. See the beginning of Michel Foucault, Discipline and punish: the birth of the prison, trans.Alan Sheridan, second edition (New York: Random House, 1995); see also theConstitutio criminalis Carolina in Friedrich-Christian Schroeder (ed.), Die Carolina:die Peinliche Gerichtsordnung Kaiser Karls V. von 1532 (Darmstadt: WissenschaftlicheBuchgesellschaft, 1986).

1

which was already systematically employed as an interrogation method,was often a component of the punishment, as well as constituting anintensification of the death sentence. Even though torture as a meansfor investigation and security is currently being propagated again3 andeven though the death penalty is still supported,4 there are no theoriststo be found who would come out in support of a return to earlymodern practice. All contemporary theorists show themselves to beguided by the humanitarian spirit of the Eighth Amendment to theUnited States Constitution: “Excessive bail shall not be required, norexcessive fines imposed, nor cruel and unusual punishments inflicted.”

What exactly this humanism consists of, and where the limits of thishumanism might lie, are still controversial issues. However, the funda-mental discrepancy is generally seen to be somewhere between thetheories of retributive justice and general deterrence.

Retributivist theories justify the punishment of a criminal on thegrounds that retribution is demanded by justice to compensate for theinequities created by the crime. On this basis, one may make a subsid-iary distinction and ask whether what needs to be compensated for isthe gravity of the offense itself or the malevolence of the criminal thatwas illustrated by the said offense. Theories of general deterrence, onthe other hand, justify the punishment on the grounds that all of thecitizens will be deterred from carrying out an offense before it occurs,either through the threat of a certain punishment or through theenforcement of the said punishment; the latter option relies on theexample it displays.

On the one hand, the modern advocates of general deterrence(examples include Thomas Hobbes, Samuel von Pufendorf, ChristianWolff, Cesare Beccaria, Anselm Feuerbach and Arthur Schopen-hauer) consider that the uselessness inherent in punishments thatdo not serve to deter others from committing crimes is inhumane.The current system of positive criminal law also requires that eachpunishment contains elements of general prevention as stated in theinitial paragraphs of the German Penal Code:

3. See, for example, Winfried Brugger, “Darf der Staat ausnahmsweise foltern?,” Der Staat,35 (1996), 67–97; and “Vom unbedingten Verbot der Folter zum bedingten Recht aufFolter?” Juristenzeitung, 35, no. 4 (18 February 2000), 165–73. For a detailedrepudiation of torture, see, for example, Henry Shue, “Torture,” Philosophy and PublicAffairs, 7, no. 2 (1977–88), 124–43.

4. See Ernest van den Haag, “Why capital punishment?,” Albany Law Review, 54, nos. 3–4(1990), 501–14.

2 german idealism and the concept of punishment

By serving a prison term, a prisoner should eventually become able tolead a life in which he or she commits no crimes (which is the goal of theexecution of punishment). The completion of a term of imprisonmentalso serves to protect the general public from further crimes.5

On the other hand, theories of general deterrence often draw thecriticism that they treat criminals inhumanely, because the intendedaim of punishment is conceived of solely to serve the interests ofthe other citizens without taking into consideration the dignity ofthe criminal. This criticism is leveled in its most intense form whenit is postulated that general deterrence allows the punishment of aninnocent person.6

This objection can be understood in at least two different ways.General deterrence can be objected to on the one hand, for the reasonthat the aim of punishment ignores the interests of the convicted, or onthe other hand, in accordance with retributivism, for the reason thatany kind of interest – whether of the criminal or of the fellow citizens –should be disregarded by the penal sentence, because a punishmentjustified in a retributivist way is merely about inflicting on the criminalwhat he or she intrinsically deserves because of the deed. A punishmentsituated in the retributive model should be concerned with inflicting apunishment that is in line with what the prisoner merited because acertain crime was perpetrated. The latter objection is raised by retribu-tivism. The former objection is raised by positions that hold that therehabilitation of the perpetrator should be the punishment’s goal.Admittedly, the latter position recognizes that for the purpose ofreaching rehabilitation a certain period of time of specific deterrencemay become necessary, in which society is protected from furthercrimes through incapacitation of the criminal.

5. From the German National Code of Enforcement of Sentences, the “Strafvollzugsgesetz”(StVollzG), published by the German Federal Ministry of Justice, } 2. The above passageis the precept to the later statutes in the codex. Compare as well with the judgmentof the German Federal Court of Justice (Bundesgerichtshof) on December 8, 1970, 1 StR353/70. (Translation mine.)

6. Cf. Peter Koller, “Probleme der utilitaristischen Strafrechtfertigung,” Zeitschrift fur dieGesamte Strafrechtswissenschaft, 91 (1979), 45–95; Kristian Kuhl, Die Bedeutung derRechtsphilosophie fur das Strafrecht (Baden-Baden: Nomos, 2001), p. 29; Peter Landau,“Karl Christian Friedrich Krauses Rechtsphilosophie,” in Klaus-Michael Kodalle (ed.),Karl Christian Friedrich Krause (1781–1832): Studien zu seiner Philosophie und zumKrausismo (Hamburg: F. Meiner, 1985), pp. 80–92 (p. 29). For a refutation of thisobjection, compare Fred Rosen, “Utilitarianism and the punishment of the innocent:the origins of a false doctrine,” Utilitas, 9, no. 1 (March 1997), 23–37.

introduction 3

Retributivism criticizes theories of general deterrence as well astheories of rehabilitation considered as the aim of punishment formuch the same reason, that is, because these theories treat punish-ment as a mere means to an end. Retributivists themselves hold thatpunishment ought to be justified without any reference to furthergoals, invoking the rationale that the criminal deserves it because heor she knowingly violated the law. It is for this reason that legaltheorists term retributivism an absolute theory, because, accordingto retributivism, punishment represents a good that does not dependon any goal. On the contrary, in the “relative” theories the justificationof punishment always depends on its relation to a goal. Accordingto the proponents of retributivism,7 it derives its superiority fromthe fact that it alone – as the only theory of criminal justice that viewsthe punishment solely as a goal in itself – treats the malefactor not asa simple means to an end, but as a subject possessing human dignity.In this work, I will attempt to refute these theses. I hope to show thatit is not retributivism but rather rehabilitation that meets thisrequirement.

Proponents of rehabilitation obviously consider the aim of punish-ment to be to grant the criminal the best “possible” status, by whichit is understood that this is the way of treating the criminal that isboth the most benevolent and still compatible with the protection ofsociety against further crimes. Thus, with rehabilitation as an aim inpunishment, clear limits are set for specific deterrence. Without spe-cific deterrence, rehabilitation would be unthinkable, for if there wereno public enforcement of the law, there could in turn be no reinte-gration back into society, for there would be no rule of law into whicha criminal could be reintegrated after the rehabilitation has beencompleted.

Unlike rehabilitative and specific deterrent punishments, retributi-vism does not concern itself with the future of the malefactor beyondthe duration of the punishment. In this respect, the theory of rehabili-tation is the only one that can categorically exclude those sorts ofpunishment that – as mentioned at the beginning of this introduction– the proponents of all the theories of punishment reject resolutely:the “cruel and unusual punishments.” The vice president of theFederal Constitutional Court of Germany Winfried Hassemer rightly

7. See, for example, Otfried Hoffe, Gerechtigkeit: eine philosophische Einfuhrung (Munich:C. H. Beck, 2004), p. 83.


observes: “The goal of rehabilitation is necessarily contained in optingfor prison sentences. Corporal and capital punishments do not needany further justification than retribution.”8

Even when one assumes that the punishment is merited and there-fore justified as being compensation either for the crime or for thecriminal’s own wickedness, and asserts the claim that one should treatthe criminal humanely, this allegedly humane view of the criminalshould also concern itself with the situation in which the criminal willbe after the complete term of imprisonment has been served. Theabsence of appropriate rehabilitative measures leads not only directlyto the creation of a durable criminal environment with an ensuingreduction in public safety,9 but also to a state in which the criminal ispunished twice for a crime, by being continually stigmatized, insteadof enabling him or her to express remorse and to reach a reconcili-ation with society.10 Therefore, concern for the convict’s future afterthe sentence necessitates appropriate treatment while the sentence isbeing served. Retributivism may attempt to fulfill this requirement ofhumaneness, along the lines of what Paul Ric�ur attempted.11 Indoing so, retributivism stumbles upon what Hassemer terms the “anti-nomy of punitive goals,” which refers to the fact that in many cases thevarious existing theories of punishment do not allow for the sameamount of punishment.12 Hassemer observes that

A period of punishment limited by the proportionality principle andas required by the goal of retribution normally does not suffice for atreatment, so that the goal of rehabilitation itself will fail. A period of

punishment can also be too long for a reasonable treatment of the prisoner.13

In view of this antinomy of the punitive goals, a priority rule must beset. Either retributivism should be deemed to be the primary aim andrehabilitation the secondary aim – meaning that the rehabilitation willbe carried out only as far as it does not interfere with the retribution –or the rehabilitation should be given priority over the retribution.

8. Winfried Hassemer, Einfuhrung in die Grundlagen des Strafrechts, second edition(Munich: C. H. Beck, 1990), p. 286.

9. Cf. John Braithwaite, Crime, shame and reintegration (Cambridge: Cambridge UniversityPress, 1989), p. 102.

10. Cf. Braithwaite, Crime, shame and reintegration, p. 101; Hassemer, Grundlagen desStrafrechts, p. 289.

11. Paul Ric�ur, Le Juste (Paris: Editions Esprit, 1995), p. 203.12. Hassemer, Grundlagen des Strafrechts, p. 291.13. Hassemer, Grundlagen des Strafrechts, p. 291.

introduction 5

The form of retributivism that is given priority regards its moralsuperiority as stemming from its being the only genuinely humanejustification for punishment. It sees itself as better than rehabilitationfor the reason that it bases itself not upon concern for the criminal,but instead solely upon his or her merit – or for that matter, upon hisor her guilt, responsibility or malevolence. However, in doing so,retributivism ignores the following points.

First, the following differentiation is necessary. Retribution can beunderstood, on one hand, in its minimalist sense, as meaning that theguilt of the punished, without any exception, should be the prerequis-ite for any given punishment.14 According to this understanding, alltheories of punishment, that is, theories both of general deterrenceand of rehabilitation, are retributivist theories.15 On the other hand,retribution can also be understood to mean (1) a response to theoffense that strives to provide equal compensation for the criminal’smerit or guilt and (2) that this equal compensation is the sole justpunishment. Thus, it excludes any goal in punishment (for instance,general deterrence, rehabilitation and specific deterrence). This iswhat one usually understands under the term retributivism. Retribu-tivism consists in the acceptance of the latter (disputable) thesis.When I speak of retribution in this book, I will be making referenceto the latter understanding of retribution.

Secondly, in the justification of the punishment as retaliation, it isnot the criminal’s future, but rather his or her past that is taken exclusivelyinto account. As we have seen, this occurs with appeal either to thecriminal’smerit or to his or her guilt. The concept that lies at the root ofthis guilt can be from a modern, humane perspective only the conceptof responsibility. In fact, in every contemporary retributive theory – asopposed to the cases of the deterrent and rehabilitative theories – thereis the imperative to respect the responsibility of the criminal as consti-tuting his or her own dignity and to respond to this dignity with equalcompensation.

Throughout this book, I will proceed as would a retributivist,assuming that a human being’s responsibility for his or her actions,as opposed to other living beings, is actually what constitutes thespecial status of human beings (their dignitas). But then, the following

14. Cf. Otfried Hoffe, Gibt es ein interkulturelles Strafrecht? Ein philosophischer Versuch(Frankfurt a.M.: Suhrkamp, 1999), p. 72.

15. Cf. Ulfried Neumann and Ulrich Schroth, Neuere Theorien von Kriminalitat und Strafe(Darmstadt: Wissenschaftliche Buchgesellschaft, 1980), p. 6.


differentiation must be taken into account. There are always twodimensions to the assertion that a human being bears responsibilityfor his or her actions. On the one hand, this responsibility means thatdifferent actions will lead to different consequences, and especiallythat all actions, which from the perspective of morality are valueddifferently, will also lead in some cases to different consequences.Among other things, it means that actions violating the law necessarilycould or should lead to a worse situation. On the other hand, thehuman’s status as a being capable both of reason and of assumingresponsibility is an inalienable status that cannot be taken away fromthat human or any other human. The assertion that the human being’spast illegal actions should carry consequences shouldnot lead us to stoptreating this human being as a being capable of reason, except whenthis prevents his or her fellow human beings from exercising that samestatus. Otherwise, the perpetrator would be treated as a person capableof assuming responsibility only up to the crime; after the conviction, how-ever, the criminal would lose this status, that is, his or her worth. In thisway, retributivism would lack exactly the sort of behavior toward thecriminal which it regards as its moral superiority over the other theoriesof punishment. In short, retributivismdoesnot sufficiently differentiatebetween the actor and the action. In this respect, the viewpoint thatwould trace a retributivist influence back to Christianity proves notto be truly convincing.16 Rehabilitation is equally rooted in theChristian tradition, as the rehabilitation theorist John Braithwaitesuggests in his plea for “reintegrative shaming”:

It is shaming which labels the act as evil while striving to preserve theidentity of the offender as essentially good. It is directed at signifyingevil deeds rather than evil persons in the Christian tradition of “hate thesin and love the sinner.”17

Thirdly, retributivism relies on the assumption that the consequencesof the crime, for which the criminal should take responsibility, shouldconsist of providing equal compensation, because that is what justicedemands. While retributivism focuses in this manner on demandingan equivalent punishment, it misses an indisputable consequence of thecrime : at least temporarily, the existence of a commonwealth betweenthe criminal and other citizens is made impossible by the crime.

16. For this viewpoint, cf. Claus Roxin, Strafrechtliche Grundlagenprobleme (Berlin: De Gruyter,1973), p. 3; and Neumann and Schroth, Neuere Theorien, p. 13.

17. Braithwaite, Crime, shame and reintegration, p. 101.

introduction 7

A short thought experiment may illuminate this point. Let usassume the retributivist position as the starting point for this experi-ment, which is a position that holds that the severity of punishmentshould be equivalent to the gravity of the crime and that this aloneshould represent – without any specific aim of punishment – the onlyjustified solution. Then, let us assume that in this fashion we reach theconclusion that a criminal has merited a twenty-year prison sentence.It is usually the case that serving the prison term begins, at the verylatest, after the judgment has taken effect and after all recourse hasbeen exhausted (in reality, however, the criminal – still consideredofficially innocent – has already been in custody either since theinvestigation commenced or since the issue of the arrest warrant). Inthis case, retributivism should raise no objections if the enforcementof the sentence were to be postponed. If the criminal were, forexample, twenty years old, this would allow him or her to be in prisonfrom age thirty to age fifty, but to remain a free person from twenty tothirty.18 It is not only for pragmatic reasons (for instance, because ofadministrative concerns regarding prison capacity) that our societywould reject such a reform of the prison system; our society would findthis fundamentally and wholly unacceptable because it would severelyendanger public safety. Above all, such a reform would be rejectedbecause of a specific deterrent rationale. Now, the danger to society isclearly consequent to the crime. Following the specific deterrencemodel, the criminal is therefore liable for this consequence causedby his or her criminal offense. Whereas it is controversial whether themoral demand of retribution for the guilt that is addressed by retribu-tion can be seen as a consequence of the crime, the danger to societyis an unquestionable consequence arising from the actions of thecriminal. Retributivism, however, does not actually take into accountthe degree to which society is endangered by these crimes, which is atodds with retributivism’s own self-portrayal of itself as being the onlytheory of criminal justice that requires the criminal to shoulder theburden of the consequences of his or her actions.

Fourthly, the disregard shown for the consequences of the actionattests to a disregard for the legal dimension of crime as well as for thepunishment of a crime. Retributivism focuses on the guilt of theperpetrator. Admittedly, it also emphasizes that the punishment of

18. Such a postponement is actually possible under German law, though admittedly onlyin a limited number of cases with short sentences.


the criminal represents justice for the victim.19 Apart from that, con-cern for the commonwealth does not play a role in retributive theory’sjustification of punishment. Consequently, if in the retributivist justi-fication of punishment only the criminal is taken into account, thisbegs the question of why the punishment falls under the jurisdictionof the judicial powers, that is, why it belongs to the commonwealth –which, in other cases, maintains distance from the private sphere ofthe individual. If, in the retributivist justification of punishment, onlythe criminal and the victim are taken into account, then it is inevitableto ask why the punishment could not just be decided in a civil trial.Under these circumstances, we should not be surprised to observe thatretributivism is hardly ever supported by legal theorists, even though itenjoys a wide esteem among philosophers for its supposed morality, aswell as majority support.20 This discrepancy between the view ofphilosophers and the view of legal theorists does not, unfortunately,receive much attention from philosophers.

Unlike the theories of deterrence and rehabilitation, the justifiedimplementation of retributivism is not accountable to performance cri-teria, or to its “output” (put differently: “According to retributivism . . . thesignificance of punishment lies outside the realm of social reality”).21 Ifespecially high recidivism figures were noticed in relation to one sort ofcrime that was being handled with measures involving rehabilitationof the criminal, there would then be questions as to the legitimacy ofthis sort of penal mechanism. However, the implementation of retri-butivism is, by its very essence, not dependent on its effectiveness. Thejustification for a system of retributive justice is not empirically verifi-able through criminological studies, for example. Rather, the criticismdirected toward retributivism must be at the conceptual level. For this

19. This perspective is very questionable, because, in a modern constitutional state, thecriminal proceedings are differentiated from civil proceedings, among other things,by the former being carried out by the state’s attorney as the representative of theinterests of the commonwealth, while the latter transpires between two private parties.At the most, the victims appear alongside as joint plaintiffs and the punishment is inno way seen to be compensation for the infringement on their rights. Since the notionthat punishing the malefactor represents justice for the victim seems not to be centralto the core of retributivism I will disregard this aspect of it, at least in this work, andconcentrate only on the main argument of retributivism. Were this main argumentomitted, the (alleged) justice on the part of the victim for a proponent ofretributivism would also not be enough justification for the ills that the criminalwould be forced to undergo as a component of his or her punishment.

20. Cf. Roxin, Strafrechtliche Grundlagenprobleme, p. 182.21. Neumann and Schroth, Neuere Theorien, p. 11.

introduction 9

reason, the criticism of retributivism that will be made in this bookwill present hardly any empirical and interdisciplinary aspects.Instead, it will concentrate on the conceptual arguments, that is, onarguments dealing with legal ethics and moral philosophy. In thisrespect, this book will proceed no differently than do the depictionsof those legal theorists who still treat “absolute theory” in relationto Kant’s and Hegel’s argumentation – or more exactly, to its theore-tical foundation. But my conceptual critique will aim to facilitatethe convergence of the philosophical with the legal debate over thejustification of punishment.

Last but not least, the importance of empirical studies for thejustification of punishment should not be overestimated. Hassemercalls to our attention

that reliable knowledge about the successes of rehabilitation can hardlybe obtained. The favorite argument based on the recidivism figures,which because they fluctuate between 30 and 40 percent are said todiscredit the concept of rehabilitation, is untenable upon closerexamination. First, from such statistics one knows only the manifest,determined and judged criminality . . . Secondly, experiments andempirical lines of argument face a fundamental problem in penal law:one cannot isolate the intervening variables; one cannot try out how itwould be if one were to seek to achieve the reform of the offender withanother form of penal consequence.22

Even if an especially high recidivism rate were to be observed in regardto rehabilitative punishments of certain crimes, it would still be im-possible to draw a reliable conclusion from that observation. Is thelimited efficiency of the punishment still better than nothing? Or, onthe contrary, should the punishment of these crimes be completelyabolished? Should a retributivist degree of punishment be substitutedin place of a rehabilitative punishment? Or should an effective generaldeterrence punishment be introduced, such as the death penalty?The decision remains to be made, and it requires overall legal andmoral guidance regarding the justification of punishment. The con-sequences for the commonwealth should also be brought into consid-eration. I differ in this point somewhat from the following view ofKristian Kuhl, for example: “All the theories of punishment that aregeared to certain future goals must assert the suitability of the

22. Hassemer, Einfuhrung, p. 288. Also, cf. George P. Fletcher, Basic concepts of criminal law(Oxford: Oxford University Press, 1998), p. 31.


punishment for reaching these goals, and this assertion really mustbe proven with empirical data.”23 In my view, the debate about justifi-cation of punishment is first and foremost a legal and a moral discus-sion. In this work I will conduct the discussion accordingly.

The four aforementioned points urge skepticism in relation to theclaim of moral superiority that is made by retributivism, becauseretributivism, unlike rehabilitation, does not pay heed to the crim-inal’s interests, but instead only to his or her merits or responsibility forthe consequences of every act. First, equal compensation is neithernecessarily nor self-evidently the consequence of a criminal offense.Secondly, retributivism considers only one dimension of responsibil-ity, but not the inalienable status of a human being as a being capableof reason. Thirdly, retributivism would allow important consequencesof the criminal offense – such as those for the commonwealth – to bedisregarded. Fourthly, an explanation is missing in retributivism forwhy a punishment intended retributively, whose justification onlyconcerns the criminal – and possibly the criminal’s victim as well –does not, however, concern the commonwealth, but yet is stillimposed by the commonwealth.

All of these four objections form the basis of my critical debate withretributivism. Many exponents of it are already cognizant that thesefour points are at least problematic. They respond to skepticism bydispensing with the exclusivity of retributivism in justifying punish-ment and in determining the degree of punishment. The opinion isoften expressed that retributivism only represents one particular eth-ical framework for criminal law in which goals of punishment are alsoallowed to be fully pursued. When this point is made in order to limitthe application of the other theories of punishment through a kind oflimitation of power, there is a presupposition that retributivism enjoyslegal and moral superiority, for it allegedly respects the rights of theindividual human being more than the other theories of punishment.Claus Roxin, for example, states:

If guilt gives the state a right to retribute, or if it is the means by whichthe needs of the many as opposed to individual freedom can be reinedin, this seems to me to be a more important question for criminal lawthan the question about the existence of guilt in general. The answermust be compatible with the second option.24

23. Kuhl, Die Bedeutung der Rechtsphilosophie, p. 30.24. Roxin, Strafrechtliche Grundlagenprobleme, p. 21.

introduction 11

The restriction, which Roxin admits partly contradicts retributivism,proceeds, according to Roxin, by prohibiting a punishment fromexceeding the retributivist degree of punishment, while allowing itto fall short of the retributivist degree.

The punishment is not allowed to exceed the degree of guilt . . . On thecontrary, it is allowed that the punishment befitting the guilt isundercut. Actually, such a practice shows itself to be unacceptable fora consistent retributivist theory, because it means a partial relinquishmentof the suffering in punishment that provides equal compensation.25

The justification for this mixed theory is found by Roxin, as by somany other mixed theorists who have dominated the debate sincethe 1980s, in a form of division of labor between each of the theoriesof punishment. This division of labor proceeds by assuming thefollowing observation:

Each theory of punishment focuses unilaterally on certain aspects ofcriminal law – the specific deterrent theory focuses on the enforcement,the retributive notion on the judgment and the general deterrentconception on the goal of threatening punishment – and neglectsthe remaining guises of the penal power, even though, however, eachone of them implies specific interferences with the freedom of theindividual.26

According to the division of labor in the mixed theories, every theorywill allegedly confine itself to its main focus. However, such a con-struct obviously succumbs to the aforementioned “antinomy of puni-tive goals.” My investigation will therefore select the search for a singleprimary justification of punishment as the guideline. In doing so, it is notout of the question that the demands of the other theories of punish-ment may be partly met; however, if they are met at all, they will onlybe partly met. This is because the other theories of punishment areincompatible with one another. They are not, however, contrary theoriesof punishment.

In what follows, I will attempt not to lose sight of the conceptualargument’s concrete meaning and its concrete consequences. How-ever, arguing from the point of view of legal ethics, my critiquewill, whenever necessary, allow itself to diverge from positive law, asretributivism also does. In this respect, I will not, for example, accept

25. Roxin, Strafrechtliche Grundlagenprobleme, pp. 23–4.26. Roxin, Strafrechtliche Grundlagenprobleme, p. 12.


out of hand the criticism against rehabilitation theory that it allowsneither a minimum nor a maximum time-limit for the degree ofpunishment. Roxin regards it as an objection against rehabilitationthat rehabilitation should “consequently” aim for a “treatment untilan ultimate reform . . . even when it is of unforeseeable duration,” and“that every punishment has to be dispensed with, even with the gravestcrimes, if there is no danger of recurrence.”27 It may be true thatmany theorists of rehabilitation28 do not want to draw these conclu-sions, and are, on account of that, inconsistent. This does not mean,however, that rehabilitation as an aim of punishment should be aban-doned, but instead that one should defend it – if one is to defend it atall – in a consistent manner, and for this reason accept and assertthese consequences. I want to do exactly this.

Therefore, in this text, I would like to criticize and refute retribu-tivism in that very aspect which it sees as being the core of its legal andmoral superiority over the other theories of punishment: I will attemptto show that it is not retributivism, but instead solely rehabilitation, that takesthe responsibility truly seriously that a human being – and especially a criminal– has for his or her actions. For the aforementioned reasons, I will relate mycritique first and foremost to Kant’s and Hegel’s arguments.

My critique’s foundation will be the Kantian concept of right thatbuilds directly upon the very human dignity to which retributivismappeals: “Right is therefore the sum of the conditions under which thechoice (Willkur) of one can be united with the choice of another inaccordance with a universal law of freedom.”29

Kant’s concept of right can be interpreted in two ways, whichI present and critically discuss in the first chapter, that is, either as aliberal concept of right, or as a moral concept of right, which is moralin the narrowest sense of the term. According to the liberal interpret-ation, the Kantian concept of right (Rechtsbegriff ) means the coexist-ence of the freedoms of action according to the principle of equalrights, without any consideration of the content of the “choice” (Will-kur), that is, not dependent on how humans want to employ thisfreedom. However, according to the moral interpretation in thenarrow sense, the task falls to the legal system to implement the

27. Roxin, Strafrechtliche Grundlagenprobleme, p. 7. Also, cf. Fletcher, Basic concepts, p. 38.28. For example, John Braithwaite and Philip Pettit, Not just deserts: a republican theory of

criminal justice (Oxford: Clarendon Press, 1990), p. 101.29. RL Ak vi:230. Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge:

Cambridge University Press, 1996), p. 387.

introduction 13

complete categorical imperative to such an extent in the legal systemas the will of a single human being will allow itself to be moved to themorally right through coerced implementation. In the rest of Part i, theattempt will be made to show that Kant’s own retributivist theory isinconsistent with both the first, liberal interpretation of his concept ofright (see Chapter 2) and the second, moral (in the narrower sense)interpretation of it (see Chapter 3). Rather, the theory of rehabilitationis compatible with both interpretations.

In Part ii, I shall proceed from the statement that the Kantianconcept of right comes to be adopted by Fichte as his concept ofright, and by Hegel as the “abstract right.” Out of Kant’s concept ofright, Fichte and Hegel develop similar consequences to those thatKant, in my view, should have logically drawn. Whereas in Fichte thefate of the criminal beingpunished is central (seeChapter4),Hegel, onthe contrary, follows restoration of rights as a guideline and throughthis offers amore systematic presentationof aposition that, in its results,is very similar to the Fichtean theory of punishment (see Chapter 5).The retributivist way of interpretingHegel, which is frequently deemedto be axiomatic, will prove to be unfounded and wrong.

In Part iii, I shall attempt to show that, in opposition to Kant’sclaims to it, retributivism respects the human dignity in the criminal’sperson less than the alternatives of deterrence. I will deal in Chapter 6with Nietzsche’s radical critique of retributivism, according to whichthe original motive for the institution of a retributivist punishment isto be found by no means in a respect for human dignity, but instead ina cruelty to the criminal free of morality, which rather hinders thanencourages the onset of bad conscience and remorse. In contrast tothat, I advocate, in Chapter 5, out of the perspective of a combinationof specific deterrence and rehabilitation, a treatment of criminalsthat respects the dignity in human beings. In order to comply withthe promise that has been made in this introduction to accept all ofthe consequences of the alternative to retributivism that I support –and even, as the case may be, the most radical, the most unusual andthe most unpopular consequences – I shall even use, in the seventhchapter, the example of the gravest and most inhuman crimes, that is,of crimes against humanity.


PART I

DESERT AS THE SOLEJUSTIFICATION FOR PUNISHMENT

1

THE TWO KANTIANCONCEPTS OF RIGHT

1.1. The retributivist turn

The lex talionis is a rule of distribution for the degree of punishmentthat is at least as ancient as the biblical verse “Breach for breach, eyefor eye, tooth for tooth: as he hath caused a blemish in a man, so shallit be done to him again.”1 It never really was a self-evident, absoluteprinciple for all punishments that needed no further justification.Plato, along with several major schools of ancient philosophy, adaptedthat premise formulated by Seneca into the famous sentence “Nemoprudens punit quia peccatum est sed ne peccetur.”2 It contains twodecisive ideas. First, it implies that punishment is an evil.3 If punish-ment is to be allowed at all, this evil must be outweighed by the good itbrings with it. Secondly, punishment is, as any institution, a meanstoward the realization of the bonum commune, which is the good thatmakes all other goods realizable in a sustainable way.

Early modern thought in penal law drew important consequencesfrom these premises. I would like quickly to highlight the commonal-ities of the early modern theorists of penal law, however significanttheir divergences from other aspects of law and penal law may be.

First, punishment is a means to obtain compliance with the lawfrom individuals. This is independent of the question whether thepurpose of the law is primarily self-conservation, or perhaps the

1. Lev. 24:18–20; see also Exod. 21:23–25; Deut. 19:21 (KJV).2. “A sensible person does not punish aman because he has sinned, but in order to keep him

from sin.” Seneca, De ira, i.19.7 in Moral essays, trans. John W. Basore (3 vols., London:Heinemann/Cambridge, Mass.: Harvard University Press, 1928), vol. 1, pp. 106–355(pp. 158–9).

3. See, for instance, Henry Sidgwick, The elements of politics, reprint of 1891 edition(New York: Cosimo Classics, 2005), p. 109.

17

pursuit of happiness (Hobbes), or happiness through the least possiblelimitation of individual freedoms with an emphasis on the less privil-eged (Beccaria), or happiness as the greatest happiness of the greatestnumber (Bentham), or any other form of bonum commune. Indeed, self-conservation – and, in general, protection of citizens – is a condition forpursuing happiness, however one may conceive of the latter. Classicalutilitarian thinkers certainly do belong to this tradition of thought inpenal law: for example Jeremy Bentham, as well as Cesare Beccaria, whosome see as a forerunner to utilitarianism (according to him, the “coolobserver of human nature” recommends the following maxim of publicethics: “the greatest happiness shared among the greatest number”).4 However,this tradition is not limited to utilitarian authors, but also encompassesmost of the authors of the natural law tradition. Thus, I shall sketch themost prevalent elements of early modern thought without addressingthe many differences between natural law and utilitarianism that existin the realm of public ethics, public law and the foundations thereof.

Punishment is considered to be a sanction, meaning that it attachesa consequence to certain behaviors and certain deeds by means of alegal institution in order negatively to influence the motivations thatlead to these deeds. As a sanction, its purpose is to deter a citizen frombreaking the law; thus, punishment must be felt by its addressees to bean evil, and to that end the punishment must be impressive. In this case,the addressees are both the convicted and all other citizens whosepassions may incline them to violate the law5 and to disregard reasonsthat demonstrate that obedience to the law benefits the entire com-monwealth. While Grotius,6 Pufendorf7 and Bentham explicitly men-tion both goals, the latter states that “example is the most importantend of all, in proportion as the number of the persons under tempta-tion to offend is to one.”8 As a means of general deterrence, Bentham

4. Cesare Beccaria, On crimes and punishments, in Beccaria, On crimes and punishments andother writings, ed. Richard Bellamy, trans. Richard Davies (Cambridge: CambridgeUniversity Press, 1995), pp. 1–113 (p. 7).

5. Book ii, Chapter xx, Section xxix(1) in Hugo Grotius, The rights of war and peace, ed.Richard Tuck (3 vols., Indianapolis: Liberty Fund, Inc., 2005), vol. 2, p. 1003.

6. Book ii, Chapter xx, Section vi(1f) in Grotius, The rights of war and peace, vol. 2,pp. 961ff.

7. Book viii, Chapter iii(9) in Samuel von Pufendorf, De jure naturae et gentium, ed. JamesBrown Scott, trans. C. H. and W. A. Oldfather, Classics of International Law (2 vols.,Oxford: Clarendon Press / London: Humphrey Milford, 1934), vol. 2, pp. 1175–6.

8. Footnote in Chapter 13, } 1 of Jeremy Bentham, An introduction to the principles of moralsand legislation, reprint of 1823 edition (Oxford: Clarendon Press, 1907), p. 171.

18 desert as the sole justification for punishment

mentions making an example of a criminal; as a means to specificdeterrence he mentions “reformation” of the criminal’s will as well as“disablement” of his or her physical power.9

Secondly, innocents are not to be punished, for the following tworeasons. To begin with, “there can arrive no good for the Common-Wealth, by Punishing the Innocent.”10 In fact, the punishment, nolonger being linked to the infringement of the law, would not be asanction – meaning a disincentive – any longer. Thus, the punishmentof innocents would contradict the very goal of punishment: deter-rence. Furthermore, “for seeing all Soveraign Power, is originallygiven by the consent of every one of the Subjects, to the end theyshould as long as they are obedient, be protected thereby; the Punish-ment of the Innocent, is a rendring of Evill for Good.”11 Thus, thepunishment of innocents would damage the very reason for establish-ing the commonwealth, the preservation of which is the proper aimof punishment.

Thirdly, as for the degree of punishment, on the one hand, it isnot (according to Pufendorf) “necessary that a man suffer exactlywhat he has done to another, that is, that crimes be always punishedby talion.”12 Hugo Grotius stresses the point that most often penal lawdid not in practice enforce the law of talion, and that this is even thecase with Moses, who most definitely never instituted it in reality. Infact, according to Grotius, the practice of penal law inflicted, rather, adegree of punishment proportionate to the crime, rather than equalto it: most of the time, the true degree of punishment exceeds theseverity of the crime by many multiples.13 On the other hand, theearly modern thinkers make a plea for introducing more differenti-ation into the degree of punishment that is in accordance with theseverity of the crime committed. There is a sort of ordinal proportionalityto punishment of crimes, wherein the severity of each crime is peggedto an ordinal scale of punishment. The purpose of this scale, uponwhich a less severe crime is always punished more mildly than a moresevere crime, is deterrence of potential criminals from committing

9. Bentham, Principles of morals and legislation, p. 170.10. Part 2, Chapter 28 in Thomas Hobbes, Leviathan, ed. A. R. Walter (Cambridge:

Cambridge University Press, 1904), p. 229.11. Part 2, Chapter 28 in Hobbes, Leviathan, p. 229.12. Book viii, Chapter iii (27) in Pufendorf, De jure naturae et gentium, vol. 2, pp. 1213–18.13. Cf. Book ii, Chapter xx, Section xxxii(f) in Grotius, The rights of war and peace, vol. 2,

pp. 1010–13.

the two kantian concepts of right 19

more severe crimes rather than lesser crimes, as well as to detercriminals who are committing a crime from committing a furthercrime, for instance to deter robbers from murdering the victims whocould identify them.14 Yet this rule for determining the degree ofpunishment is not to be confused either with talion law – which wouldimply an absolute equality between crime and punishment – or evenwith an arithmetic equality. (An arithmetic equality would imply pro-portionality between the difference in severity among the crimes, onthe one hand, and the difference of severity between the degrees ofpunishment to which criminals are sentenced, on the other hand.)Since early modern penal theorists argue against punishmentsharsher than those required for deterrence, and since they assumethat the practice of talion law exceeds the alleged equality of talionlaw, they see themselves, in comparison with talion law, as promotingclemency toward the convicts. The characteristic of avoiding “super-fluous and needless pain” from being inflicted on the convict istermed “frugality” by Jeremy Bentham.15

Fourthly, the inner logic of the requirement for punishment tobe no harsher than deterrence requires leads Cesare Beccaria toreject punitive torture and to prefer the punishments that simultan-eously make the most efficient and durable impression on the peopleand are the least painful for the convict. The goal is not the realinfliction of an evil or of pain, but the impression made first ofall on other citizens, secondarily on the criminal himself or herself.For example, as the utilitarian Henry Sidgwick puts it, “from a utilitar-ian point of view . . . punishment should be, so far as possible, whatBentham calls ‘exemplarity,’ that is greater in appearance thanin reality, since it is chiefly appearance that deters.”16 Mill defendsthe same view when he rejects the abolition of the death penalty andits replacement by a life sentence because he reproaches the abolition-ists with being “driven to inflictions less severe in appearance, andtherefore less efficacious, but far more cruel in reality.”17

14. Cf. Chapter 27 of Beccaria, On crimes and punishments, p. 63: “The harsher thepunishment and the worse the evil he faces, the more anxious the criminal isto avoid it, and it makes him commit other crimes to escape the punishment ofthe first.”

15. Chapter 15, } 11 in Bentham, Principles of morals and legislation, p. 194.16. Sidgwick, The elements of politics, p. 120.17. John Stuart Mill, “April 1868 speech on capital punishment,” in Mill, Utilitarianism,

ed. George Sher (Indianapolis: Hackett, 2001), pp. 65–70 (p. 65).


Fifthly, punishment is not considered as following any of the trad-itional principles of justice such as commutative justice, distributivejustice or corrective justice. Grotius and Pufendorf present thefollowing arguments for rejecting the classification of punishment inany of the usual categories of justice. Since the aforementioned pro-portionality, which would correspond to a geometric equality – in anAristotelian sense – has been rejected, punishment does not followany principle of distributive justice. Does punishment follow aprinciple of commutative justice? Punishment is not owed to thecriminal as being either the payment of a debt or the fulfillment ofany other agreement.18 Nor is punishment to be thought of as com-pensation for damage. Also, there exists a duty that requires one togive back what each person merits only in the case of a positive meritand not in the case of demerit.19 Instead of only following a principleof justice, punishment is subject, first and foremost, to prudence(prudentia) and to public utility.20 Admittedly, punishment can be saidto be “just,” but rather in a negative sense, by which no injustice arisesby punishing someone who deliberately caused an evil to the othermembers of the commonwealth. Yet it is not justice but rather pru-dence that evaluates the degree of punishment necessary forobtaining a deterrent effect. Furthermore, although criminals oughtto understand that being punished is not unjust, justice does notrequire them either to inflict punishment upon themselves or tocooperate in the infliction of punishment upon their own person. Itis solely the other citizens who must cooperate in the latter case.

Kant is the first theorist of penal law who derives from talion lawnot only a rule for determining the degree of punishment, but also ajustification for punishment. In so doing, he introduces an alternativetheory that is external to the idea of deterrence, to which all compet-ing theories hitherto belonged. Therefore, Kant is not only a centralfigure in retributivism, but also its founding father.

The Kantian turn concerns all five elements mentioned. Concern-ing the first point, Kant excludes neither all the deterrent effects norall the deterrent goals of punishment. On the contrary, in the Doctrineof virtue, for example, he even rejects any “renunciation of rigorous

18. Cf. Grotius, Book ii, Chapter xx(2), in The rights of war and peace, vol. 2, p. 995; andBook viii, Chapter iii(4f) in Pufendorf, De jure naturae et gentium, vol. 2, pp. 1152–6.

19. Cf. Book viii, Chapter iii(15) in Pufendorf,De jure naturae et gentium, vol. 2, pp. 1186–7.20. Cf. Pufendorf, Book viii, Chapter iii(24), in Pufendorf, De jure naturae et gentium,

vol. 2, p. 1210.


means (rigorosa) for preventing the recurrence of wrongs by others;for then a human being would be throwing away his rights and lettingothers trample on them.”21 What Kant actually rejects is the consider-ation of punishment as merely being a means toward the fulfillment ofan end, for instance toward the fulfillment of an end of preventingcrime before it occurs. According to Kant, first, punishment is justifiedonly by the fact that the criminal violated the law, and, secondly, sothat the only such punishment that complies with this justificationis the one that corresponds to “justice” as defined by ius talionis. Inall cases in which punishment cannot be justified in this way, it isunjust and therefore prohibited by Kant, however strong the deterrenteffect may be.

The consequence of this is that, concerning the fifth point, Kantregards the justification of punishment as concerning only justice –and not prudence. According to Kant, not only is it not unjust thatcriminals are punished, but justice even requires their punishmentaccording to talion law. However, it is unclear whether for Kant thisrequirement of justice belongs to the realm of legal justice or of moraljustice, which will become apparent below (see Section 1.2).

As for the fourth point, the impression made by punishment on thecitizens and the criminal does not matter to Kant. What matters to himis, instead, solely the evil inflicted upon the criminal who is deservingof the punishment. Concerning the third point, Kant unrestrictedlyadopts talion law, although with some adaptations. He considers thattalion law inflicts more severe degrees of punishment than deterrencetheory, in particular more severe than Beccaria does – for Beccaria isallegedly “moved by overly compassionate feelings of an affectedhumanity.”22 One must note that today’s defenders of Kant’s theoryconsider that talion law imposes an upper limit on the degree ofpunishment, which, they allege, is exceeded by deterrence theories.

Interestingly enough, in regard to the second point, it was notKant himself but today’s Kantian retributivists who reproach thedeterrence theories for tolerating the punishment of innocents undercertain circumstances.

Owing to the contrast in the first point, that is, between the deter-rence theories and Kant, who comes to link the punishment and its

21. TL Ak vi:461. Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge:Cambridge University Press, 1996), p. 578.

22. RL Ak vi:335. Practical philosophy, ed. Gregor, p. 478.


degree only to the moral demerit of the criminal (something that onlya human being is capable of bearing), Kantian retributivists also seethe core of their theory as presenting the only theory that respects thehumanity in the person of the criminal. In this respect, Kantianretributivism radically diverges from the ancient function of talionlaw, as described by Henry Sidgwick: “at an earlier stage of social andintellectual development, th[e] distinction [between retribution andreparation] is obscure, or but faintly perceptible; the penal loss of an‘eye for an eye’ or a ‘tooth for a tooth,’ was commonly regarded as akind of reparation to the person originally maimed.”23 Thereby,respect for humanity is not to be understood in the utilitarian ornatural law sense, that is, as keeping the evil or the pain inflictedupon the criminal as low as possible. Rather, in Kant’s view, respect forthe humanity in human beings considered as beings capable of reasonconsists in two dimensions. The first dimension is treating the human-ity in the criminal’s person as an end in itself. In this book, I shall notenter into Kant’s complex theory of autonomy and free will. Nor doI need to deal with this theory. I limit myself to observing that therecannot be any respect for the free will of rational beings without firstallowing them external freedom, provided that it conforms to thelimits imposed by the freedoms of the others. The second dimensionin which respect for humanity consists is the infliction of as much evilon the criminal as is merited; it is respectful of the humanity in thecriminal’s person because this demerit is the choice of the criminal,and therefore everybody ought to have the foreseeable consequencesof their actions expressed, that is, of their merits and demerits.In } 49e of the Doctrine of right, which Kant devotes to penal law, heassumes that the latter requirement – meaning that everybody oughtto accept the consequences of their actions – belongs to the formerone. Yet he does not provide any argument supporting this point,and I think that none is possible in the Kantian system. Thus I shallassess the questions of whether retributivism can be grounded in thefirst dimension and of whether it can be grounded in the seconddimension as two separate issues.

The importance of proceeding through these two separateapproaches is enhanced by the link between the two dimensions ofrespect for humanity, on the one hand, and the controversial issue ofthe relationship between law and morality, on the other hand. There

23. Sidgwick, The elements of politics, p. 107.


are two theses that pertain to this relationship. Either the legal systemis largely independent from morality and its moral premise does notgo beyond preserving the first dimension that involves respect forhuman dignity by mutually limiting external individual freedom ofeach member of the legal system, or law is subordinated to morality,which it intends to enforce to some extent – which may be very wide.Whereas the first dimension involves respect for human dignity andbelongs, incontrovertibly, to the competence of the law, the answerto the question whether the second dimension can belong to thecompetence of the law depends on the answer to the question aboutthe relationship between law and morality in general. Now, if itappears that retributivism cannot be justified by respect for humandignity as taken in its first dimension, one should still inquire into thepossibility of justifying it by respect for human dignity in the seconddimension, for the case of an extensive subordination of the legalsystem under morality.

In what follows, this twofold question of whether retributivism canbe supported by either of the two dimensions of respect for humandignity mentioned above will eventually be answered in the negative inboth dimensions. Thus, unlike the case in which only one of the twoquestions is answered in a negative way, the other being answeredpositively, I will not be obliged to choose either between the twodimensions or between the two theses mentioned above. I shall thuscontent myself with exploring the consequences of both dimensionsof respect for human dignity for the justification of punishment. It willultimately appear that they both support the same justification ofpunishment and the same degree of punishment.

1.2. Two Kantian concepts of right

Some like to advance the notion that Kant created a revolution inlegal philosophy just as much as he did in moral and theoreticalphilosophy, and that he made a significant contribution to today’sdiscussion. According to this notion, we have Kant’s Doctrine ofright24 to thank for a liberal concept of right: right (Recht) as the

24. Throughout the rest of this work, an italicized and capitalized Doctrine of right will referto the title of the work, whereas the doctrine itself, that is, Kant’s conception of right,will be lower-case roman. According to the common usage in Kant’s age, Kant employsthe term doctrine of right in the sense of ius (RL Ak vi:229), referring therefore to law(Recht).


coexistence of empirical freedoms according to the universal law ofequal rights for all legal persons. This concept of right offers twoadvantages. On the one hand, it claims to comply not with ademanding moral justification, but only with the mere reciprocityin the coexistence of empirical freedoms. On the other hand, itappears not to contradict morality, because it demands, just as thecategorical imperative does, a universal law. In this sense WolfgangKersting, for example, writes:

The rational principle of right requires from everyone the verycurtailment of freedom to which everyone who influences each otherthrough actions in their freedom would agree to under fair conditions ina situation free of threats, namely, a strictly universalizable curtailmentof freedom that limits everyone in the same way.25

Likewise, Otfried Hoffe writes about the Doctrine of right that, on theone hand, “the political and legal theory that derives from it [from thecategorical legal imperative] consists of a political . . . liberalism.”26

On the other hand, according to Hoffe, the Metaphysics of moralsorients “the principles of moral philosophy in the (metaphysical)doctrine of right toward an external, and in the (metaphysical) virtuetheory, toward an internal lawgiving.”27 I believe, though, that theseaspects are not compatible with one another. In believing so, I am notin any way disputing that Kant advanced them. He just did notadvance them in the same work.

The first aspect, that is, right as coexistence of empirical freedomswith equal rights, is advanced by Kant in, for example, Toward perpetualpeace (1795) as “a republican constitution,” while requiring that the“civil constitution in every state” should “be republican,”28 and definingthe republic in the following manner:

A constitution established, first on principles of the freedomof themembersof a society (as individuals), second on principles of the dependence of all

25. Wolfgang Kersting, Wohlgeordnete Freiheit: Immanuel Kants Rechts- und Staatsphilosophie,second edition (Berlin: De Gruyter, 1993), p. 27.

26. Otfried Hoffe (ed.), Immanuel Kant: metaphysische Anfangsgrunde der Rechtslehre (Berlin:Akademie Verlag, 1999), p. 8.

27. Otfried Hoffe, “Der kategorische Rechtsimperativ: Einleitung in die Rechtslehre,” inO. Hoffe (ed.), Immanuel Kant: metaphysische Anfangsgrunde der Rechtslehre (Berlin:Akademie Verlag, 1999), pp. 41–62 (p. 48).

28. ZeF Ak viii:349. Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge:Cambridge University Press, 1996), p. 322.


upon a single common legislation (as subjects), and third on the law oftheir equality (as citizens of a state) . . . is a republican constitution.29

Let it be noted here that equality before the law is by no meansjustified by the categorical imperative.

The second aspect is advanced by Kant in his Doctrine of right, whichis why the Doctrine does not offer a liberal concept of right. Withregard to the relation of right to morality in Kant, one typicallydistinguishes three main options: (1) the thesis of complete inde-pendence (initially proposed by Julius Ebbinghaus); (2) the depend-ence thesis of the necessary derivation of right from the principle ofmorality (initially advanced by Kantian legal theorists in the years inbetween the Groundwork of the metaphysics of morals and the Doctrine ofright30); and (3) the thesis of limited dependence, arguing depend-ence on morality for the validity (Geltung) or for the adjudication ofright, and, in contrast to that, arguing independence from moralityfor the realization or for the execution of right. Independence in theexecution of right has the essential feature that, whereas moralityenvisions the execution – that is, the carrying out of actions – out ofduty (Pflicht), right relies on coercion (Zwang) as the incentive. Thisthird thesis appears to me to be philologically correct. The systematicevaluation connected to it overlooks, however, the scope of right’sdependence on morality in the adjudication.

Interpreters of Kant concentrate, though, on what differentiatesKant from the Kantian legal theorists, who, even before the publica-tion of the Doctrine of right (1797), had tried to develop a theory ofright on the basis of the Groundwork of the metaphysics of morals (1785).Most of the Kantian legal theorists, such as Gottlieb Hufeland andTheodor Schmalz, derived right from the principle of morality, and,in fact, from the principle of morality as an adjudicative principle.In doing so, they understand right, just as Kant does, to mean thatwhich regulates the external relations between humans. Externalactions and relations are already objects of moral duty. The differencebetween moral duty and right is placed by these Kantian legal theoristsin the differentiation between, on the one hand, what is morally com-manded or forbidden and, on the other, what is morally allowed.

29. ZeF Ak viii:349f. Practical philosophy, ed. Gregor, pp. 322–3.30. Cf. W. Kersting, “Sittengesetz und Rechtsgesetz: die Begrundung des Rechts bei Kant

und den fruhen Kantianern,” in Reinhard Brandt (ed.), Rechtsphilosophie derAufklarung: Symposium Wolfenbuttel 1981 (Berlin: De Gruyter, 1982), pp. 147–77.


The derivation from the principle of morality is admittedly twofold.On the one hand, a part of right consists of the morally allowed or is inthe realm of the moral adiaphoron. On the other hand, the Kantianlegal theorists appeal to the implication that whatever is morally com-manded is therefore morally allowed. Thus, moral duty, next to themoral adiaphoron, belongs to right: I have the right to do the morallyallowed as well as the morally commanded. Depending on the author,right is therefore classified into absolute or conditional, or into alien-able and inalienable, or even into complete and incomplete rights.Onecould possibly distinguish a first step to the contemporary differenti-ation between human rights and the other rights, in so far as humanrights count as being absolute rights, whereas the other rights differfromcommonwealth to commonwealth and from situation to situation.

Such a twofold derivation of right from the principle of moralityleads, however, to three major problems. The first of these is thederivation that assembles right out of two heterogeneous parts relyingon two very different normative justifications, which calls into doubtwhether right as such is derived, that is, if right really forms a wholeand can be differentiated from morality, or whether it is, rather, aweakened version of morality. Secondly, the authority to coerce islacking. Thirdly, the authority to coerce in the case of the merelymorally allowed appears more difficult to justify, if at all, than in thecase of the morally commanded.

1.3. Kant’s concept of right derived from the principleof morality in the Doctrine of right

The division of Kant’s Metaphysics of morals into the Doctrine of right andthe Doctrine of virtue obviously attempts to propose a doctrine of rightthat is not subject to the Kantian legal theorists’ aforementionedproblems. The core relation between the doctrine of right and thedoctrine of virtue consists of two aspects. First, the doctrine of rightand the doctrine of virtue can be derived from the same principle,that is, from the categorical imperative; this derivation, in the case ofthe law, implies no weakening of morality: the obligation remains thesame. Secondly, the difference between the doctrine of right and thedoctrine of virtue rests on the different motives for their respectiveexecution. The execution of the doctrine of right only has to deal withcoercion as its motivation, whereas the doctrine of virtue is responsiblefor the execution out of duty.


It arises from the second aspect that “the substantive boundaries ofthe duties of right are defined . . . through the range of the externalmotivation that is related to them in juridical lawgiving: wherevercoercion fails to motivate, juridical lawgiving does not rule.”31 Likemost of the other interpreters, Kersting highlights the second point:

Therefore, every duty of right is always also an indirectly ethical duty;the ethical mode of assigning obligation a priori inherent in every duty isnot rescinded for the category of duties of right just because they can beissued in a juridical manner.32

I would, however, like to accentuate the first point: the area of rightencompasses every ethical duty that can also be fulfilled through coer-cion. In other words, right should provide for the fulfillment of everyethical duty that could be fulfilled through coercion by coercion itself(that is, through both the threat of coercion and the application ofcoercion).

But which duties can be fulfilled through coercion? Here too Kant’sanswer appears to be clear: not maxims, but actions. Bernd Ludwigformulates this in the following manner:

Now, what is the fundamental substantive difference between theprinciple of right and the categorical imperative? . . . When the“disposition” (Gesinnung) [Ak vi:393] of the actor – which is notaccessible to external lawgiving – both cannot and is not allowed to beincluded, it is also not possible to address demands to it. However, thathis or her actions at least can coexist with a universal lawgiving is indeedexternally enforceable – and without having to exert influence for it onthe maxims of the person concerned.33

It appears to me that there is an inaccuracy in Ludwig’s formulation.The third formula of the categorical imperative, to which Ludwigobviously refers, reads: “act in accordance with a maxim that can at thesame time make itself a universal law.”34 If right is derivable, on the onehand, from the categorical imperative, and it is, on the other hand,responsible for that part of this imperative that allows it to be imposedthrough coercion, then the procedure for determining the content ofright should look something like the following. As a first step, maxims

31. Kersting, Wohlgeordnete Freiheit, p. 176.32. Kersting, Wohlgeordnete Freiheit, p. 176.33. Bernd Ludwig, Kants Rechtslehre (Hamburg: F. Meiner, 1988), p. 95.34. GMS Ak iv:436f. Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge:

Cambridge University Press, 1996) p. 86.


that survive the test should be determined. As a second step, theactions (by which I mean both active actions and abstaining fromrealizing actions) that these maxims lead to should be investigated.As a third step, an investigation must really seek to determine which ofthose actions can be realized by means of coercion; right must thensee to it that these actions are met with a threat of coercion or anexecution of coercion. Therefore, Ludwig’s formulation should bemodified thus: “However, that his or her actions at least can coexistwith a universal lawgiving of the maxims, that is, with the categoricalimperative, is indeed partly externally enforceable – and without havingto exert influence as a counterpart on the maxims of the personconcerned”; and it should also be enforced by means of coercion inso far as it is enforceable by coercion. In contrast, Ludwig’s formula-tion (“that his or her actions at least can coexist with a universal law-giving”) leaves open the possibility that law is nothing other than theequality of all legal persons with regard to actionable rights. Such aconception of right is not compatible with Kant’s Doctrine of right, asI will show in what follows.

The premises stating that the actions, at the very least, can coexistwith the universal lawgiving of the maxims, that is, with the categoricalimperative, and that this is partly externally enforceable by meansof coercion, and also should be enforced by this, in so far as it isenforceable, imply the conclusion not only (1) that rights are action-able by the persons concerned, but also (2) that action must be taken,and (3) that the enforcement should be compelled by coercion fromthose who would like to abstain from it. I shall content myself withonly one example of this here; Kant writes:

Thus ethics commands that I still fulfill a contract I have entered into,even though the other party could not coerce me to do so; but it takesthe law (pacta sunt servanda) and the duty corresponding to it from thedoctrine of right, as already given there. Accordingly the giving of thelaw that promises agreed to must be kept lies not in ethics but in Ius.

All that ethics teaches is that if the incentive which juridical lawgivingconnects with that duty, namely external constraint, were absent, theidea of duty itself would be sufficient as an incentive.35

It is certainly the case that no legal system would ever want to enforcefulfillment of every promise. Not even the relevant specifications fromKant’s Doctrine of right fulfill this challenging requirement (this is often

35. RL Ak vi:219–20. Practical philosophy, ed. Gregor, pp. 383ff.


overlooked by most of the interpreters of Kant).36 The law – theKantian right as well – deals with promises in very different ways.It is possible to differentiate between the following conceivable cases:

1 Always, and without action by the person concerned, the fulfillmentof promises is enforced by means of coercion, or, alternatively,compensations are compelled from a promise breaker.

2 Only by the suit brought by the person concerned – but in this casealways – is either the fulfillment of the promise enforced, or, alterna-tively, thedeliveryof compensations compelled fromapromisebreaker.

3 By the concerned person’s suit, only is the fulfillment of somepromises (for example, only the fulfillment of contracts that fulfillcertain conditions) enforced, or are remedies compelled for non-fulfillment of some promises (for example, for the non-fulfillmentof contracts that fulfill certain conditions).

4 The fulfillment of some contracts that fulfill certain states of laws is,under certain circumstances and despite the action brought, notenforced, nor are compensations compelled.

When we take Kant’s formulation cited above seriously, then only thefirst, or at most also the second, option should be possible in hisdoctrine of right. But Kant advances different options, which is exactlywhat occurs in every legal system. Here is one example for each optionthat appears in Kant’s writings.

1 The preliminary articles of Toward perpetual peace are in no sense atthe disposal of the contractual partners, even if all partners were tobe in agreement to abrogate an article.37

2 Kant claims that “if one of the partners in a marriage has left orgiven itself into someone else’s possession, the other partner isjustified, always and without question, in bringing its partner backunder its control.”38

3 The adherence to contracts with “servants” is only compelled inlimited variations: “The contract of the head of a household withservants can therefore not . . . [be concluded] for life, but at mostonly for an unspecified period of time, within which one party maygive the other notice.”39

36. Cf. Kuhl, Die Bedeutung der Rechtsphilosophie, p. 41.37. ZeF Ak viii:343–7. Practical philosophy, ed. Gregor, pp. 317–21.38. RL Ak vi:278. Practical philosophy, ed. Gregor, p. 427.39. RL Ak vi:283. Practical philosophy, ed. Gregor, p. 432.


4 The sentence “sale breaks a lease”40 provides for the termination ofrental contracts without compelling compensation for loss (butwith adherence to the cancellation period). In Kant’s view, as withthe current legal system, only in very few cases (for example, in thecase of lying in court or in the case of breach of trust, etc.) are lieslegally proscribed; still less is the truth compelled, even though withKant, as is generally known, the moral prohibition against tellinglies applies without exception.

Even in penal law, in which Kant counts as being the strict supporterof a categorical retributive imperative, he does not accept all theconsequences of his concept of right. On the one hand (to name awell-known case) if all the inhabitants of an island leave their home-land, he nevertheless requires that all convicts, including those whohave been condemned to death, should actually serve their sentencesout, but then, on the other hand, he concedes a power of pardon tothe sovereign. In this vein, he writes, on the one hand:

Punishment by a court . . . can never be inflicted merely as a means topromote some other good for the criminal himself or for civil society.It must always be inflicted upon him only because he has committed a crime

. . . The law of punishment [that is, of criminal law] is a categoricalimperative.41

On the other hand, however, he claims:

With regard to crimes of subjects against one another it is absolutely notfor him [the sovereign] to exercise it [the power of pardon]; for herefailure to punish (impunitas criminis) is the greatest wrong against hissubjects. He can make use of it, therefore, only in case of a wrong doneto himself (crimen laesae maiestatis).42

Here the enforcement of the legal–ethical consequences of a crimestands at the disposal of other persons. At this point, the categoricalimperative turns into – in opposition to Kant’s concept of right – amere hypothetical imperative.

Such isolated inconsistencies muddle not only the effective rangeof Kant’s concept of right, but also the systematic differentiation thatultimately proves to be untenable and leads to the situation that withKant some actions, which, according to his Doctrine of right, should be

40. RL Ak vi:361. Practical philosophy, ed. Gregor, p. 496.41. RL Ak vi:331. Practical philosophy, ed. Gregor, p. 473.42. RL Ak vi:337. Practical philosophy, ed. Gregor, pp. 477–8.


the concern of a legal command or proscription, turn out not to be so,because Kant separates many duties toward external actions from theircorresponding actions in an inconsistent way.

As is generally known, Kant differentiates between duties of rightand duties of virtue. He claims that duties of virtue cannot be subjectto “external lawgiving,”

because they have to do with an end which (or the having of which) isalso a duty. No external lawgiving can bring about someone’s setting anend for himself (because this is an internal act of the mind), although itmay prescribe external actions that lead to an end without the subjectmaking it his end.43

But if external actions are actually “commanded” by the correspond-ing categorical imperative or by the corresponding duty, then theseactions – because they are external, and provided that they areenforceable by means of coercion as such – are legally dictated, thatis, legally enforceable by means of coercion.

In this respect, we should not allow ourselves to be led astray eitherby the asymmetry in the presentation of duties in the Groundwork of themetaphysics of morals – for instance, in the commentary on the categoricalimperative’s formula of humanity as an end in itself – or by qualifyingas duties of right the duties toward the preservation of humanity asan end in itself.

The classification of duties follows two dichotomic criteria: “dut[ies] to oneself” are opposed to “dut[ies] to others,” just as the“preservation of humanity as an end in itself” is opposed to a “further-ance” of humanity as an end in itself. The asymmetry arises out of Kantreferring to the proscription of false promises and the “assaults onfreedom and property” as duties for the preservation of humanity as anend in itself, whereas no duties toward its furtherance are named. Atleast three arguments speak against combining the asymmetry in thepresentation offered by Kant with a conceptual asymmetry. First, itwould be absurd to believe that the proscription of false promises andthe attack against freedom and property apply unconditionally. Ifproperty or freedom were to be abused as the means of a criminalaction, both could then be vastly curtailed or suspended for longerperiods of time. Secondly, the proscription of false promises is to beunderstood as a duty of right. A false promise is a promise that the



promiser does not intend to keep. This intention is, however, not anexternal action. An external action is present if, instead of observing afalse promise, one observes the non-fulfillment of a promise; in thisway, however, the moral prohibition of false promises becomes a legalproscription of non-fulfillment of promises. Last but not least, thirdly,the duties toward the furtherance of humanity as an end in itself canbe formulated as duties toward external action, for instance the dutiestoward establishing basic schooling or basic medical care for all. Beingduties toward external – and enforceable – actions, these dutiesshould definitely belong to the duties of right, whereas in Kant they,being duties toward the furtherance of humanity, count as beingduties of virtue. For these reasons, the identification of the differenti-ation between duties toward the preservation of humanity and dutiestoward the furtherance of humanity, on the one hand, with a differenti-ation between duties of right, in the legal sense of enforceable rights,and duties of virtue, in the sense of non-legally enforceable duties, onthe other hand, is generally untenable.

But how far does the range of Kantian right actually extend?The categorical imperative cannot surely determine, a priori, either

the single proscribed, dictated and allowed maxims or the singleproscribed, dictated and allowed actions, because the moral appraisalof maxims and actions presupposes knowledge of their empiricalattributes and of their context. In contrast, however, it is alwayspossible to determine with the categorical imperative whether a givenmaxim or action is contrary to duty or in conformity with it, and, inthe latter case, whether it is morally dictated or just morally allowed.Apart from that, every maxim, being as general and indirect as it canbe, is focused on external actions. If we have a general maxim forguiding our lives, we can really determine for all cases that this or thatexternal action is compatible or incompatible with our maxim, and ifwe necessarily have to carry out the action when following our maxim.That is why one can say about all external actions – taking intoconsideration their entire context – whether, in each case, they aremorally proscribed, dictated or merely allowed.

All external actions can potentially be influenced by coercion tosome extent. Therefore, all external actions are not only objects ofmoral judgments, but also objects of right in the sense of Kant’sconcept of right. Apart from that, the moral and the legal judgmentsof external actions should be decided identically every time. The onlything in the moral judgment that does not have any influence on the


legal judgment is the answer to the question whether the morally andlegally dictated action was executed out of duty or merely in conform-ity with duty.44 In this respect, morality is certainly more demandingthan law.

The range of right presented here, nevertheless, greatly surpasseswhatever belonged to the area of law. The proximity of law andmorality is likewise here much closer than it ever was in a legal system.Especially if one considers that the categorical imperative issues,amongst other things, dictates and proscriptions about aspects ofpersonal lifestyle pertaining neither to the rights nor to the legitimateinterests of other individuals, then Kant’s conception of right appearsto be hardly liberal and, in fact, very odd.

1.4. The liberal interpretation of the Doctrine of right

In view of the fact that Kant, like Locke and Mill, belongs to that groupof authors one tends to refer to for the definition of what the liberalstate is, these implications may astound. Yet, the liberal interpretationof Kant is based on two elements.

Regarding the first element, the derivation of the concept of rightfrom the categorical imperative is usually underestimated. MarcusWillaschek, for example, goes so far as to claim:

Kant nowhere really says that the principle of right can be derivedfrom, or is based on, the categorical imperative. The moral law andthe categorical imperative are not even mentioned in }}A–E of the“Introduction to the doctrine of right,” where Kant introduced theprinciple of right.45

And Thomas Pogge begins his enquiry into the concept of rightdirectly with the last lines from } B: “Kant defines Recht as ‘the wholeof the conditions under which the choice of one can coexist . . . withthe choice of the other according to a universal law of freedom.’”46

44. Otfried Hoffe, in “Konigliche Volker”: zu Kants kosmopolitischer Rechts- und Friedenstheorie(Frankfurt a.M.: Suhrkamp, 2001), p. 11, correctly remembers that the differentiationsbetween morality and legality and between the duties of law and virtue are two differentdifferentiations.

45. Marcus Willascheck, “Why The doctrine of right does not belong in The metaphysics ofmorals,” Annual Review of Law and Ethics, 5 (1997), 205–27 (p. 230).

46. Thomas W. Pogge, “Is Kant’s Rechtslehre a ‘comprehensive liberalism’?” in MarkTimmons (ed.), Kant’s Metaphysics of morals (Oxford: Oxford University Press,2002), pp. 133–58 (p. 137).


In the “Introduction to the metaphysics of morals,” the division ofthe Metaphysics of morals is undertaken according to the motive of thefulfillment of the dictates of the categorical imperative. The decisivedefinition of right is found in the following formulation: “The doc-trine of right and the doctrine of virtue are therefore distinguishednot so much by their different duties as by the difference in theirlawgiving, which connects one incentive or the other with the law.”47

This definition consists of (1) the duties that are common to moralityand to right, which result from the categorical imperative,48 and (2)the difference in the respective lawgiving and incentive.

In } b of the “Introduction to the doctrine of right” the second andthird paragraphs give a definition of right. Right at the beginning, theduties that are common to morality and to right are recalled: “Theconcept of right, insofar as it is related to an obligation correspondingto it (i.e., the moral concept of right).”49 The three following pointsdo indeed draw on the difference between right and morality (rightpertains only to, first, “the external and indeed practical relation” ofpeople; secondly, choice [Willkur] in contrast to mere wishes; and,thirdly, the form of choice in contrast to its matter). One usuallyoverlooks, though, that at the end of this very same paragraph dutiesthat are common to right and to morality are again mentioned:Kant speaks of the form of choice “insofar as choice is regardedmerely as free, and whether the action of one can be united with thefreedom of the other in accordance with a universal law.”50 Hence,it is about choice as form, that is, about choice “insofar as choice isregarded merely as free.”

This clearly means that the freedom we are talking about here isnot the freedom of action in choice, but instead the pure freedomrequired in Kantian morality, that is, a freedom of will in autonomy.Therefore, the “universal law” concerned here cannot just be equalityof rights, but instead must be a law suited to the actor’s freedom of

47. RL Ak vi:220. Practical philosophy, ed. Gregor, p. 384.48. Nothing changes with these duties if, like Willaschek, one presupposes an additional

level of the imperative (“Act according to duty from duty,” p. 214), which is whyWillaschek’s “alternative thesis” is much closer to the official thesis than he himselfbelieves, and why it is not really relevant for the treatment of our problem. Willaschekhimself believes his thesis to be a thesis of limited independence of the law frommorality; in his own words, “rather an independent, basic law of practical rationality”(p. 223).

49. RL Ak vi:230. Practical philosophy, ed. Gregor, p. 387.50. RL Ak vi:230. Practical philosophy, ed. Gregor, p. 387.


will: it is the same “universal law” as the one in the third formula of thecategorical imperative (“act in accordance with a maxim that can at thesame time make itself a universal law”).51 Because the individual formulasof the categorical imperative are versions of the same imperative, andtherefore equivalent, it also says: “So act that you use humanity, whether inyour own person or in the person of any other, always at the same time as anend, never merely as a means.”52 The “reciprocal relation of choice” is,therefore, more demanding in relation to Kant than the mere coexist-ence of the powers of choice or the freedoms of action under theprinciple of equality as in the liberal understanding of a just legalsystem.

The reference to a demanding moral freedom is indeed present inthe concluding formulation of } B, in the “Introduction to the doc-trine of right,” though not emphasized: “Right is therefore the sum ofthe conditions under which the choice of one can be united with thechoice of another in accordance with a universal law of freedom.”53

One should, at this point, take note that Kant uses the expression “inaccordance with a universal law of freedom,” which is neither the sameas just saying “in accordance with a universal law” nor the same as“according to a universal law of choice.” In the former case, “freedom”(Freiheit) would be redundant, and should have been “choice”(Willkur), as in the middle of the sentence. In the latter case, “offreedom” would be superfluous. At the beginning of } C, we find avariant of the very same definition of right in which it is stated more

51. GMS Ak iv:436f. Practical philosophy, ed. Gregor, p. 86. Allen W. Wood rejects thispoint with the following argument: “as to the ‘universal principle of law’ itself, it ishard to sustain the view that it can be derived from the moral imperative. Thisprinciple is: ‘any action is right if it can coexist with everyone’s freedom inaccordance with a universal law, or if on its maxim the freedom of choice of eachcan coexist with everyone’s freedom in accordance with a universal law’. . . Thisprinciple may bear a superficial resemblance to the Formula of Universal Law: ‘Actonly in accordance with that maxim which you can at the same time will that it becomea universal law . . .’ Like that formula, the principle of right provides us with a test onlyof the permissibility . . . But the principle of right says nothing about willingmaxims asuniversal laws.” A. W. Wood, “Kant’s Doctrine of right,” Introduction to Otfried Hoffe(ed.), Immanuel Kant: Metaphysische Anfangsgrunde der Rechtslehre (Berlin: AkademieVerlag, 1999), pp. 19–39 (p. 35). Wood’s argument against the derivation of law fromthe categorical imperative is, however, only valid against a derivation of his “execution”from the categorical imperative; it is not convincing against an adjudication of actionsfrom the categorical imperative.

52. GMS Ak iv:429. Practical philosophy, ed. Gregor, p. 80. For the meaning of the term“humanity” see below p. 64.



precisely which aspect of choice is meant (“the freedom of choice”54

[Freiheit der Willkur], that is, the form of the freedom, or the freedom ofwill in choice), and this extraneous element “according to a universallaw of freedom” (emphasis mine) becomes dispensable and disappears:“Any action is right if it can coexist with everyone’s freedom in accord-ance with a universal law.”55 In the remaining part of } C, as well asthroughout the entire Doctrine of right, the “freedom” of “choice” and“universal law” ismentioned at every repetitionof thedefinitionof right.Explicit references to the categorical imperative and to duty, however,appearmore seldom, so that the framework, and hence the exact mean-ing of “freedom” and “universal law,” can easily be overlooked.

Regarding the second element: the liberal interpretation of Kantalso rests on a minimalist understanding of the second formula of thecategorical imperative. Kant in fact applies (as shown above) only onepart of the categorical imperative to law, because he is of the opinionthat the duties toward the furtherance of humanity are allowed torelate only to an “end” (Zweck) and not to an external action so thatthey cannot constitute any “external lawgiving.” In contrast, for Kant,the duties toward the preservation of humanity as a goal in itself canconstitute an external lawgiving, and indeed all of these duties as awhole can constitute such an external lawgiving. Now, the minimalistunderstanding leads one to believe that, even if one neither overlooksnor forgets the derivation of the Kantian concept of right from thecategorical imperative, the requirement addressed by the categoricalimperative to the legal system is strictly limited to the protection of thefellow citizens or to the prohibition of homicide, slavery, withdrawal ofcitizenship and the like.

If law means as much as the coexistence of the freedom of action ofall humans according to the law of equality of right, then what is beingdealt with here is a liberal definition of law, which is derived from theharm principle without any reliance on Mill’s utilitarian backgroundor on any comprehensive moral justification. In such a context,the requirements of just such a formula as the second formula ofthe categorical imperative, which has been conceived of minimally,are not especially difficult to fulfill (one of the exceptions consists ofthe case, for example, of suicide, which is forbidden by the categoricalimperative).



A liberal concept of right and a genuine independence of law frommorality can be won not through the adoption of coercion as themotivation for the execution, but instead, first through an independ-ent adjudication, that is, first when the content of law is no longerderived from the categorical imperative. It should be obvious thatsuch a conception of right amounts by no means to legal positivism.The liberal concept of right comprising the set of conditions for thecoexistence of external freedoms, according to the law of equality,contains completely normative requirements (of protection of theexistence of all individuals, of equality before the law, etc.). Manylegally positive and valid legal systems do not, however, conform tothese criteria; rather, they infringe them.

1.5. The liberal concept of right

In the introduction I argued that both of the following aspects canbe found in Kant’s writings: (1) a liberal concept of right and (2) aderivation of right from the categorical imperative that leads toanother concept of right, which is incompatible with the liberal con-cept of right. After I have developed the second aspect, I would like toreturn to the first. Here I shall content myself with presenting twopassages that document Kant’s liberal concept of right:

1. On the concept of right as found in Plato’s Republic, Kant writes:

A constitution providing for the greatest human freedom according to lawsthat permit the freedom of each to exist together with that of others (not oneproviding for the greatest happiness, since that would follow of itself) isat least a necessary idea, which one must make the ground not merely ofthe primary plan of a state’s constitution but of all the laws too.56

2. Kant designates as “the highest task which nature has set formankind” the “establish[ment of] a society in which freedom underexternal laws would be combined to the greatest possible extent withirresistible force, in other words [the establishment of] a perfectly justcivil constitution.”57 Kant states more precisely that it concerns that sortof society that “has not only the greatest freedom, and therefore acontinual antagonism among its members, but also the most precise

56. KrV B 373. Immanuel Kant, Critique of pure reason, ed. and trans. Paul Guyer and AlanW. Wood (Cambridge: Cambridge University Press, 1997), p. 397.

57. Idee, Proposition 5, Ak viii:22. Immanuel Kant, Political writings, ed. Hans Reiss, trans.H. B. Nisbet, second edition (Cambridge: CambridgeUniversity Press, 1991), pp. 45–6.


specification and preservation of the limits of this freedom in orderthat it can co-exist with the freedom of others.”58

The context of both of these quotes, as well as further passages inKant’s works where this concept of right is mentioned, is a teleology ofhistory, what Kersting terms “a conception of right belonging to moralteleology.”59 Kersting rejects, however, that selfsame teleological per-spective as being normatively irrelevant. He raises the following objec-tion against the teleological perspective on right, which he diagnoses ashaving too strong an “affinity” with Wolff’s idea of human perfection:

What kind of sense, though, does that sort of teleological, moral–pragmatic justification of right make if the right of coercion towardthe freedom to fulfill duty relates neither to striving for perfection norto self-preservation, but instead must relate itself to a duty that demandsthe formally perfect treatment of others, and if the moral duty were notto be grounded in a lex naturalis but in an internal lawgiving that wouldnot require the realization of natural designs, but instead require an actout of respect for the law, and for that reason cannot depend onexternal assistance?60

Kersting’s rhetorical question is astonishing for three reasons.First, Kersting overlooks that the natural design in Kant’s moral

philosophy is precisely none other than that the human “natural cap-acities” are “developed” “completely” and “in conformity with theirend” “directed towards the use of his reason.”61 Since a reasonablebeing who follows reason – and not, for instance, his or her impulses –is really obeying the categorical imperative, the natural design inhumans, according to Kant in his Idea for a universal history from a cosmo-politan point of view (1784), is clearly not only the cultivation of his or herrelation to nature, the civilization of his or her behavior toward fellowhumans and the formation of a civil society or state of law, but also themoralization of humans themselves, in short, exactly the “acting out of arespect for the law” required by Kersting, as well as the “moral duty” asan “internal lawgiving,” which is likewise required by him.

Secondly, “the right of coercion toward the freedom to fulfill duty”is surely sufficiently present, as Kersting believes, because Kant speaks ofthe “greatest human freedom according to laws” without guaranteeing

58. Idee, Proposition 5, Ak viii:22. Political writings, ed. Reiss, p. 45.59. Kersting, Wohlgeordnete Freiheit, p. 142.60. Kersting, Wohlgeordnete Freiheit, p. 151.61. Idee, Propositions 1 and 2, Ak viii:18. Political writings, ed. Reiss, pp. 42–3.


this freedom to the individual only under the condition that he or shefulfills any duties: individuals only need to heed the laws of coexist-ence in order to enjoy these freedoms guaranteed by coercion. Thatmeans that they enjoy them equally, no matter if they act morally orimmorally, just not contrary to right. However, freedom is enablingfor an individual striving for the fulfillment of moral duty to fulfill inan appropriate way the duties, which Kant derives from the categoricalimperative, toward the preservation and the furtherance of humanityas an end in itself. The individual receives explicitly through freedommore and better means to preserve and to advance humankind as anend in itself. In this respect, the freedom to fulfill duty contains neithera duty to achieve perfection nor a duty toward self-preservation (other-wise we would not be dealing with a liberal concept of right); rather thefreedom to fulfill duty is contained in the duty toward self-preservationand the duty to achieve perfection (in one of the passages mentionedabove, Kant maintains, for example, that “happiness would followof itself”).62 For this reason, right actually belongs to the “moral–pragmatic” without thereby betraying Kant’s system of reason.

Thirdly, Kersting assumes that “right of coercion toward the free-dom to fulfill duty . . . must relate itself to a duty that demands theformally perfect treatment of others . . . grounded . . . in an internallawgiving . . . requiring an act out of respect for the law, and for thatreason cannot depend on external assistance.” Kersting shows herethat he locates the motivation for compliance to law in coercion, yetsees the source of law as an institution entitled to coerce in the moralduty to establish “the right of coercion toward the freedom to fulfillduty.” In no case do I deny that, according to Kant, it is a part of dutyto enter into a state of laws wherever the state of nature still exists andto contribute to the establishment of a just legal system to the best ofone’s ability. Kant does not say, however, that the legal system shouldoriginate from a human’s conscious fulfillment of duty. On the con-trary, he mentions misery in the state of nature, that is, nature’scoercion being the drive for the establishment of a legal system, whichis opposed to the inclination of humans. Immediately after one of thecitations introduced above, we find the following remark:

Man, who is otherwise so enamoured with unrestrained freedom, isforced to enter this state of restriction by sheer necessity. And this is

62. KrV B 373. Critique of pure reason, ed. Guyer and Wood, p. 397.


indeed the most stringent of all forms of necessity, for it is imposed bymen upon themselves, in that their inclinations make it impossible forthem to exist side by side for long in a state of wild freedom. But onceenclosed within a precinct like that of civil union, the same inclinationshave the most beneficial effect.63

Kant repeats this conception in his Critique of the power of judgment andin the text Toward perpetual peace:

For the problem is not the moral improvement of human beings butonly the mechanism of nature, and what the task requires one to know ishow this can be put to use in human beings in order so to arrange theconflict of their unpeaceable dispositions within a people that theythemselves have to constrain one another to submit to coercive lawand so bring about a condition of peace in which laws have force.64

1.6. A legal and a moral task

If there are really two entirely different concepts of law in Kant, asI believe, which one should count as being the more “Kantian,” that is,which of the two is more consistent with Kant’s system, and which themore convincing concept? The answer to the second question appearsclearly to be: the liberal conception of right accords not only with thereality of the legal systems we rate as just and possibly even the realityof legal systems generally, but also with our normative intuitions.

The answer to the first question is more difficult. We can, however,begin with observing, in Kant’s Doctrine of right, the numerous devi-ations from those actions that, from the perspective of the categoricalimperative, should be held to be necessary, for instance those con-cerning the proscription of lying, the dictate of the fulfillment ofpromises, etc. In Kant’s Doctrine of right, for instance, the right toproperty entails the right to allow fruit to rot without regard for otherpeople actually needing the fruit or being able to make better use ofit. The categorical imperative does forbid such waste, whereas Kant’sright does not. Furthermore, the liberal concept of right is, in contrastto the other concept of right in Kant, not found only in one work:standing alone opposed to the Critique of pure reason (1781), the Ideafor a universal history from a cosmopolitan point of view (1784), the Critiqueof the power of judgment (1790), On the common saying (1793) and the

63. Idee, Proposition 5, Ak viii:22. Political writings, ed. Reiss, p. 46.64. ZeF Ak viii:366. Practical philosophy, ed. Gregor, p. 335.


text Toward perpetual peace (1795) is theMetaphysics of morals (1797); allthe works listed belong to the critical section of Kant’s oeuvre.That the concept of right is more comprehensively developed in theDoctrine of right than in the other works should count less than therepetition of the other concept in systematic contexts. Seen systemat-ically, the liberal concept of right is also more convincing than theothers, because it reinforces the independence of right from morality,expands the range of coercion from compliance with the law to theestablishment of a legal system, and saves theMetaphysics of morals fromdichotomization into a Doctrine of right and a Doctrine of virtue.

Just to name some external evidence: it is striking that many strange– no, even embarrassing – legal regulations found in Kant’s Doctrine ofright, and derivable from a direct application of the categoricalimperative, are shamefully, or out of respect for such an otherwiseimpressive philosophy, left unconsidered. In this context, Kant’s con-ception of marriage can be named, for instance, or his rigorousproscription of “public prostitution (venus volgivaga),” the islandexample, the remark on bestiality, etc.65 Fortunately, such embarrass-ments cannot justifiably be found in the passages of the aforemen-tioned works. Not even Vittorio Hosle, who, on the basis of Kant,Fichte and Hegel, also wants to punish “those crimes whose impunity. . . dissolves the metaphysical dignity of the person,”66 and in doing soadvances a non-liberal concept of right, goes so far as to draw suchconclusions.

Last but not least, many Kant interpreters appear to me usually tomisunderstand the Doctrine of right as a depiction of a liberal conceptof right and thereby they choose more or less consciously the inter-pretation that they best hold to fit into Kant’s system. In this respect,I am more sympathetic to this – in my opinion, false – interpretation.Should I not be mistaken and were it to be correct, I would further-more regret that Kant in the Doctrine of right did not derive thecorresponding legal regulations from the liberal concept of right.

Owing to this, because the interpretation of Kant is still open, I will,in what follows, inquire into the respective consequences that arisefrom both of the interpretations. A further reason compels me to suchan investigation. The argument from } 49e in the Doctrine of right in

65. RL Ak vi:277–80; RL Ak vi:325; RL Ak vi:333; RL Ak vi:363. Practical philosophy, ed.Gregor, pp. 426–30; p. 467; p. 474; p. 498.

66. Vittorio Hosle, “Was darf und was soll der Staat bestrafen?” in Hosle, Rechtsphilosophiedes deutschen Idealismus (Hamburg: F. Meiner, 1989), pp. 54–5.


favor of a theory of retributivism cannot be classified, at first glance,into a liberal legal framework. Just as often, we find in the Doctrine ofright elements67 with which the liberal concept of right is incompat-ible. If Kant speaks, for example, of a proportionality of punishment“to his [the criminal’s] inner wickedness” he obviously means aninternal wickedness as opposed to an external wickedness, that is, toa mere contrariness to right of external freedom.

It is completely possible that such elements actually do not belongto the liberal concept of right, though yet being compatible with it,and that we are dealing with two heterogeneous yet consistent theor-ies of penal law. In this case we would have to observe that Kant doesnot keep to the division into the Doctrine of right and the Doctrine ofvirtue; a Kantian retributivist penal theory would still, however, remaintenable, although in this case it would have to be based on a moralargument. It would then simply reinforce the non-liberal interpret-ation of Kant’s concept of right. Therefore, after we have first investi-gated the arguments that are neutral to the moral theory as wellas those arguments that are influenced by moral theory in supportof retributivism, we can count Kant’s retributivist penal theory asbeing refuted.

In what follows, I will accordingly test initially the arguments inde-pendent of the moral theory in favor of retributivism (the secondchapter of this book) and then the arguments influenced by moraltheory (the third chapter). Both the liberal concept of right andKant’s moral theory will prove to be incompatible with his retributivistpenal theory.

67. Cf. Jean-Christophe Merle, “Il punto di vista educativo e religioso dei Contributidentinati a rettificare il giudizio del pubblico sulla Rivoluzione francese: ladimensione politica del Saggio di una critica di ogni rivelazione,” in Aldo Masulloand Marco Ivaldo (eds.), Filosofia trascendentale e destinazione etica (Milan: Guerini,1995), pp. 303–25.


2

KANT’S LEGALJUSTIF ICATION OF PUNISHMENT

Utilitarianism and deontological ethics have traditionally dominatedthe debate about the justification of punishments and have therebybeen associated respectively with the theories of deterrent and retri-butivist punishment. Nowadays, the debate about the justification ofpunishment is increasingly dominated by mixed theories.1

Most of these mixed theories represent the attempt of deonto-logical philosophers, that is, mainly those influenced by Kant, to breakwith the traditional conception of the deontological, that is, the ratherKantian justification of punishment, as a purely retributive theory.Such theories were actually – with good reason – suspected of restingmore on private morality than on legal principles.

In what follows, I will assess the success of these attempts tostrengthen the Kantian retributivist theory through the integrationinto the justification of punishment of an element of some form ofdeterrence or prevention. First, I hope to show that these attempts areonly superficially mixed, whereas in reality they are completely basedon a retributive foundation. Secondly, I will inquire into the argu-ments that Kant employs to justify the principle of retribution andthe type of retribution he adopts, and attempt to determine theweaknesses of this argument. Finally, I will propose a principle ofpunishment that, in my opinion, fits in better with Kant’s principle

1. Cf. Don E. Scheid, “Kant’s retributivism,” Ethics, 93 (1983), 262–82; Sharon Byrd, “Kant’stheory of punishment: deterrence in its threat, retribution in its execution,” Law andPhilosophy, 8, no. 2 (1989), 151–200; Thomas E. Hill, “Kant on punishment: a coherentmix of deterrence and retribution,” Annual Review of Law and Ethics, 5 (1997), 291–314;Sarah Holtman, “Toward social reform: Kant’s penal theory reinterpreted,” Utilitas,9 (1997), 3–21; Otfried Hoffe, “Vom Straf- und Begnadigungsrecht,” in Hoffe (ed.),Metaphysische Anfangsgrunde der Rechtslehre, pp. 213–33.

44

of right than his own conception of punishment. In so doing, I willconsciously and consistently employ Kant’s principle of right as acriterion for judging the correctness of that principle of punishment.The “Introduction to the doctrine of right” formulates the principle ofright in the following manner in } B: “Right is therefore the sum of theconditions under which the choice of one can be united with thechoice of another in accordance with a universal law of freedom.”2

Upon this foundation, I will criticize Kant’s theory of punishment, andattempt to construct one anew.

2.1. The interpretation of Kant’s penaltheory as a mixed theory

For a long time the philosophical debate about penal theory wasdominated by the dichotomy between retribution, exemplarily illus-trated by the Kantian school, and general deterrence. By the latterterm, I mean deterring the commission of future crimes by deterringnot only the criminal but also other citizens. Recently, however, thesituation has radically changed: the discussion is now dominated bymixed theories, which conceive of the theories of retributivism and ofdeterrence as reciprocally restricting and complementing each other.

On the one hand, general deterrence theorists clearly profess culp-ability as being a prerequisite to every punishment and endorse somekind of proportionality between a crime and its punishment. On theother hand, retributivists generally agree with the deterrence theoristson the observation that although the principle of retribution in itselfindeed forms a moral principle it does not qualify as a principle ofright in the cases in which punishment demanded by it obviouslywould not deter from the commission of any crime.

Each of these positions complements the other, because the elem-ent that is stronger in one is weaker in the other, and vice versa. Thegeneral deterrence theories remedy a weakness in the principle ofretribution by providing a rationale for punishment that clearly con-nects it not with a private morality, but instead with a system of publiclaw. Retributivism improves the theory of deterrence by furnishing aprinciple that gives a simple answer to the question of proportionalityof punishment and seems to be unobjectionable with relation tojustice toward individuals. Sharon Byrd’s important article about


kant’s legal justification of punishment 45

Kant’s conception of penal law3 brilliantly contrasted this combin-ation with the traditional view of Kant’s penal theory according towhich it is through and through a retributivist penal theory. Byrdargues that legal systems must be secured against the inclination ofcitizens toward breach of the law and that they obtain this guaranteethrough threatening punishment. Hence, according to Byrd, deter-rence constitutes the fulfillment of the public right to coerce. But oncethe aim of punishment is determined in that way, the execution – thatis, the type and the degree – of the punishment no longer followsthe principle of deterrence, but instead the principle of retribution.The main reason Byrd proposes for this shift from deterrence toretribution is that only the latter treats human beings not merely asmeans but also as ends in themselves. Thus, Byrd points out, in Kant’sDoctrine of right deterrence is found in the threat of executing thepunishment, whereas the retributive principle is found in the actualexecution of the punishment, so that deterrence and retributionreciprocally “limit” each other.

Even though Byrd is of the opinion that threatening and retribu-tion reciprocally limit each other in this relation, there is, however, anobvious asymmetry between them, which, in my opinion, privilegesretribution and explains the appeal of Kant’s retributivism in recentyears. Indeed, even though in Byrd’s reconstruction general deter-rence constitutes the aim of punishment, it is retribution that deter-mines the “amount” of punishment. I would like to express this in thefollowing manner. Citizens should be punished for their crimes if andonly if:

1 the threat of punishment could deter them (deterrence condition);and

2 the punishment punishes the crime (retributivist condition); and3 the amount of punishment is determined by the retributive

principle (I will define this principle more precisely below).

The first two conditions are purely negative prerequisites: ifthey are not satisfied then both the punishment and the threat of itare prohibited. They do not, however, positively determine the typeand amount of punishment. For clarification of this point I wouldlike to look at the strongest example upon which Byrd grounds hernew interpretation of Kant, that is, Kant’s expounding on “right of

3. Byrd, “Kant’s theory of punishment.”


necessity (Ius necessitatis)” by way of what Kant terms “the plank ofCarneades”:

There can be no penal law that would assign the death penalty tosomeone in a shipwreck who, in order to save his own life, shovesanother, whose life is equally in danger, off a plank on which he hadsaved himself. For the punishment threatened by the law could not begreater than the loss of his own life. A penal law of this sort could nothave the effect intended, since a threat of an ill that is still uncertain(death by a judicial verdict) cannot outweigh the fear of an ill that iscertain (drowning), but only unpunishable (impunibile).4

Kant adopts two premises. The first premise states that the penal law“would assign the death penalty” (retributive determination of thedegree of punishment, cf. the third point above). The second premiseis formulated indirectly, that is to say, negatively: “A penal law of thissort could not have the effect intended.” The logical opposite to this isour requirement that penal law prevent some crimes, which meansthat the threat cannot be proven to have no effect. The logical oppos-ite does not mean that the effect of deterrence must be either provenor certain or maximal. This prerequisite is so weak that I have to askmyself if there is any imaginable punishment that could ever fail tosatisfy it. Is there truly no one who would prefer a certain drowning toa possible death by a judicial verdict? Provided that the risk of beingcaught (for example because there were witnesses to the event or forsome other reason) is not negligible: is there really no person whoprefers drowning and being mourned because of this tragic accidentover the public dishonor of a judicial sentence possibly or probablyleading to death? To stress this point, let us assume that the court willprobably accept mitigating circumstances and sentence the criminalto a prison sentence of twenty years, so that the person who has thechoice between murder and drowning could be sure that the deathsentence would not be imposed. Can we exclude the possibility thatsome persons might prefer to die tragically rather than face a life inprison without honor and career chances?

If that is so, then the threat of the death penalty (even of a prisonsentence) really does prevent some murders even in such “a case ofnecessity.” I admit that a “not uncertain” death by judicial verdictwould increase the deterrent effect, that is, it would prevent some

4. RL, “Appendix to the Introduction of The doctrine of right,” at Ak vi:235f. Practicalphilosophy, ed. Gregor, pp. 391–2.


murders upon which the threat of an uncertain death or prisonsentence by a judicial verdict would have no effect. At least in somecases, “a threat of an ill that is still uncertain” truly can and will“outweigh the fear of an ill that is certain.” The logical opposite tothe expression “could not have the effect intended” requires only somedeterrent effect, but by no means the maximal deterrent effect. That iswhy there can be a penal law that also imposes a death sentence (or alesser sentence) for murder in such a case of necessity. In moregeneral terms, I do not see any punishment, any imposition of painon guilty persons, of which one can claim that it does not have in oneway or another a deterrent effect (the punishment of innocentpeople, on the contrary, would probably completely fail, becausesuch a punishment would not be related to a crime). Even if deathwas sanctioned with something as ridiculous as either a one-monthperiod in jail or a 10,000 dollar fine, these punishments would stilldeter some potential murderers; therefore, they would still have adeterrent effect.

Because the principle of retribution prescribes a punishment for allcrimes, it seems that the first condition listed above (that the punish-ment must have a deterrent effect) will always be satisfied. Thus, thecondition of deterrence does not in any way limit the principle ofretribution. Nor does the principle of retribution limit the principleof deterrence, because the latter only requires some deterrent effect.However if we (in opposition to Byrd’s interpretation of Kant) were tounderstand the deterrence principle in such a way that it also deter-mined the degree of punishment, then it would aim to maximize thedeterrence effect. In this case, the principle of retribution wouldactually limit the principle of deterrence. In our case, of necessity,the maximization of the deterrence effect would require not only thedeath penalty, but also death after a lengthy and varied torture.

One might object by contending that the prohibition of tortureonly means that a punishment should not surpass a certain threshold,which, for instance, would be established by the prohibition of a cruelor inhumane treatment, though below this threshold the principleof deterrence can still require the adoption of the punishment thatis the most deterrent. This could be an alternative approach, but itis certainly not Kant’s or Byrd’s, because the principle of retributiondoes not allow any latitude to apply any sort of leximin. Rather, theprinciple of retribution allows only one solution, for instance thedeath penalty for murder as in the example given above.


If, in Byrd’s essay, no incompatibility is present between the retribu-tive and the deterrent elements, then it is due only to the weakness ofthe criterion that is adopted for the latter: the punishment must atleast deter some future crimes. The retributive principle satisfies thiscriterion and is thus not really limited by it. Hence, all that this mixedtheory adds to the classical theory of retribution is the condition thatclassical retributivism would not be justified if the prescribed threat ofpunishment would not actually prevent at least one crime. The claimthat the aim of punishment for Kant rests in deterrence shows just oneway in which the case for retributivism could be made stronger. Forthe rest, the mixed theory prescribes the same degree of punishmentas the classical theory of retribution.

In order to express more clearly my point that mixed theory displaysno actual relationship between retributivism and deterrence, I wouldlike to differentiate the possible meanings of a retributive theory.Retributivism can correspond to at least the four following theses:

1 All criminals, and only criminals, should be punished.2 The punishment of criminals serves as retribution for the crimes

committed.3 The degree of punishment should be proportional (ordinally, not

cardinally) to the crime, which means that the relation of the crimesamongst themselves should correspond to the relation of the punish-ments amongst themselves. By that I mean that a serious crimeshould be punished more harshly than a minor crime, and that twoequally serious crimes should bepunishedwith equivalent harshness.

4 The degree of punishment must be equivalent to the crime.

The third thesis compares two classes of relations, that is, the relationbetween different crimes and the relation between different punish-ments. In contrast to that, the fourth thesis directly links a crime toa punishment without consideration of proportionality. The thirdthesis obviously does not imply the fourth thesis, it merely prohibitspunishing a shoplifter more harshly than a murderer.

Now consider the following possibilities. The first possibility is thatthe thief is sentenced to a week of community service, and the mur-derer to twenty years behind bars. The second possibility is that thethief is sentenced to one year in jail, and the murderer is sentenced todeath. The third possibility is that the thief is sentenced to one week ofcommunity service, and the murderer is sentenced to death. All threeexamples comply with the third thesis. However, they are not just


representing very different approaches to penal legislation. The firstpossibility definitely does not comply with the fourth thesis, becausethe murderer is not sentenced to death. The criterion for the fourththesis appears to me to be better fulfilled by the third possibility thanby the second, but I do admit that this might be controversial. I do notwish to discuss here the question whether a different penal law couldbe imagined that complies with the fourth thesis without complyingwith the third thesis, even though this appears to me not to beimpossible. The only reason for my differentiation between the thirdand the fourth thesis is to reject the fourth thesis as being incompat-ible with Kant’s concept of right, even though Kant defends this thesis.Yet, I admit that the third thesis is statistically true (I will explain laterwhat I mean by “statistically”).5

I observe that theses one through four remain untouched by Byrd’smixed theory. I would like to distinguish now the possible thesesconcerning the theories of deterrence.

1 Future crimes are prevented by the punishment of actual criminals(in contrast to the following two theses, this descriptive thesisbelongs to no normative theory of deterrence).

2 Future crimes should be prevented by the punishment of actualcriminals.

3 Citizens should be punished in such a manner as to provide themost effective deterrent to future crimes.

4 Criminals and only criminals should be punished, and only in suchmanner that provides the most effective deterrent to future crimes.

I observe that the third and fourth theses are not supported by mixedtheory. Because the first thesis is a purely descriptive thesis, the onlygeneral deterrent thesis within the mixed theory that remains is thesecond thesis.

Therefore, I will now concentrate on the principle of retribution.As a first step, I will inquire into Kant’s justification and attempt toshow that his justification is clearly insufficient. As the second step,I will show that Kant’s principle of retribution contradicts his principleof right.

5. I take “statistically” to mean that the seriousness of the offense correlates with thedegree of punishment actually carried out, even though there will also be cases wheremore serious crimes are punished more mildly than less grave ones. This will beexplained at the end of this chapter as well as being illustrated in Chapter 7 , usingthe example of punishment of crimes against humanity.


2.2. Retributivism as a justification of punishment

Before we inquire into Kant’s justification of the principle of retribu-tion we will have to distinguish between two meanings of retribution, astronger one and a weaker one: these are Vergeltung (the second retri-butivism thesis, or mere retribution) and Wiedervergeltung (the fourthretributivism thesis, or ius talionis).6 Most of the current Kantian retri-butivists argue for Vergeltung and rejectWiedervergeltung.7

In } 49e of the Doctrine of right, Kant argues in a two-step process:first, he attempts to justify the right to punish in general; secondly, heattempts to justify Wiedervergeltung, or ius talionis, as a principle fordeciding the degree of punishment. In the first step, Kant rejects theutilitarian conception of penal law:

Punishment by a court . . . can never be inflicted merely as a means topromote some other good for the criminal himself or for civil society . . .The law of punishment is a categorical imperative, and woe to him whocrawls through the windings of eudaimonism.8

Most interpreters implicitly derive the principle of retribution (Vergel-tung) from this rebuttal. Yet neither the word retribution (Vergeltung)nor its concept is present before the second step. Even then, onecannot find mere retribution (Vergeltung) but only equality in retribu-tion (Wiedervergeltung). Therefore, I suspect that the interpretation ofthe first step as a justification of mere retribution comes from theinterpretation of the second step, that is, from the justification ofretaliation. Illustratively, the second step contains yet again the samerejection of the utilitarian theory as the first step: according to Kant,all principles except ius talionis “are fluctuating and unsuited for asentence of pure and strict justice because extraneous considerationsare mixed into them.”9 In part five of the Appendix, Kant makes a pleafor rejecting the utilitarian concept of punishment in favor of iustalionis.10

6. The German terms Vergeltung (that is, requiring comparable recompense for a wrong)andWiedervergeltung (that is, demanding wholly equal recompense) correspond roughlyto the English terms retribution and retaliation (etymologically: ius talionis),respectively.

7. Cf. Schied, “Kant’s retributivism”; Jeffrie G. Murphy, “Does Kant have a theory ofpunishment?” Columbia Law Review, 87 (1987), 509–32; and Hoffe, “Vom Straf- undBegnadigungsrecht.”


10. RL Ak vi:363. Practical philosophy, ed. Gregor, pp. 497f.


Kant’s formulation of retaliation, that is, ius talionis, confirms mysuspicion:

Whatever undeserved evil you inflict upon another within the people,that you inflict upon yourself. If you insult him, you insult yourself;if you steal from him, you steal from yourself; if you strike him, youstrike yourself; if you kill him, you kill yourself.11

The first sentence cites an “undeserved evil” and therefore does notbase itself on degree of punishment; thus it cannot concern retali-ation, that is, equality in retribution. Furthermore, whoever wishesto understand the passage as a plea for mere retribution will have toderive this concept from the second sentence. In fact, there are noexamples to be found in the second sentence for the “like-for-like”principle of punishment. For instance, the following interpretation ofthe first sentence would be completely sufficient: if you commit acrime, you throw society back to the state of nature in which you arenot protected against undeserved evils that are committed against you.In this condition, there is no retribution, not even “mere” retribution.Such an interpretation obviously falls short of explaining the secondsentence.

Therefore, the interpretation of the first step as the justification ofretribution is based merely on retaliation. This interpretation, though,bases itself singularly and alone on the rejection of the utilitarianposition in which it is tacitly assumed that there cannot be a thirdalternative to retributivism and the utilitarian theory of deterrence.This assumption appears wrong to me. In order to demonstrate this,I will suggest another solution.

First, however, I would like to make clear that Kant himself allowsexceptions with regard to retaliation. The first exception rests in thealleged right of necessity in the example listed above in Section 2.1:

There can be no penal law that would assign the death penalty to someonein a shipwreck who, in order to save his own life, shoves another, whoselife is equally in danger, off a plank on which he had saved himself.For the punishment threatened by the law could not be greater thanthe loss of his own life . . . Hence the deed of saving one’s life byviolence is not judged inculpable (inculpabile), but only unpunishable(inpunibile).12



This case is supported by neither the second, the third or the fourththesis of retributivism. There is a further exception, if the number ofa murderer’s accomplices

is so great that the state, in order to have no such criminals in it, couldsoon find itself without subjects . . . then the sovereign must alsohave it in his power . . . [to] pronounce a judgment that decreesfor the criminals a sentence other than capital punishment, such asdeportation.13

The first retributivism thesis must therefore be modified to: “Allcriminals and only criminals should be punished by the state, unlesseither the crime occurred in the state of nature [as in the first excep-tion] or the punishment would lead to reversion back to that state.”Or, formulated differently: “All criminals, and only criminals, shouldbe punished, presupposing that the state can ubiquitously enforce thelaw from the time before the deed till after the execution of thepunishment.” This means: “In a stable state, all criminals and onlycriminals should be punished.”

From this we can conclude that for Kant even the weakest retribu-tion thesis, that is, the first thesis – and therefore also the strongestthesis, that is, the fourth retribution thesis (the principle of retali-ation, or Wiedervergeltung) – possesses no absolute validity, but insteadis subject to the realization of his principle of right.

2.3. The ambiguity of the conceptsof retribution and of right

Now I would like to explain why the principle of retribution hasnothing to do with Kant’s principle of right, and furthermore whyeven its application in the Doctrine of right can stand in the way of hisprinciple of right.

It has already been demonstrated how the principle of retaliationsuddenly arises in } 49e, without Kant attempting to deduce it fromhis principle of right. If we search for an argument for this principle inthe Doctrine of right, then we find the following in the “Preliminaryconcepts of the metaphysics of morals (philosophia practica universalis).”If someone does the duty “he can be constrained by law to do”



he does what is owed (debitum) . . . if what he does is less than the lawrequires, it is morally culpable (demeritum). The rightful effect of what isculpable is punishment (poena) . . . conduct in keeping with what is owedhas no rightful effect at all.14

The debt (debitum) results from the criminal not fulfilling what the lawrequires. It is here that the ambiguity of the German term Schuld(“guilt”) comes into play. It can mean either debt (debitum) or culpabil-ity (culpa). Accordingly, retributivism can be schematically described inthe following manner:

1 Whoever makes him- or herself culpable of a crime has not paid hisor her debt to society.

2 Unpaid debts must be paid off.3 Punishment is the paying off of debts.

To take the two meanings of the word Schuld as being equivalent ismisleading and leads to fallacies in the case of crimes that causedamage for which no compensation is possible, that is, for crimesfor which there is no reparation. All infringements that are not irrep-arable are private crimes not falling under the jurisdiction of criminalcourts, but rather under that of civil courts. Kant gives the example of“embezzlement, that is misappropriation of money or goods entrustedfor commerce, and fraud in buying and selling when committed insuch a way that the other could detect it.”15 Such crimes endanger theexistence not of the commonwealth, but only of the individuals whotrusted and freely entered into a contract with the criminal. On theother hand, public crimes do not damage private legal contracts, butinstead the commonwealth itself. For such crimes, as in the case of theexercise of possibly deadly force upon a victim, there is no possiblecompensation. How can, for instance, a prison or death sentence everprovide compensation for the victim of such a crime? What could everprovide compensation for the insecurity of all citizens caused by amurder? The debts (debita) to the commonwealth can only be paid offwhen there is abstention from committing crimes in the future: assoon as a crime occurs, though, the criminal is no longer able to payoff his or her debts. Once the criminal has lost civil personality, thenthis person has lost even the possibility of paying off the debt to thecommunity in the future, that is, to abide by the law in the future.

14. RL Ak vi:227f. Practical philosophy, ed. Gregor, p. 382.15. RL Ak vi:331. Practical philosophy, ed. Gregor, p. 472.


The objection could be raised that there is no real possible com-pensation between the commonwealth and the criminal, but at leastan internal compensation is possible. An internal compensation ofthis sort stands possibly connected with the problem of the “highestgood” – that is, with the “exact correspondence of happiness withmorality.”16 Kant seems to support himself on such an idea when hewrites:

This fitting of punishment to the crime, which can occur only by a judgeimposing the death sentence in accordance with the strict law ofretribution, is shown by the fact that only by this is a sentence of deathpronounced on every criminal in proportion to his inner wickedness.17

There are two kinds of objections that can be raised against such ajustification of retributivism.18 The first and most obvious objection isthat, according to Kant, the state exercises control in a state of laws,that is to say, it regulates the relation between the external freedom ofthe individuals to one another. The highest good is, on the contrary,part of an internal and therefore individual relation.19 Secondly, evenif procurement of the highest good falls under the purview of thestate, then the requirements for the highest good would never justify aretributivist penal theory, but would rather refute it. The highest goodwould require that the relative relation between immorality andunhappiness, or pain (or remorse), should be the same. Actually,Kant defines punishment as “the right a ruler has against a subjectto inflict pain upon him because of his having committed a crime.”20

What should the pain consist of so that it could be comparable withthe amount of immorality? Let us rely on Kant’s definition from theCritique of practical reason:

Happiness is the state of a rational being in the world in the whole ofwhose existence everything goes according to his wish and will, and rests,therefore, on the harmony of nature with his whole end as well as withthe essential determining ground of his will.21

16. KpV Ak v:125. Practical philosophy, ed. Gregor, p. 240.17. RL Ak vi:333. Practical philosophy, ed. Gregor, p. 474.18. For a more detailed refutation of theories of punishment as restoration of harmony,

cf. Jean-Claude Wolf, “Strafe als Widerherstellung eines Gleichgewichts,” Jahrbuch furRecht und Ethik, 11 (2003), 199–216.

19. Cf. Thomas E. Hill, “Kant on wrongdoing, desert, and punishment,” Law andPhilosophy, 18 ( 1999), 407–41 (p. 429); see also below, Chapter 3.

20. RL Ak vi:331. Practical philosophy, ed. Gregor, p. 472.21. KpV Ak v:124. Practical philosophy, ed. Gregor, p. 240.


If pain is the opposite of happiness, then pain must be the state of arational being to whom, in the whole of existence, nothing occursaccording to his or her “wish and will.” Therefore, pain should consistin the whole absence of harmony between the person’s nature and hisor her end, as well as between the end and the “essential determiningground of his will.” Owing to the punishment, the criminal, who isforced to undergo stringent discipline in order to make the “wish andwill” spotless again, does not profit from the criminal deed. Therefore,one must grant that punishment for the purpose of reform bringswith it much pain. Ius talionis also does not make the criminal happy.It does, however, allow that something in the world occurs accordingto the “wish and will” of the criminal. The principle of punishmentwould really be decided according to the criminal’s maxim: ifthe criminal really wanted to kill, then the state will kill the criminal;if the property of another is taken by a thief, then the property of thethief will be taken, etc. Paradoxically, the punishment reaches some-thing like a universalization of the criminal’s maxim. Now, for Kant,the universalization of a maxim poses something of a test for morality.Even the merely partial universalization of an evil maxim as a maximfor action is invariably forbidden. This should be understood to be awarning signal against ius talionis. It is also, coincidentally, the maximupon which a vendetta, that is, a personal revenge or blood-feud,is based.

My arguments are indifferent to the debate as to whether ius talionisis to be taken literally regarding Kant. Everyone knows Hegel’s ironiccommentary against the literal application of ius talionis.22 However,Kant declines punishing sex offenders with rape – he rather demandsthat they be castrated.23 He sentences the bandit to slave labor, butsurely not in order to allow the bandit to fall victim to the sameviolence felt by the victims of banditry.24 Because enough Kant inter-preters have emphasized this point,25 it is not necessary to explain anyfurther why my argument does not have to distinguish between thethird thesis and the fourth thesis, but instead allows me to reject thesetheses along with the second thesis.

The only retribution taking place in the punishment is purely anegative type. Kant writes: “Whoever steals makes the property of

22. Cf. GPhR } 101 Anm.23. RL Ak vi:363. Practical philosophy, ed. Gregor, p. 498.24. RL Ak vi:333. Practical philosophy, ed. Gregor, pp. 474–5.25. For example, Hoffe, “Von Straf- und Begnadigungsrecht,” p. 227.


everyone insecure and therefore deprives himself (by the principle ofretribution [Wiedervergeltung]) of security in any possible property.”26

This means nothing other than the criminal being shut out of thecommonwealth and sent back into the state of nature. Therefore, onlycriminals can and should be punished, under the condition given bythe first modified retributivism thesis27 that a stable state exists. Retri-butivism is correct then only if we understand it in that way. The use ofthe word “retribution” in this case is indeed misleading. In fact, boththe reciprocal limiting of rights in the commonwealth and reciprocalprotection of the citizens of a state could be called reciprocal retribu-tion. Recognition of the rights of another and the protection of themare actually a type of reciprocal contribution or “tribute.” In contrastto that, a crime is merely negative and in no respect a contribution.For this reason, the terms “retribution” and “retaliation” appear to beill suited to describe the modified first thesis.

The simplicity and strictness of ius talionis may be fascinating forsome people. The strictness of a punishment ought not to be mis-taken, however, for strictness regarding a concept. Ius talionis cannotbe deduced from Kant’s concept of right, but rather contradicts thatvery concept; it does not even originate from common sense for it isdisputed in equal measure between experts on penal law and betweencitizens. Above all else, it is possible to reconstruct Kant’s theory ofpunishment directly from his concept of right without ius talionis.

2.4. Exclusion from the commonwealth

In fact, besides the dichotomy between the utilitarian theory of deter-rence and the so-called Kantian theory of retribution, there is a third,authentically Kantian and legal approach.

Neither the prevention of repetition of a crime nor the criminal’sreform, which could be called specific deterrence, is given muchattention in today’s debate about punishment, because both ideasare alleged to replace the punishment’s true meaning with socialconsiderations that have nothing to do with the crime. I regard suchan objection as being completely baseless, and believe that this type ofspecific deterrence could provide rationale for the punishment itself.

26. RL Ak vi:333. Practical philosophy, ed. Gregor, p. 474.27. The modified thesis: “All criminals and only criminals should be punished by the

state, unless either the crime occurred in the state of nature [as in the first exception]or the punishment would lead to reversion back to that state.”


In the following, I will describe the theory of punishment that Kantshould have built upon his principle of right, in order then to criticizethe theory he actually constructed upon it, which, in my opinion,contradicts this principle. I will refer as far as possible, though, toindividual passages in Kant’s Doctrine of right so long as they appear tome to be compatible with his principle of right. I will argue for thefollowing theses:

1 Criminals ought to be punished in order to prevent them fromcarrying out further crimes. This ought to be accomplished bymeasures (2) and (3).

2 Punishment ought to deprive the criminal of the status of a citizenuntil (3) is accomplished.

3 The punishment ought to reeducate the criminal so that the crim-inal’s status as a citizen can be recognized again.

In the following, I will connect the principle of right with the categor-ical imperative in order to reconstruct a Kantian theory of punishment.

Besides the rejection of utilitarianism, Kant names two other argu-ments in his first step (see Section 2.2), two affirmative arguments:first, the criminal must be punished “because of his having committeda crime”; secondly, the penal law is a “categorical imperative.”28 Boththese points contain a descriptive and a normative element: onaccount of the crime committed by the criminal, the said criminalmust be punished for the sake of the categorical imperative. There is,however, in the Doctrine of right no mention of an imperative thatspecially relates to the penal law. In the proper meaning of the term,only one single categorical imperative can exist, for which the Ground-work of the metaphysics of morals offers different formulations. TheDoctrine of right develops an imperative of right as a particularizationof the categorical imperative by factoring in the circumstance that alarge number of people share the same limited world. I propose toconceive of penal law’s “categorical imperative” as being a particular-ization of the categorical imperative of right, under the premise that acrime has been committed.

By definition, because of the crime, a state of nature prevails overthe relationship between the criminal and the rest of the common-wealth. Kant regards the state of law and the state of nature as adichotomy that excludes any third possibility. In fact, Kant terms the



situation after the crime’s occurrence thrice as “state of nature”: in thecase of a murder with many accomplices, in the case of a duel and inthe case of infanticide.29 In an original state of nature before theestablishment of a commonwealth, as well as during periods of regres-sion back to the state of nature, the categorical imperative demandsthe creation of a legal system. Should the establishment of the insti-tutions constitutive of a state of law not be immediately possible, thenpermissive law30 demands achievement of it for everyone – therefore,for the criminal as well – in the quickest way possible. I am arguingthat nothing other than punishment can rehabilitate and make thecriminal a member of the commonwealth again, because one mustconsider that the criminal is not immediately able to reenter thecommonwealth.

By committing a crime the criminal has obviously accepted theregression back to a state of nature. In a state of nature, nobody hasany rights and external freedom is continually threatened, whereas,however, what the categorical imperative demands is that the latter berespected. The criminal accepts the disappearance of his or her exter-nal freedom as well. This means that the will of the criminal commits atype of suicide, which makes this person “unfit to be a citizen.”31 Thecriminal cannot be treated any longer by the state as a free person, as arational being. The type and the degree of punishment that Kantrequires demonstrate this in a clear way. The most famous exampleis surely the death sentence for murder. Hermann Cohen and somemodern interpreters32 have raised a convincing objection against it,saying that this punishment contradicts the Kantian principle ofmorality because the punishment irrevocably destroys a rationalbeing. For this reason they have proposed alternative punishments.Their argument has been so successful that nowadays there is hardly a

29. RL Ak vi:334, 336. Practical philosophy, ed. Gregor, p. 475, pp. 476–7.30. ZeF Ak viii:347f. “Toward perpetual peace,” in Practical philosophy, ed. Gregor, p. 321.

Cf. Reinhard Brandt, “Das Erlaubnisgesetz, oder: Vernunft und Geschichte in KantsRechtslehre,” in R. Brandt (ed.), Rechtsphilosophie der Aufklarung: Symposium Wolfenbuttel1981 (Berlin: De Gruyter, 1982), pp. 233–85.

31. RL Ak vi:331. Practical philosophy, ed. Gregor, p. 472.32. Hermann Cohen, Schriften zur Philosophie und Zeitgeschichte, ed. Albert Gorland and

Ernst Cassirer (Berlin: Akademie Verlag, 1928), p. 341. Cf. Robert A. Pugsley,“A retributivist argument against capital punishment,” Hofstra Law Review, 9 (1981),1501–23 (p. 1516); Steven S. Schwarzschild, “Kantianism and the death penalty,”Archiv fur Rechts- und Sozialphilosophie, 70 (1985), 343–77; Attila Ataner, “Kant oncapital punishment and suicide,” Kant-Studien, 97, no. 4 (2006), 452–82.


philosopher who dares to endorse this part of Kant’s penal law. (Aninteresting exception would be Tom Sorell, who argues in favor ofboth Kantian retributivism and the death penalty.)33

If one accepts the Kantian theory of penal law, as Cohen does, thensuch objections are, however, not convincing. They become so onlywhen one reconstructs an alternative theory of penal law from Kant’sconcept of right, which is explicitly counter to Kant’s theory of penallaw.34 In fact, Kant would deal with these objections in the followingway. First, a rational being ceases to be such as soon as a crime hasbeen committed, and not only after he or she has been punished. Thepunishment only derives the consequence from the factor thatthrough the crime a rational being denies its rational character.Secondly, the degree of punishment that Kant imposes for all othercrimes he mentions shows clearly that he no longer sees the criminalas a rational being. In this vein, I wish to provide the followingexample:

But what does it mean to say, “If you steal from someone, you steal fromyourself”? Whoever steals makes the property of everyone else insecureand therefore deprives himself (by the principle of retribution) ofsecurity in any possible property. He has nothing and can also acquirenothing; but he still wants to live, and this is now possible only if othersprovide for him. But since the state will not provide for him free ofcharge, he must let it have his powers for any kind of work it pleases(in convict or prison labor) and is reduced to the status of a slave fora certain time, or permanently if the state sees fit.35

At this point, Kant fails to differentiate between two issues: (a) thatsomeone without property has to work for his or her livelihood and(b) that the work for one’s own livelihood means enslavement, insteadof, for example, working as a day laborer. Actually, there is a common-ality between this enslavement and the death penalty, just as betweenevery other punishment mentioned in the Doctrine of right, for exampledeportation,36 “permanent expulsion from civil society,” or castra-tion.37 To Kant, castration is a “partial” murder, as is self-castration:“To deprive oneself of an integral part or organ (to maim oneself) . . .

33. Tom Sorell, Moral theory and capital punishment (Oxford: Blackwell, 1987), p. 162.34. Cf. Section 2.5.35. RL Ak vi:333. Practical philosophy, ed. Gregor, p. 474.36. RL Ak vi:334. Practical philosophy, ed. Gregor, p. 475.37. RL Ak vi:363, Appendix 5. Practical philosophy, ed. Gregor, p. 498.


are ways of partially murdering oneself.”38 Even if we leave aside thequestion whether suicide is punishable or not, then castration byanother person still remains a type of partial death.

With all of these kinds of punishment, the criminal is neither seenas being a constituent of the commonwealth nor treated as beingsuch. Kant himself speaks against slavery, just as he does againstvoluntary slavery as well as against selling one’s own children oragainst inherited slavery. Yet he makes a single exception expresslyagainst a person who has “forfeited his personality by a crime.”39

The exact consequence of this exclusion appears to have beenexpressed most clearly by Fichte: either the criminal is deported likean “outlaw” into the desert, where an even more horrible death awaitsthan the legal execution of a death penalty, or the criminal becomessomeone who anyone may kill, as dangerous animals are killed. Thisextreme conclusion resulting from exclusion from civil society hasonly a heuristic function, however, since just as it is nowhere nearFichte’s last word on the subject, so little shall it be the last word in ourreconstruction of Kant’s theory of punishment.

2.5. Rehabilitation by discipline

Even though the criminal may have lost his or her “personality,” it isforbidden from the Kantian standpoint to treat the criminal as onewould a thing or a (possibly dangerous) animal, for the followingreasons.

The most important provided by Kant in } 49e of the Doctrine of rightseems to rely on the following formulation of the categorical impera-tive: “So act that you use humanity, whether in your own person or in theperson of any other, always at the same time as an end, never merely as ameans.”40 At first glance, the Doctrine of right seems to apply thisformulation to the instance of the criminal:

For a human being can never be treated merely as a means to thepurposes of the other or be put among the objects of rights to things:his innate personality protects him from this, even though he can becondemned to lose his civil personality.41

38. TL Ak vi:423. Practical philosophy, ed. Gregor, p. 547.39. RL Ak vi:283. Practical philosophy, ed. Gregor, p. 431.40. GMS Ak iv:429. Practical philosophy, ed. Gregor, p. 80.41. RL Ak vi:332. Practical philosophy, ed. Gregor, p. 473.


Although these sentences resemble one another, they do not appearto use exactly the same dichotomy: whereas the Groundwork of themetaphysics of morals speaks of “humanity” and “personality,” the Doc-trine of right uses the words “innate personality” and “civil personality.”However, if we inquire into the meaning of the latter dichotomy, theformulation of the Groundwork would then appear relevant forexplaining the sentence from the Doctrine of right.

Kant defines personality in relation to the concept of imputability.In the “Preliminary concepts of the metaphysics of morals (philosophiapractica universalis)” he gives the following definition:

A person is a subject whose actions can be imputed to him. Moral

personality is therefore nothing other than the freedom of a rationalbeing under moral laws (whereas psychological personality is merelythe ability to be conscious of one’s identity in different conditionsof one’s existence). From this it follows that a person is subject tono other laws than those he gives to himself (either alone or at leastalong with others).

A thing is that to which nothing can be imputed. Any object of freechoice which itself lacks freedom is therefore called a thing (rescorporalis).42

Thomas Pogge reads this passage – in my opinion, correctly – asconsisting of two definitions:

By italicizing “moral,” Kant flags that “moral personality” is morespecific than “personhood.” The most plausible specification, whichwould also vindicate the “therefore,” is this: Having moral personalitymeans being a subject whose inner actions are capable of imputation,a subject with (transcendental) freedom of the will . . . Persons inthe wider, weaker sense are then subjects whose external actions canbe imputed to them as expressive of their will, choice, or intentions.43

I shall term the latter a subject with freedom of action, in contrastto the former, which, in addition, possesses freedom of the will.According to Pogge’s view, a moral personality presupposes the other,weaker personality and, hence, freedom of action as well, whichcan be guaranteed only by a legal system. This means that moralpersonality presupposes civil personality. On the other hand, one

42. RL Ak vi:224. Practical philosophy, ed. Gregor, p. 378.43. Thomas Pogge, “Is Kant’s Rechtslehre comprehensive?” Southern Journal of Philosophy, 36

supplement (1997), 161–87 (p. 163).


can construct a legal system that simultaneously ensures freedomof action and neither presupposes nor demands that one dispose offreedom of the will.

Of what does the “innate personality” consist to which Kant refers?Since he separates it from the civil personality, it can concern eitherthe moral personality or a third type of personality. But since themoral personality presupposes the civil personality, but the “innatepersonality” is opposed to the civil personality, the “innate personal-ity” cannot then be the moral personality. From the term “innatepersonality” we can derive that this personality is original (that is,not acquired) and inalienable. From the term “personality” we candraw that it is somehow dependent on or relates to a type of imput-ability. Since Kant, while distinguishing persons from things, identifies“imputation” with “free choice,” that is, identifies it as being freedomof action, one could conclude that the innate personality, as well asthe civil and indirectly the moral personality, should be guaranteed bya legal system. Actually, the innate personality receives the protectionof penal law. However, the innate personality is not the same as thecivil personality, and one can be the bearer of the former withoutbeing the bearer of the latter.

I now turn to the reason why the “innate personality” according to} 49e of the Doctrine of right cannot simply be treated as a mere means.According to the Groundwork of the metaphysics of morals,44 the reasonconsists in the humanity in either my person or the person of some-one else. In this context, the concept of “humanity” can nowbe understood in a misleading way, that is to say, by linking it to adichotomy that only plays a role where the differentiation betweenmorality and right is at stake.

Now the human being as a natural being that has reason (homophaenomenon) can be determined by his reason, as a cause, to actions inthe sensible world, and so far the concept of obligation does not come intoconsideration. But the same human being thought [of] in terms of hispersonality, that is, as a being endowed with inner freedom (homo noumenon),is regarded as a being that can be put under obligation and, indeed,under obligation to himself (to the humanity in his own person).45

In thedoctrine of duties a humanbeing can and should be representedin terms of his capacity for freedom, which is wholly supersensible, and

44. GMS Ak iv:429. Practical philosophy, ed. Gregor, p. 80.45. TL Ak vi:418. Practical philosophy, ed. Gregor, p. 544.


so too merely in terms of his humanity, his personality independentof physical attributes (homo noumenon), as distinguished from thesame subject represented as affected by physical attributes, a human

being (homo phaenomenon).46

In both of these citations, “humanity” clearly means moral person-ality or freedom of the will. Understanding “humanity” in the formu-lation of the categorical imperative in this way would then make itsstatement redundant: “So act that you use [moral personality],whether in your own person or in the person of any other, always atthe same time as an end, never merely as a means.”47 The Groundworkconfirms this reading:

A human being, however, is not a thing and hence not something thatcan be used merely as a means, but must in all his actions always beregarded as an end in itself. I cannot, therefore, dispose of a humanbeing in my own person by maiming, damaging or killing him.48

In the specific context of the Doctrine of right, the following readingprevails:

Every human being has a legitimate claim to respect from his fellowhuman beings and is in turn bound to respect every other. Humanityitself is a dignity; for a human being cannot be used merely as a meansby any human being (either by others or even by himself) but mustalways be used at the same time as an end. It is just in this that his dignity(personality) consists, by which he raises himself above all other beingsin the world that are not human beings and yet can be used, and so overall things.49

Forbidding that human beings be “used” can be read in two differentways. Seen from the standpoint of the categorical imperative, it issurely justified, since absolutely every human is able to develop auton-omy, that is, free will. From the standpoint of the legal system, itmeans that this system in no way protects only those citizens whoabide by it and limit their freedom of action in such a way as to becompatible with the freedom of action of every other citizen. Rather,the legal system protects all humans, even those who are actuallyunable to abide by it, for example children and criminals. By analogy

46. RL Ak vi:239. Practical philosophy, ed. Gregor, p. 395.47. GMS Ak iv:429. Practical philosophy, ed. Gregor, p. 80.48. GMS Ak iv:429. Practical philosophy, ed. Gregor, p. 80.49. TL Ak iv:462. Practical philosophy, ed. Gregor, p. 579.


to the manner in which Kant deals with children, I propose thatcriminals are differentiated from citizens by the use of the term“citizens of the world.” In fact, Kant writes about parents:

They cannot destroy their child as if he were something they had made

(since a being endowed with freedom cannot be a product of this kind)or as if he were their property, nor can they even just abandon him tochance, since they have brought not merely a worldly being but a citizenof the world into a condition which cannot now be indifferent to themeven just according to concepts of right.50

Neither children nor criminals are in a position to abide by the law,that is, to limit their freedom of action by the freedom of action ofothers. Yet both of them are capable of acting; therefore, they areendowed with freedom, though at the same time lacking the ability todevelop, in a way compatible with the freedom of all the other persons(borrowing Pogge’s phrasing), “their will, choice, or intentions.”Therefore, the legal system cannot allow them freedom of action,though it must at the same time provide the possibility of obtainingthis freedom, as soon as and for as long as they are in the position torespect the freedom of others. In my opinion, that is the meaning ofthe term “citizen of the world” in the aforementioned passage.

At the very least, I see two important confirmations of my interpret-ation. First, since Idea for a universal history with a cosmopolitan purpose(1784), Kant has assumed, in human beings, a “natural disposition”(Naturanlage) for developing reason within a historical process till a“perfectly just civil constitution” is established.51 In this context,“reason” is not to be conflated with the transcendental meaning ofautonomy or freedom of the will, but instead to be seen as being“a faculty which enables the creature to extend far beyond the limitsof natural instinct the rules and intentions it follows in using itsvarious powers, and the range of its projects is unbounded.”52 Thus,according to Kant, the innate character belonging to all agentsrequires their own development toward the establishment of andmembership in a commonwealth, which is a notion that finds furtherdevelopment in the Doctrine of right.Other writings – first and foremostToward perpetual peace – contain this perspective.

50. RL Ak vi:281. Practical philosophy, ed. Gregor, p. 430.51. Idee, Proposition 5, Ak viii:22. Political writings, ed. Reiss, p. 46.52. Idee, Proposition 2, Ak viii:18. Political writings, ed. Reiss, p. 42.


Secondly, even criminals sentenced to slavery are expressly notsubjected to regulation by the “right of things” (or the law of prop-erty). For Kant, the difference between a slave and any animal consistsof the slave being able to fulfill duty, even when he or she possesses norights whatsoever. In the “Division of the metaphysics of morals as awhole,” Kant defines the legal status of things and animals as being“the relation in terms of rights of human beings toward beings thathave neither rights nor duties.” Kant says, with regard to “the relationin terms of rights of human beings toward beings that have only dutiesbut no rights,” that it is missing (“vacat”) in the doctrine of right “forthese things would be human beings without personality (serfs,slaves).”53 Though these human beings lack status as citizens, theycannot, however, be either used for “shameful purposes” or “depose[d] of . . . life and members.”54

One could object to such an interpretation on the grounds that it issupported by metaphysical assumptions. Indeed, Kant obviouslymakes metaphysical assumptions in the Doctrine of right, for examplerelating to child-rearing: “the offspring is a person, and it is impossibleto form a concept of the production of a being endowed with freedomthrough a physical operation.”55 Yet one need not accept suchassumptions to recognize agents as potential citizens or as actualcitizens, as soon as and for as long as they are able to respect therights of others for whatever reasons they choose, be it due to theobservance of the categorical imperative or be it simply due to self-interest for having their own rights recognized by the commonwealth.Every “citizen of the world” has a claim on the state maintaining thepossibility for him or her to become an actual citizen of the said state.Moreover, the establishment of a commonwealth amongst all humanbeings is a postulate or a prerequisite. It is a prerequisite not only fromthe standpoint of the categorical imperative, but also when we acceptwith Pogge’s “one-way thesis” that as soon as we accept Kant’s categor-ical imperative we also have to accept his system of right, though notthe other way around. Actually, Kant’s doctrine of right is not onlydescriptive, but also normative – it clearly contains the criterion ofuniversalizability and applies it to all subjects of external freedom, forexample agents. Therefore, Kant’s system of right does not create

53. RL Ak vi:241. Practical philosophy, ed. Gregor, p. 396.54. RL Ak vi:330. Practical philosophy, ed. Gregor, p. 471.55. RL Ak vi:280. Practical philosophy, ed. Gregor, p. 429.


agency, but instead presupposes it, and ought to incorporate as muchas possible the principle of the reciprocal restriction of freedom. Inmy opinion, this requires that the legal system promote and developthe ability to follow the law.

In fact, Kant explicitly deals with the development of this ability inchildren and in “savages.” Now, there is no difference between crim-inals and children in respect to their legal status. Therefore, I will viewthe treatment of criminals as being analogous to Kant’s treatment ofchildren.

In Pedagogy, Kant conceives of child-rearing as a twofold task:“discipline” and “culture.”56 This goal is to be followed regardingchildren as well as “savages.” Kant’s concept of right consists in thereciprocal coercion of citizens guaranteed through public coercionexercised by the state. In contrast thereto, the criminal is like a childand a “savage” and an object of unilateral coercion, which is termedby Kant in Pedagogy “discipline” or “breeding.” “Discipline submits thehuman being to the laws of humanity and begins to let him feel thecoercion of the laws.” Kant conceives of discipline as being a pre-requisite for culture, that is, for the positive part of education. Respectfor the law must first be acquired through an external force appliedagainst human beings, before the ability is gained to obey the lawvoluntarily and – most importantly – to be one’s own legislator, whichconstitutes freedom under law, which is equivalent to the civil state.When we coerce a dog to obey a rule what we really want is the dog toobey our rule. With a human being, however, we ought to try topromote humanity as being the goal, that is, encourage externalfreedom under the law. Punishment ought to promote the criminal’sreform.

Kant considers the possibility of a time-limit on the punishment.The thief “is reduced to the status of a slave for a certain time, orpermanently if the state sees fit.”57 Unfortunately, however, this time-limit is found nowhere else in the Doctrine of right. The reason for thismight be that Kant assumes that the education of “savages” is muchmore difficult than educating children. He explains: “But the humanhas from nature such a large tendency to freedom that if he has grownused to it for a while he sacrifices everything for it.”58 However

56. Pad, Ak ix:449. (Translation mine)57. RL Ak vi:333. Practical philosophy, ed. Gregor, p. 474.58. Pad, Ak ix:442. (Translation mine)


nowhere does Kant explicitly exclude for a criminal the possibility ofrehabilitation. In Kant’s Doctrine of right, a murderer does not alwayshave to die. The cases listed above (see Section 2.2) are not the soleexceptions: not only in the case in which a murderer has too manyaccomplices, for example, but also in the case in which a sovereignexercises “his right of majesty” to grant “clemency . . . only in individ-ual cases.”59 Therefore, the murderer can eventually be rehabilitatedand released, unless his or her release would threaten the security ofthe other citizens, i.e. unless he or she is not yet “disciplined.”

Now, the possibility of rehabilitation is completely incompatiblewith the death penalty. The alternative to retributivism that I ampresenting here is in no way utilitarian. It absolutely does not viewpunishment as being “merely a means to promote some other goodfor the criminal himself or for civil society.”60 Admittedly, the mur-derer certainly profits more from his or her rehabilitation than fromthe death penalty. Yet the rehabilitation model that I have sketchedout here supports itself solely on the goal of restoring the civil statedestroyed by the crime in such a way that the criminal can be reintro-duced into that very state. The only available means to this endis punishment with unilateral coercion. During the time when the“discipline” is being carried out, the commonwealth is protected fromthe risk of the criminal repeating the crime. In the model of punish-ment I have sketched out, punishment therefore carries out an inten-tion that is not only inherent in the concept of right but also inherentin the goal of preserving the humanity in every person. I do concedethat the theory of general deterrence also follows a goal inherent inthe concept of right, in so far as the punishment motivates the othercitizens to respect the law; however, a theory that requires deterringthe rest of the population from committing a crime does not reallytake the goal of preserving the humanity in every person seriously. Inorder to maximize the deterrent effect, the theory of general deter-rence could end up increasing the punishment in duration and indegree beyond what is necessary to rehabilitate the criminal back intoa full-fledged citizen.

One may object to specific deterrence, which I propose as beingan alternative to the principle of retribution, on the grounds that itinfringes the fairness principle, according to which a punishment is



not allowed to be lengthened or formulated anew once the criminalhas already been sentenced. Such an objection would, however, relyon the false assumption that the punishment’s unrevisability impliesthat it is set to an exact duration, that is, to a sentence of a certainamount of either months or years. But the duration is clearly definedin my proposal: the punishment ends with the criminal’s rehabili-tation, that is, with the reestablishment of his or her ability to abideby the law like other citizens, so that the criminal poses no more ofa threat to the law than they do. A weightier objection might lie inthe difficulty in determining exactly when the criminal is able toreenter the commonwealth as a full-fledged citizen. In order to assesswhether a criminal is rehabilitated one certainly needs, in most cases,a complex decision in which there always exists a danger of error. Yetthis danger is also present in every court sentence. Moreover, today’sjudicial system does actually apply to a certain extent specific deter-rence: most prisoners are released on parole, which reduces theeffective term of imprisonment, and recidivists are often punishedmore harshly than first-time offenders, and so on.

Let us imagine for a moment how a mechanism might look withwhich one could eliminate the risk of arbitrariness in assessing thedegree of rehabilitation. One could imagine an institutional systemthat would gradually remove the burden of proof from the criminaland place it on competent judges. After a while, the burden of proofwould shift and it would become increasingly more difficult for theappropriate judges to prove that the punishment should continue.Aside from that, such measures as parole, community service andmore severe sentences against recidivists could constitute a sort ofsecond “test.” The second test would be found in “real” life and wouldconstitute at the same time both a threat to and an incentive for theprisoner to reintegrate sincerely into society and respect the legalsystem. Finally, the punishment could also include specific measuresof specific deterrence. The judge Marianna Pfaelzer not only sen-tenced the hacker Kevin Mitnick to exactly four years in jail, but alsoforbade him from using computers and mobile telephones withoutwritten approval of his probation officer.

Certainly specific deterrence and rehabilitation – the latter mean-ing strict discipline – prevent numerous crimes from being committedby other citizens. Therefore, both theories always include a certaindimension of general deterrence, even though they do not targetits maximization. Both specific deterrence and rehabilitation, in


accordance with Kant’s definition of punishment, “inflict pain upon”criminals61 and threatening pain for a breach of the law certainly is animportant component of deterring potential criminals from crimes.Yet general deterrence does not become the rationale for punishmentsince we would still punish criminals even when not a single othercitizen would be deterred from committing a crime by the punish-ment. Nonetheless, the effect of general deterrence caused by specificdeterrence certainly does constitute a strong and very fortunate sideeffect of the punishment. I would, however, like to make it clear thatthe concept of punishment as being rehabilitation is in no way expres-sive of “overly compassionate feelings of an affected humanity,” forwhich reason Kant reproaches Beccaria’s theory of punishment.62

Discipline, “breeding,” enslavement and unilateral coercion certainlyappear to criminals as being neither attractive nor enjoyable ways tobe punished.

Specific deterrence and rehabilitation may also frequently coincidewith the proportionality of punishment required by the retributivistthesis formulated above on p. 49. Although one might not be able todetermine in principle that a thief probably might be rehabilitatedmore quickly than a hostage-taker or even a murderer, one can do itempirically. However, there are certainly some cases in which just theopposite may be true.

Last but not least, retributivism and general deterrence do notexclude the death penalty. In fact, Kant even explicitly prescribes iteven though many Kantians either deplore this or maintain a carefulsilence on the subject. For a long time, there was a debate about theusefulness of the death penalty as a means of general deterrence.63

Yet all the participants in this debate agreed that it would be justified ifit could be proven that the death penalty was effective in preventingcrimes by other citizens. In opposition to both retributivism andgeneral deterrence, my alternative radically excludes the death pen-alty. It also excludes sentences condemning the criminal to life inprison without the opportunity for a reduction, although it does notforbid keeping a criminal behind bars for life if all attempts atrehabilitation should fail. The solution that I have proposed appears

61. RL Ak vi:331. Practical philosophy, ed. Gregor, p. 472.62. RL Ak vi:334f. Practical philosophy, ed. Gregor, p. 475.63. A classic – and skeptical – representation of the debate is provided in H. L. A. Hart,

Punishment and responsibility (Oxford: Clarendon Press, 1968).


to me to reach in the best possible way what the twofold requirementof all theories of penal law ought to be: ensuring the safety of allcitizens and respecting the humanity of all criminals.

2.6. Summary

The “mixed theories” of penal law fail in their attempt truly to linkretributivism and general deterrence. They necessarily fail becauseboth theories clearly differentiate themselves from each other asjustification for punishment. I have therefore proposed we undertakea twofold modification of the terms in the debate between deonto-logical ethics and deterrence theory. First, I have attempted to showthat Kant’s retributivist theory of punishment does not draw on hisconcept of right. Secondly, I have consciously avoided attempting tomix the Kantian concept of right with a principle of general deter-rence, but instead have argued that the Kantian concept of right iscompletely compatible with a theory of specific deterrence, whichencompasses both incapacitation and the rehabilitation of criminals.This theory is not a mixed theory, and nor does it have to be one.

Therefore, if Kant’s theory of punishment is justifiable on the basisof the Kantian system, then this justification could only occur whileattempting a moral argument from which legal consequences arise. Inthe following chapter, I wish to show that Kant’s moral theory cannotafford this justification. Quite to the contrary, it will confirm theend result of this chapter, which had as its point of departure thetheory of right.


3

KANT’S MORALJUSTIF ICATION OF PUNISHMENT

Now that I have rejected Kant’s theory of retribution on the basis ofa liberal interpretation of his concept of right in favor of a combin-ation of specific deterrence and rehabilitation, I would like to investi-gate whether the alternative, stronger or moral interpretation of hisconcept of right can substantiate Kant’s theory of retaliation. As wasshown in the first chapter, this interpretation consists of an imple-mentation of the categorical imperative in the legal system so far asthe commands and prohibitions of the categorical imperative are ableto be implemented by force.

3.1. Which type of moral proportionality is at issue?

The current retributivist theories of punishment often radically differfrom Kant’s own theory because they assume that punishment isjustified by a goal, however modest this goal may be. The positionheld by expressivists, for instance, states that punishment should beseen retributively so that society’s moral judgment about the punish-able action can be expressed.1 Another theory of punishment, such asthe one held by Jean Hampton and Jeffrie Murphy, regards thepunishable action as comprised of the criminal putting his or hervalue over the worth of the victim.2 The retributive punishment thatis then stipulated is a punishment that restores the proper relationbetween the worth of the victim and the worth of the criminal.

1. Cf. Thomas E. Hill, “Kant on punishment: a coherent mix of deterrence and retribution,”Annual Review of Law and Ethics, 5 (1997), 291–314 (pp. 320–1).

2. Cf. Jean Hampton and Jeffrie Murphy, Forgiveness and mercy (Cambridge: CambridgeUniversity Press, 1988), pp. 45–53.

72

In opposition to such theories as well as similar ones, and alsoagainst the legal tradition following Seneca’s dictum (as I have men-tioned before), “nemo prudens punit quia peccatum est sed ne pec-cetur,”3 is Kant’s view, according to which the punishment should notbe allowed to have a goal (Zweck) but rather should be inflicted only“because of his having committed a crime.”4 Kant regards this require-ment as a moral justification of punishment, by which he defines thecriminal’s worthiness of punishment. In the following, I would like toconduct an interpretation of Kant’s moral retributivism that satisfiesthis requirement. I shall attempt to show that although one can derivefrom Kant’s moral theory a retributivist conception of a criminaldeserving of punishment, one can, however, justify not a retributivistpunishment in the legal system, but rather that combination of spe-cific deterrence and rehabilitation that already ensues from the liberalconcept of right (see Section 2.5).

At this point, I would like to introduce a classification through anadditional differentiation, in which I will enumerate several possibleretributive theses. H. L. A. Hart divides the retributive position intothree theses:

First, that a person may be punished if, and only if, he has voluntarilydone something morally wrong; secondly, that his punishment must insome way match, or be equivalent of, the wickedness of his offence; andthirdly, that the justification for punishing men under such conditionsis that the return of suffering for moral evil voluntarily done, is itself justor morally good.5

3. “A sensible person does not punish a man because he has sinned, but in order to keephim from sin.” Seneca, De ira, i.19.7. Seneca uses as a basis Plato’s Laws 11.933e–934b,where it is said that every thief should not only provide compensation for the item thatwas taken, but also suffer a more lenient or a severer punishment depending on hismotives: “This additional penalty is to be inflicted not because of the crime (what’sdone can’t be undone), but for the sake of the future: we hope that the offenderhimself and those that observe his punishment will either be brought to loathe injusticeunreservedly or at any rate recover appreciably from this disastrous disease.” Plato,Laws, trans. Trevor J. Saunders, in John M. Cooper and D. S. Hutchinson (eds.), Plato:Complete works (Indianapolis: Hackett, 1997), pp. 1318–1616 (pp. 1585–6). Therefore,Plato’s – as well as Seneca’s – intention is clearly directed, not only toward thecompensation of the victim, but toward the reform of the criminal as well as towardgeneral deterrence, which is conceived of as a collective reform. Seneca, De ira, inMoralessays, ed. John W. Basore (3 vols., London: Heinemann/Cambridge, Mass.: HarvardUniversity Press, 1928), vol. 1, pp. 106–355 (pp. 158–9).

4. RL Ak iv:331. Practical philosophy, ed. Gregor, p. 472.5. H. L. A. Hart, Punishment and responsibility (Oxford: Clarendon Press, 1968), p. 231.

kant’s moral justification of punishment 73

The first thesis does not constitute a specific characteristic of retribu-tivism, because, in my opinion, no alternative theory of criminaljustice ever allowed the punishment of the innocent.6 The third thesiscan correspond to as many as the four following subtheses:

1 All criminals – and only criminals – should be punished.2 Thepunishment of criminals acts as retribution for a past criminal act.3 Thedegreeofpunishment shouldbeproportional (inanordinal,not

in a cardinal way) to the crime, that is, the proportion of punishmentsto one another should correspond to the proportion of the crimes toone another.

4 The degree of punishment must be equal to the crime.

The third thesis connects a legal descriptive observation (the descrip-tive observation of a punishable action) to a legal consequence (thepunishment of an action). In contrast, the second thesis connects alegal consequence (the punishment of an action) to a descriptivemoral observation (the wickedness of the perpetrator). Thus, retribu-tivism’s moral justification pertains to the second thesis; however, itcan support or contradict the third thesis. In what follows, I will showthat it contradicts the third thesis. Even the point of reference islacking for proportionality between the moral wickedness and thepunishment. Kant distinguishes, in fact, actions in conformity withduty from actions contrary to duty, and with the latter he distinguishesactions that are merely in conformity with duty from actions that aredone out of it. However, neither a gradation of how much an actionstands in conformity with duty nor a gradation of contrariness to dutycan be found in Kant’s works. Kant merely states:

If someone does more in the way of duty than he can be constrained bylaw to do, what he does is meritorious (meritum); if what he does is justexactly what the law requires, he does what is owed (debitum); finally if whathe does is less than the law requires, it is morally culpable (demeritum).7

Virtue ¼ þ a is opposed to negative lack of virtue (moral weakness) ¼ 0as its logical opposite (contradictorie oppositum); but it is opposed to vice¼ – aas its real opposite (contrarie s. realiter oppositum).8

6. Cf. Fred Rosen, “Utilitarianism and the punishment of the innocent: the origins of afalse doctrine,” Utilitas, 9, no. 1 (March 1997), 23–37.

7. RL Ak vi:227. Practical philosophy, ed. Gregor, p. 382.8. TL Ak vi:384. Practical philosophy, ed. Gregor, p. 516. (There is a small error in the

translation that I have corrected.)


Indeed, the gravity of the punishable action should not be confusedwith the correspondence between the intention and the duty.Whereas a punishable action is more or less serious, any injury of dutycounts as being contrary to duty, and Kant does not differentiate anyfurther than that. Kant even emphasizes that an exception from themoral law or from the fulfillment of duty should never be allowed tobe trivialized, but should be considered as really representing a com-plete misconduct.9 The individual instances of misconduct can betraced back to a supreme maxim, which does not obey the categoricalimperative. Because the assertion of a gradation of evil is problematicwith Kant, I will then, in what follows, confine Kant’s retributive moraltheory to the assertion stating that the punishment is not imposed inorder to fulfill an aim, but instead merely because the criminal didinjury to the law.10

3.2. The highest good and the necessaryrelationship between guilt and punishment

The requirement for a relationship between the moral evil of theindividual who acts contrary to duty and his or her punishment isderived from the postulate of the highest good as it is formulated inthe Critique of practical reason. It is generally known that this postulaterequires a necessary connection between the two components of theidea of the highest good, that is, between virtue and happiness (that is,between Tugend and Gluckseligkeit); however, Kant further refines thisconnection into a subordination of happiness beneath virtue, whichserves as a prerequisite for happiness. Kant writes:

Now, inasmuch as virtue and happiness together constitute possessionof the highest good in a person, and happiness distributed in exactproportion to morality . . . constitutes the highest good of a possible world,the latter means the whole, the complete good, in which, however,virtue as the condition is always the highest good, since it has nofurther condition above it, whereas happiness is something that,though always pleasant to the possessor of it, is not of itself absolutelyand in all respects good but always presupposes morally lawful conductas its condition.11

9. Cf. TL Ak iv:424. Practical philosophy, ed. Gregor, p. 76.10. For example, Hugo Bedau, in “Retribution and the theory of punishment,” Journal of

Philosophy, 75, no. 11 (1978), 601–20, shows that it is also problematic in general.11. KpV Ak v:110f. Practical philosophy, ed. Gregor, p. 229.


The highest good means postulating the reign of the principle of rightover the physical world, which usually abides by the laws of physics.The highest good can only be realized under the reign of theprinciple of right. Retaliation against evil, for the time being, playsno role in this; instead, it is only the reward to the moral dispositionthat plays a role. How can the relation between having wicked dispos-itions and the worthiness of punishment be derived using such adefinition of the highest good?

Next, we must refer to the status and the substance of the highestgood, whose negative variant is the worthiness to be punished. Being apostulate of the practical reason, the highest good has the status of anoumenal, synthetic and necessary connection between virtue andhappiness; in the sensible world this connection – wherever it mighthappen to be present – can only incidentally come into existence.12

Kant even finds that,

it must seem strange, that philosophers both of the ancient and moderntimes could nevertheless have found happiness in precise proportion tovirtue already in this life (in the sensible world), or [let themselves be]persuaded that they were conscious of it.13

If the relation between guilt and punishment were to show the inverseof the highest good, and therefore have the same status as the highestgood, then the relation must belong exclusively to the noumenalworld. In the Doctrine of virtue, we find an acknowledgment of this veryassumption:

Punishment is not an act that the injured party can undertake on hisown private authority but rather an act of a court distinct from him,which gives effect to the law of a supreme authority over all those subject toit; and when (as we must in ethics) we regard human beings as in arightful condition but in accordance only with the laws of reason (not civillaws), then no one is authorized to inflict punishment and to avenge thewrongs sustained by them except him who is also the supreme principleof right giver; and he alone (namely God) can say “Vengeance is mine;I will repay.”14

As already outlined in the Lecture on ethics (written sometime between1775 and 1780), deterrent punishment inheres in the worldly rulerand retributive punishment inheres in the moral ruler:

12. Cf. KpV Ak v:114f. Practical philosophy, ed. Gregor, pp. 231ff.13. KpV Ak v:115. Practical philosophy, ed. Gregor, p. 232.14. TL Ak vi:460. Practical philosophy, ed. Gregor, p. 578.


All punishments are either deterrent or retributive . . . All punishmentsby authority are deterrent, either to deter the transgressor himself, or towarn others by his example. But the punishments of a being whochastises actions in accordance with morality are retributive.15

There are two reasons for the highest good’s noumenal status. First,for us human beings, actions done out of duty and actions that aremerely in conformity with duty can never be discerned from oneanother.16 Secondly, we will never find a worldly ruler of whom wecan be completely sure that he or she will always abide by the categoricalimperative.17

One could argue against assigning this exclusively noumenal statusto the inverse of the highest good, that is, to the relation between guiltand punishment, by saying that with an action that is contrary to thelegal law one is able to suggest – wholly and with absolute certitude –the existence of wicked dispositions; and with that, one dispenses withthe first reason against a retributive punishment imposed by theearthly authorities. The second reason still persists, however, becausethere is no individual to be found, who acts out of duty, to be thehighest judge. In what follows, I would like to show that, even whenthis second reason is disregarded, the earthly authority is still notauthorized to impose retributive punishments, principally for reasonsthat result from the substance of the highest good.

The highest good consists of the necessary connection betweenvirtue and happiness, where virtue is the prerequisite for happiness.Therefore, the converse of happiness should be derivable from theconverse of virtue, that is, fromwicked dispositions. This appears primafacie to suggest that the converse of happiness is unhappiness, in thesense of an unhappy condition. However, this would be a fallacy,because the logical converse of happiness is the lack of happiness as asimple privatio. Now, Kant defines happiness in the following fashion:

Happiness is the state of a rational being in the world in the whole ofwhose existence everything goes according to his wish and will, and rests,therefore, on the harmony of nature with his whole end as well as withthe essential determining ground of his will.18

15. VE Ak xxvii:286. Immanuel Kant, Lectures on ethics, ed. Peter Heath andJ. B. Schneewind, trans. Peter Heath (Cambridge: Cambridge University Press, 1997).

16. GMS Ak iv:407. Practical philosophy, ed. Gregor, p. 62, and VE 43.17. Cf. Idee, Proposition 6, Ak viii:23. Political writings, ed. Reiss, pp. 46–7.18. KpV Ak v:124. Practical philosophy, ed. Gregor, p. 240.


According to this, the converse to happiness is a rational being’scondition in the world for whom nothing in his entire existence “goesaccording to his wish and will,” and therefore rests on the completelack of correspondence between nature and his aim, that is, “theessential determining ground of his will.” In short, the will of thewicked human being should never be able to realize itself. The com-plete lack of effect of the evil will is something other than inflictingsome sort of harm on the evil individual. In the clearest possiblefashion, Kant rejects the latter as immoral and stringently distinguishesit from actual punishment:

It is, therefore, a duty of virtue not only to refrain from repayinganother’s enmity with hatred out of mere revenge but also not even tocall upon the judge of the world for vengeance, partly because a humanbeing has enough guilt of his own to be greatly in need of pardon andpartly, and indeed especially, because no punishment, no matter fromwhom it comes, may be inflicted out of hatred.19

From the Kantian perspective, the punishment should be derivedfrom the crime itself according to a moral law that reigns over theworld. This means a necessity similar to that of the physical laws ofnature. Kant himself uses the analogy:

Before reason awoke, there were no commandments or prohibitions, sothat violations of these were also impossible . . . From the moral point ofview, therefore, the first step beyond this state was a fall; and from thephysical point of view, this fall was a punishment, for it led to a host ofhitherto unknown evils.20

Regarding crimes, this means that the moral legislator should preventtheir perpetration, and do so, in fact, before the implementation of thecriminal intention. Kant also did not ignore the fact that it does notusually happen like that in the real world. However, one is allowed topostulate that the necessary connection between guilt and punish-ment might be compensated in the noumenal world. When one hasfinally determined that, despite this postulate, the intention that iscontrary to duty led to a deed that is contrary to duty, the questionarises as to why the evil intention, whose realization could not beprevented, is allowed to be punished ex post facto. At this point, it is

19. TL Ak vi:460f. Practical philosophy, ed. Gregor, p. 578.20. Conjectures on the beginning of human history, in Immanuel Kant, Political writings, ed.

Hans Reiss, trans. H. B. Nisbet (Cambridge: Cambridge University Press, 1970),pp. 221–34.


appropriate to differentiate between two cases: either the perpetratorremains an evil person after the punishable action or he or sheconverts to the side of the good.

If the perpetrator were to remain an evil person, then he or shewould want further to realize intentions that are contrary to duty. Thenecessary connection must lead to the perpetrator being preventedfrom realizing those intentions. This prevention occurs not for thesake of punishing the specific crime that he or she committed, butinstead because of the evil manifested along with the crime’sperpetration.

According to the Kantian perspective, in the case of the perpetratorconverting to the side of the good since that time, one would have tolook at this case from two different angles. In Kant’s work Religionwithin the boundaries of mere reason, he analyzes the relation of moralconversion with punishment as follows. In the case of conversion, thepunishment takes place within a situation in which “he now leads anew life and has become a ‘new man.’”21 Kant locates punishment“in the situation of conversion itself”: the “conversion is an exit fromevil and an entry into goodness,”22 and he sees that “by means of thevery concept of moral conversion, we can think that situation asentailing such ills as the new human being, whose disposition is good,can regard as having been incurred by himself . . . and, [therefore],as punishment.”23 Kant describes the conversion as a simultaneousexistence between the old and the new human:

As an intellectual determination, however, this conversion is not twomoral acts separated by a temporal interval but is rather a single act,since the abandonment of evil is possible only through the gooddisposition that effects the entrance into goodness, and vice-versa. Thegood principle is present, therefore, just as much in the abandonmentof the evil as in the adoption of the good disposition, and the pain thatby rights accompanies the first derives entirely from the second.24

The punishment of an evil human being consists wholly of the goodhuman being’s birth pains; in short, it is comprised of that throughwhich the person is made into a good human being. This displays,

21. Rel Ak vi:73. Immanuel Kant, Religion within the boundaries of mere reason, in Kant,Religion within the boundaries of mere reason and other writings, ed. AllenWood and Georgedi Giovanni (Cambridge: Cambridge University Press, 1998), pp. 31–192, (p. 89).

22. Rel Ak vi:73f. Religion, p. 90.23. Rel Ak vi:73. Religion, p. 114.24. Rel Ak vi:74. Religion, p. 90.


par excellence, the evil human being’s lack of happiness, because in theconversion he or she breaks with the evil intentions and new, goodintentions lead to happiness, that is, to their full effect. Kant evidentlydescribes punishment in a clear form with similarities to the PaulineEpistles in the New Testament:25

The emergence from the corrupted disposition into the good is in itselfalready sacrifice (as “the death of the old man,” “the crucifying of theflesh”) and entrance into a long train of life’s ills which the new humanbeing undertakes in the disposition of the Son of God, that is, simply forthe sake of good, yet are still fitting punishment for someone else, namelythe old human being (who, morally, is another human being).26

What appears to be “moral happiness” from the new point of view isperceived from the old point of view as ill-being and punishment.27

Here, the appeal of the usual interpretation of retributive punishmentto Kant proves to be false. Thomas Hill sees a motive for future moralbehavior in the moral human being’s pangs of conscience about his orher earlier delinquency.28 Exactly the converse is the case with Kant.

3.3. Worthiness of punishment and rightful punishment

What consequence are we to draw, though, for rightful punishment inthe phenomenal world from this moral punishment in the noumenalworld? It is appropriate, owing to the case of the human being whocontinues to remain an evil human being, to use incapacitation29 as ameans so that his or her evil intentions will not come to fruition in thefuture. In the case of the human being who has converted, there is nolonger any reason for punishment provided that one can discern thatthere has truly been actual reform on the part of the human being.Now, a conversion in the phenomenal world – a conversion which wascompleted in the noumenal world – does not lend itself to beingsubstantiated with absolute certitude. Admittedly, one could increas-ingly conjecture such a conversion in the case of a continued behaviorin conformity with duty, so that, in the course of time, and in view of

25. Rom. 6:2, 6; Gal. 5:24; Eph. 4:17–24; Col. 3:1–17.26. Rel Ak vi:74. Religion, p. 90.27. Cf. footnote in Kant, Rel Ak vi:75. Religion, p. 91.28. Cf. Hill, “Kant on punishment,” 358–60; similar to John Deith, “On the right to be

punished: some doubts,” Ethics, 94 (1984), 191–211 (pp. 210–11).29. Cf. Arthur Ripstein, Equality, responsibility and the law (Cambridge: Cambridge

University Press, 1999), p. 144.


such behavior, reintegration into society becomes more and moreappropriate. In general, specific deterrence – in the form of incapaci-tation – with subsequent rehabilitation is most appropriate, because(at least in a first period of time) one cannot discern if one is dealingwith the first or the second case when it comes to a convicted criminal.

Neither specific deterrence nor rehabilitation is at the disposal ofthe public authorities, but they are their duties. On the one hand, theduty to provide assistance to those in need in itself also encompassesthe duty to provide assistance to any person who may be a futurevictim of crime. On the other hand, however, we are not allowed todo harm or to deny freedom to any human beings – who became goodin the meantime – owing to their previous transgressions, if either thisharm or this denial of freedom is not necessary to public safety. At thispoint, we have reached the same point that our Kantian critique ofKant’s theory of criminal justice derived from the Kantian liberalconcept of right reached.

Different from the conclusion that Kant reached in the sameyear (1797) in his Doctrine of right, in the Doctrine of virtue Kant cameto draw the same conclusion that we just have:

It is, therefore, a duty of human beings to be forgiving (placabilitas). Butthis must not be confused with meek toleration of wrongs (mitis iniuriarumpatientia), renunciation of rigorous means (rigorosa) for preventing therecurrence of wrongs by others; for then a human being would bethrowing away his rights and letting others trample on them, and sowould violate his duty to himself.30

The retributivist might want to celebrate half a victory because of thisresult, because – while it may not be the legal punishment – it ishowever the case that the moral punishment is conceived of retribu-tively. As early as in the Lecture on ethics, Kant defines “Retributivepunishments” as “those pronounced because the evil has occurred.”31

Unlike deterrent punishments, “moral punishments are imposedbecause a sin has been committed; they are consectaria of moral trans-gression.”32 By celebrating half a victory, however, the retributivistoverlooks the fact that Kant, even though he does not say it explicitly,understands “moral punishments” to be the worthiness of punish-ment, whereas he understands the other punishments to be legal

30. TL Ak vi:461. Practical philosophy, ed. Gregor, p. 578.31. VE 79.32. VE 79.


punishments, that is, punishments in the proper sense. Kant com-ments on the legal punishment in the Lectures on ethics, just as he does,at a later point, in the Doctrine of virtue :

Authority punishes, not because a crime has been committed, but sothat it shall not be committed. But every crime, in addition to thispunishment, has a property of deserving to be punished, because it hastaken place. Such punishments, which must therefore necessarily followupon the actions, are moral in character, and are poenae vindictivae; just asa reward follows upon a good action, not so that further good actionsshould be done, but because there has been a good action done.33

Because the worthiness of punishment – which is to say, the moralpunishment – belongs to the noumenal world, the other theories ofpunishment do not compete with this moral retributivism for a simplereason, that is, because the other theories are merely legal theoriesof punishment, which make no claim on the noumenal world. There-fore, the alternative theories do not express themselves against thesecond thesis in the manner in which it was formulated byH. L. A. Hart (see Section 3.1). They simply ignore the subject ofretribution against moral evil and merely controvert the third thesis,which retributivism intends to substantiate with help from the secondthesis. The retributivists want to derive the third thesis from theretributive moral thesis, which we have just examined, as a result ofwhich a necessary proportionality between the punishable action andthe punishment will come to exist. We have, however, just shown withour reconstruction of Kant’s moral argumentation that the thirdthesis by no means follows from the second thesis, but rather that,from it, a combination of specific deterrence and rehabilitation arises.Consequently, the third thesis can only ground itself on itself, as hasbeen argued in this chapter in order to refute it.

As a provisional conclusion, we are allowed to assert that neitherthe liberal (see Chapter 2) nor the moral (see Chapter 3) conceptof right, which one can derive from two different interpretations ofKant (see Chapter 1), can justify a retributive theory of punishment.And there is more: if both paths of argumentation were to pay atten-tion to the consistency of their deduction of punishment from theKantian concept of right they would then lead in our sketch of thereconstruction of the arguments to a combination consisting ofspecific deterrence and of rehabilitation – which strictly excludes

33. VE 79.


retributivism – without arguing for general deterrence. At this point, itfinally comes into full view that the retaliative theory is not the onlyalternative to general deterrence, which is a view that is contrary toboth Kant and many of the authors inspired by him. Fichte and Hegelthen reconstructed what should have been Kant’s theory of punish-ment, if it had been consistent with Kant’s practical philosophy.Although this is a controversial issue, I believe that none of theseauthors was a retributivist.


PART II

PUNISHMENT AS AMEANS OF REHABILITATION

I will proceed with my reconstruction of what should have been Kant’stheory of penal law, had it been derived from Kant’s concept of right,through an interpretation of the Fichtean and the Hegelian theory ofpunishment. Since the degree of punishment that both Fichte andHegel support (incapacitation and resocialization, as we shall see) isnearly the same, interpreting both authors may seem redundant. Thusone may ask oneself whether treating the more famous of the two, whois at the same time the one often considered as being a retributivist,would not suffice. However, the ways in which they come to this sharedconclusion could hardly be more divergent. As we shall see, whereasFichte begins with the unilateral breaking of the social contract by thecriminal and with its consequences (the exclusion from the common-wealth), and then introduces a normative obligation to realize theconcept of right, Hegel starts with the validity and the enforcement oflaw that has to be manifested, without drawing on the figure ofexclusion – on the contrary, the criminal does not cease to belong tothe commonwealth and to enjoy by the same token respect for his orher human dignity. Whereas Fichte focuses on the fate of the criminal,Hegel takes the validity of the law as the basis for his approach.

The fact that both authors eventually adopt the same degree ofpunishment on the basis of the same concept of law in spite of the verydifferent ways of reaching it should make it clear that it is not themethod but the adoption of this shared concept of right that mattersfor reconstructing a Kantian theory of punishment. Nor do the differ-ences between Kant, Hegel and Fichte regarding the foundation oftheir respective systems and of their respective conceptions of auton-omy matter for our investigation. Therefore, I shall deliberately putthem aside in the following argumentation.

85

4

F ICHTE’S “EXPIATION CONTRACT”

It is well known that the liberal interpreters of Kant’s concept of right(see Chapter 1) and Fichte share the same concept of right, eventhough they deduce it in fundamentally different ways:1 the reciprocalrestriction of the external freedoms according to the universal law ofequal rights. Nevertheless, Fichte develops a theory of penal lawclearly incompatible with that of Kant. That is why I wish not simplyto compare both of the theories in what follows, but rather – now thatI have confronted Kant’s theory of penal law with his concept of rightin Chapters 2 and 3 – also now to compare Fichte’s theory with theconcept of right common to Kant and Fichte.

In Chapter 2, I have already criticized Kant’s retributivism, andindeed both the classical interpretation of this retributivism as thesole justification of the existence of punishment and the newer inter-pretation of this retributivism as justification of the degree of punish-ment. In both interpretations, Kant’s retributivism cannot be derived,or deduced, from his concept of right: moreover, his retributivismeven contradicts his concept of right. That is why I will now, as a leadin, sketch with what rationales Fichte rejects Kantian retributivism as ajustification of the degree of punishment, even though he defines thedegree of punishment according to the criterion of talion law.

Fichte published the first part (}} 1–16) of his Foundations of naturalright in 1796, half a year before the publication of Kant’s Doctrineof right. Half a year after the publication of Kant’s Doctrine of right,the second part (}} 17–24) of Fichte’s Foundations of natural rightappeared, in 1797. In 1796, in the chapter in the Foundations of natural

1. Cf. Jean-Christophe Merle (ed.), “Eigentumsrecht,” in Merle (ed.), Johann Gottlieb Fichte:Grundlage des Naturrechts (Berlin: Akademie Verlag, 2001), pp. 159–72 (p. 161).

87

right about the right of coercion (}} 13ff.), Fichte developed thisaspect of the Kantian mixed theory, which was then neglected, butsince the 1980s it has again been receiving attention: general deter-rence through threatening punishment. Fichte’s actual theory, as itwas formulated in 1797 in } 20 of the Foundations of natural right,seems, however, to be a mixed theory: Fichte interpreters have shownthat in the Foundations of natural right a combination of elements ofgeneral deterrence through threatening punishment, elements of spe-cific deterrence and elements of reform and of rehabilitation canbe found. In this respect, Kaufmann rightly sees a commonalitybetween the Kantian penal law interpreted as mixed theory andFichte’s actual theory of punishment:

Here, in the principles of his theory of coercion, Fichte is very modern,in the respect that he propagates the unification theory overwhelminglyaccepted nowadays of the connection between the theories of deterrenceand retribution. The justification of punishment, on the whole, occursthrough deterrence; in the determination of who should be punished,not only the damage plays a role, but also the extent to which a bad oran inadequate will played a role. The same is valid for determining thedegree of punishment.2

Likewise, Lazzari observes “that both pillars of the Fichtean theoryof punishment, the ideas of performing expiation and of deterrence,are widely in harmony with the diffuse requirement of a more humanesystem of punishment.” Lazzari praises the mixed character of theFichtean theoryofpunishment, even thoughhe is awareof theproblemsconnected with it:

A “mixed” attempt of that sort appears in many respects reallyilluminating and befitting the multiplicity of the aspects having to betaken into account. What is problematic in Fichte’s argument is, first andforemost, the incompatibility of the legal and theoretical assumptions,which substantiate the deterrent part of Fichte’s theory of punishment inthe first part of the Foundations of natural right and the restrictingframework in } 20.3

The objection that I raised earlier against the Kant interpretation asa mixed theory, however, holds ipso facto against Fichte’s mixed theory.Therefore, I will attempt to show in what follows that general

2. Matthias Kaufmann, Rechtsphilosophie (Freiburg i.Br.: K. Alber, 1996), p. 132.3. Alessandro Lazzari, “Eine Fessel, die nicht schmerzt und nicht sehr hindert: Strafrecht,”

in Merle (ed.), J. G. Fichte: Grundlage des Naturrechts, pp. 173–86 (pp. 183–4).

88 punishment as a means of rehabilitation

deterrence through threatening punishment is not compatible withthe Fichtean – or Kantian – concept of right. Thus, it is no coinci-dence that one year later, in the actual chapter in the Foundations ofnatural right about penal law (} 20), Fichte contradicts his chapter onthe right of coercion and, in my opinion, justifies punishment primar-ily with specific deterrence, even though he never explicitly renounceshis earlier mixed theory.

4.1. Retributivism’s lack of justificationfrom the legal perspective

I will now briefly remind the reader of Kant’s argumentation in favorof a retributivist grounding of penal law. In } 49e of the Doctrine ofright, Kant advocates a retributivist position, more precisely talion law:“like for like.”4 Kant delivers no direct justification of talion law, butmerely claims that there are no acceptable alternatives.5 An alternativecan be acceptable only if it passes the test of the categorical imperative,especially its third formula: “So act that you use your humanity, whether inyour own person or in the person of any other, always at the same time as an end,never merely as a means.”6 Kant applies this test to the criminal:

For a human being can never be treated merely as a means to thepurposes of another or be put among the objects of rights to things:his innate personality protects him from this, even though he can becondemned to lose his civil personality.7

Kant declares the retributivism of talion law as being the only theory ofpenal law that passes the test of this formula; all other theories ofpenal law treat the criminal merely as a means. Now, not only do theindividual punishments determined by the retributivist talion lawclearly infringe the aforementioned formula of the categoricalimperative,8 but Kant’s negative argumentation in favor of retributi-vism also clearly applies to that theory, which regards general deter-rence – whether by threatening punishment or by an example carriedout – as being the primary justification of punishment, while leaving,in my opinion, the thesis of specific deterrence untouched.9 It is

4. RL Ak vi:332.5. See Section 2.2.6. GMS Ak iv:429. Practical philosophy, ed. Gregor, p. 80.7. RL Ak vi:331.8. See Section 2.4.9. See Section 2.5.

fichte’s “expiation contract” 89

exactly this neglected option that Fichte reaches as the result of hisdeduction of penal law.

Fichte proceeds under the assumption of the same concept ofright: “Every relation of right is determined by this proposition: eachperson is to limit his freedom through the possibility of the other’sfreedom.”10 He obviously regards Kant’s negative argumentation infavor of retributivism as a crude fallacy. In the second part of hisFoundations of natural right, Fichte criticizes with sharp language Kant’sDoctrine of right, which had appeared nine months earlier:

Punishment is not an absolute end. The claim that it is (whether statedexplicitly or through propositions that implicitly presuppose such apremise, e.g. the unmodified, categorical proposition that “he whohas killed, must die”) makes no sense.11

Some pages later, regarding the “claims” he is attacking, Fichte refersexplicitly to Kant, “the great, though not infallible, man”12 whoaccuses Beccaria of alleged “overly compassionate (Empfindelei) feel-ings of an affected humanity.”13 Fichte turns against

an absolute right of punishment, according to which judicial punishmentis regarded not as a means, but as itself an end, which is said to begrounded on a categorical imperative that is itself not furtherexaminable . . . By relying on what is supposedly unexaminable, thistheory allows its proponents to exempt themselves from the need toprove their claims and so to charge those who think differently withsentimentality (Empfindelei) and an affected humanitarianism.14

Thereby, Fichte does not wish to contest, in any way, that retaliation isa principle of justice. In the second edition of his Attempt at a critique ofall revelation (1793), Fichte definitely sees a motivation to moral actionin the morally necessary proportionality between virtue and happiness,that is to say, in God as the infinite rational being whom Fichte declaresas just.15 In the Foundations of natural right, Fichte claims furthermore:

10. GNR i/3 411. Johann Gottlieb Fichte, Foundations of natural right: Grundlagen desNaturrechts nach Principien der Wissenschaftslehre, ed. Frederick Neuhouser, trans.Michael Baur (Cambridge: Cambridge University Press, 2000), p. 109.

11. GNR i/4 60. Foundations of natural right, p. 228.12. Footnote at GNR i/4 76. Foundations of natural right, p. 245.13. RL Ak vi:334–5. Practical philosophy, ed. Gregor, p. 475.14. GNR i/4 76. Foundations of natural right, p. 245.15. Cf. Jean-Christophe Merle, “Il punto di vista educativo e religioso dei Contributi

destinati a rettificare il giudizio del pubblico sulla Rivoluzione francese: ladimensione politica del Saggio di una critica di ogni rivelazione,” in Aldo Masullo


“There is absolutely no dispute about whether a murderer has beentreated unjustly, if he, too, should lose his own life in a violentmanner.”16 However Fichte immediately stresses that the question ofpenal law is an “entirely different question,” that is, the question

from where . . . [does] a mortal get the right of this moral world-order,the right to render the criminal his just deserts? and it was this purelyjuridical question that the noble Beccaria (who was certainly notunfamiliar with that kind of moral judgment) had in mind. Whoeverascribes this right to a worldly sovereign will surely be required . . . toregard . . . every government as a theocracy.17

4.2. Penal law as a specific type of the right of coercion

In his chapters about the right of coercion (}} 13–16), Fichte identi-fies this with the right of coercion exercised by the state. Now, such anequation is by no means self-evident, as a comparison with Kant’s rightof coercion and penal law shows.

Numerous Kant interpreters proceed under the assumption, whichthey consider as obvious, that in Kant’s thought penal law naturallyand directly belongs to the right of coercion, even though Kantneither claims nor suggests this interpretation. Next, let us investigatewhat exactly the right of coercion consists of for Kant. The right ofcoercion exercised by the state is derived and deduced from that veryright of coercion inhering in every person as soon as his or her rightsare infringed. This individual right of coercion entitles the batteredperson to curtail the freedom of a fellow human being, because,according to Kant,

Whatever is wrong is a hindrance or resistance to freedom in accordancewith universal laws . . . Therefore, if a certain use of freedom is itself ahindrance to freedom in accordance with universal laws (i.e., wrong),coercion that is opposed to this . . . is consistent with freedom inaccordance with universal laws, that is, it is right.18

The consummate example of this individual right to coercion is theright to self-defense. The state, to which this right is assigned, employs

and Marco Ivaldo (eds.), Filosofia trascendentale e destinazione etica (Milan: Guerini,1995), pp. 303–25.

16. GNR i/4 77. Foundations of natural right, p. 246.17. GNR i/4 77. Foundations of natural right, p. 246.18. RL Ak vi:231. Practical philosophy, ed. Gregor, p. 388.


the equivalent of the individual right to self-defense when seeking toprevent the commission of an infringement of the law by means ofpolice at the actual moment when the violation of rights is still just anattempt. The state is also authorized to exercise a right of coercion inthe forms of nullification or annulment, seizures, and the revocationsof licenses (withdrawing a concession or a privilege, in the Hohfeldianmeaning, for example nowadays a driving license). Apart from that,the state is endowed with a right of coercion after the commission of arights infringement (as with the forced restitution of stolen goods)and with the court-ordered delivery of compensation for damages, ifthis is possible (with some rights infringements – typically with murder– there is no fitting substitute restitution). These cases concern retri-bution in the narrow sense. Where it is possible (with some rightsinfringements – typically, once again, with murder – it is fully impos-sible)19 this retribution completely satisfies the legitimate individualclaims of the victim of a rights infringement. It is in no case a punish-ment. Whereas Kant presents the right of coercion immediately in his“Introduction to the doctrine of right”20 and links each right with aright of coercion, he first treats penal law only much later and merelyas one of the elements of “public law.”21

Fichte diverges from Kant not only in introducing penal law as earlyas in the deduction of the right of coercion in general, but also in hisdefinition of the right of coercion as well as in the end he ascribes toit. Whereas the Kantian right of coercion merely resists an existing“hindrance to freedom in accordance with universal laws,” coercion,according to Fichte’s chapter about the right of coercion, alreadyprevents the formation of such a hindrance even before the attempt:

Now if things could be arranged so that the willing of any unrightfulend would necessarily, and in accordance with an ever-operative law,result in the opposite of what was intended, then any will that is contraryto right would annihilate itself . . . It was necessary to present thisproposition in its full, synthetic rigor, since all laws of coercion, or penallaws, (the entirety of penal legislation) are grounded on it.22

According to the respective definitions of the right of coerciondelivered by Kant and Fichte, the right of coercion appears for Kant

19. See Section 2.3.20. RL Ak vi:231. Practical philosophy, ed. Gregor, p. 388.21. RL Ak vi:331–7. Practical philosophy, ed. Gregor, pp. 472–8.22. GNR i/3 426. Foundations of natural right, p. 126.


initially in the perpetration of a rights violation, while for Fichte, onthe other hand, it occurs already, in its very essence, in the intent ofperpetration. Fichte states clearly:

The security of the two parties is not supposed to depend on acontingency, but on a near-mechanical necessity that excludes everypossible exception. There can be such security only if the law of rightis the inviolable law of both parties’ wills.23

Incidentally, it is worth noticing that, in Kant’s justification ofpunishment, deterrence through threatening punishment plays anexplicit role only in the passage about the right of necessity, and thatit plays this role only in an indirect and negative way; therefore, withFichte, deterrence receives a much larger importance.

4.3. General deterrence is not the true justification of penal law

In } 20 of the Foundations of natural right – in the actual paragraphabout penal law – Fichte speaks no longer about the threat ofpunishment, but about the punishment itself,24 which presupposesthat deterrence by threatening punishment has failed – at leastpartly: the crime has already taken place. In view of the demanding,declared goal of the general deterrent threat of punishment – “Thethreat of punishment aims to suppress bad wills . . . in which casepunishment will never be necessary”25 – the failure is obvious. Eitherthe humbler task of general deterrence through example must inherein the punishment of a criminal or it must have another end. In thefirst case, and in the absence of any other goal, the criminal would betreated as a mere means for general deterrence, which is, however,forbidden.26 Therefore, another primary justification of punishmentmust be searched for, which would in turn substantiate the normative

23. GNR i/3 424. Foundations of natural right, p. 124.24. In Rainer Zaczyk, Das Strafrecht in der Rechtslehre J. G. Fichtes (Berlin: Duncker and

Humblot, 1981), Zaczyk emphasizes that the chapter about the right of coercion“does not deal with penal law in the narrow sense, but instead only with a ‘coercion ofwill,’ in a meaning which remains to be defined more precisely.” The threat ofpunishment is nevertheless necessarily connected – though not in so wide a sense –with punishment. There is an important characteristic of the Fichtean theory here:the attempt to conceive of punishment as being a component of the right of coercionin general, instead of its being a right of coercion meant for special cases and being aspecial form of the right of coercion.

25. GNR i/4 60f. Foundations of natural right, p. 228.26. See Section 4.1.


acceptability of deterrence by threatening punishment – indeed nolonger being the primary aim of punishment, but instead, at best,being the secondary goal thereof.

Now, whoever at the beginning of } 20 more closely examines thethesis and the antithesis of that deduction leading to the “expiationcontract” (Abbußungsvertrag) will notice, first, that in the thesis there isno indication of the punishment to be found, and, secondly, that theantithesis no longer draws on the same concept of coercion as theearlier chapters about the right of coercion.

The thesis depicts the crime as a breach in the civil contract on thepart of the criminal, who excludes him- or herself ipso facto fromsociety. Even though the Antithesis, in the strict sense of the term –by which I mean the paragraph that begins with “Antithesis,” not thefollowing paragraph dedicated to the synthesis – refers to “thesepunishments” in contrast to “other punishments,”27 it obviously doesnot concern a punishment in the legal sense. This becomes clear assoon as one reads the passage about the destiny of the criminal, who canbe considered to be incorrigible and in whose case the thesis of beingexcluded from the commonwealth remains valid. Fichte stresses that

the criminal’s death is not a form of punishment, but only a means toensure security. This gives us the entire theory of the death penalty. Thestate as such, as judge, does not kill the criminal; it simply cancels thecontract with him . . . If, afterwards, the state also kills the criminal, then itdoes so not by virtue of its judicial authority, but through the police. Thistakesplace,not inconsequenceofanypositive right,butoutofnecessity.28

The substance of the synthesis begins with the sentence: “This can bearranged only through a contract of all with all,”29 even though thereis, from the start of the Antithesis to the end of the chapter, noheading for “Synthesis.” The synthesis immediately and explicitlyconcerns a punishment in the legal sense. Yet, “public security” isnot to be reached anymore by the right of coercion, in the sense ofthe chapter about the right of coercion, that is, by a 100 percenteffective right of coercion in the form of general deterrence throughthreatening punishment.30 The synthesis rather deals, on the onehand, with specific deterrence – both by the criminal’s incapacitation

27. GNR i/4 59f. Foundations of natural right, p. 227.28. GNR i/4 74. Foundations of natural right, p. 243.29. GNR i/4 60. Foundations of natural right, p. 227.30. GNR i/4 426. Foundations of natural right, p. 228.


(the convicts live on their own island) and by the threat of irrevocableexclusion from the commonwealth in the absence of any reform –and, on the other hand, with general deterrence by example:

The [penal] law’s first aim was to prevent the criminal from committinga crime. Since this goal was not achieved, the state’s punishment ofthe criminal serves another purpose: to prevent other citizens, and toprevent the criminal in the future, from committing the same offense.31

If general deterrence – whether through threatening punishment orexample – were the end of the punishment, then it would be betterserved by sentencing people to a death in the desert or summaryexecution than by a prison sentence or even by forced labor. BecauseFichte provides for the possibility of reform and the rehabilitation ofthe criminal, criminals will possibly be released one day. In view ofpublic security and general deterrence either by threatening punish-ment or by making an example of the criminal, every instance ofrecidivism means that the punishment recommended by Fichteappears to be less effective than exclusion from society, which,according to Fichte, equally leads either to the death of the criminalin the desert or to shooting the criminal like a wild animal. In theAntithesis, Fichte makes it clear that the prerequisite for legal punish-ment to replace the exclusion of the criminal from the common-wealth is that by legal punishment “the mutual security of the rightsof all in relation to all others”32 is met as effectively as throughexclusion from the commonwealth. By the acceptance of a possibleincidence of recidivism, the conclusion can be drawn that publicsecurity – which in the chapter about the right of coercion stilldemanded the almost mechanical, full prevention of every crime evenbefore the stage when it is attempted – is now understood only in amuch more relative sense: public security may be an important good,but not an absolute good anymore.

Whoever reads more closely the chapter about the right of coercionadmittedly finds a similar inconsistency already there. Fichte deter-mines a degree of punishment for general deterrence through thethreat of punishment that is in no way optimal for that goal: iustalionis, which is supposed to countervail the criminal’s impetus forcrime (“the opposite to A” – “A” being the criminal goal – so that the

31. GNR i/4 61. Foundations of natural right, p. 228.32. GNR i/4 59. Foundations of natural right, p. 227.


opposite to “A” means the “counterpoise”).33 In } 20, Fichte explicitlyrefers back to } 14, in order to recommend that the “punishment mustbe equal to the offense: poena talionis.” According to him, it should bethe “material principle of positive punishment within a state,” in short,a criterion. Fichte does not fortify his claim with any argument – otherthan the vague image of a counterpoise, that is, of the pans of a scalebalancing – that both general deterrence, whether by the threat ofpunishment or by example, and specific deterrence are best achievedby way of talion law, that is, by a retributivist degree of punishment.Therefore, the conflict between the different theories of penal lawpertains, allegedly, only to the grounds for punishment, not to itsdegree. Nevertheless, Fichte emphasizes that “exclusion from the stateis the most terrible fate a human being can encounter”;34 he evenrejects any form of torture before killing the criminal.35 The objectioncould be raised that killing the criminal without any court sentence,that is, as a mere police measure, provided that it is complemented bytorture before the homicide and by making public both the tortureand the killing, might deter even in cases in which talion law does notsuffice to curb the criminal impetus. In his Antithesis, Fichte himselfmentions that talion law is a much milder punishment than exclusionfrom the commonwealth.36 If one were to proceed under the assump-tion of Fichte’s psychological premise that the more severe the pun-ishment is, the greater the criminal impetus drops off, then it followsthat exclusion from the commonwealth is a more efficient deterrentthan ius talionis. Nonetheless, Fichte ultimately rejects exclusion fromthe commonwealth. Moreover, not once does he attempt to show thatthis exclusion from the commonwealth would be ultimately linked toa loss of deterrence.

Only because of Fichte’s unsubstantiated and implausible assump-tion that general deterrence is best reached by ius talionis do we havea paradox before us: legal punishment (retaliation as in talion law)meant to serve general deterrence by the threat of punishment is

33. GNR i/4 426f. Foundations of natural right, p. 229.34. GNR i/4 68. Foundations of natural right, p. 237.35. Cf. GNR i/4 74. Foundations of natural right, pp. 242–3.36. One could argue that exclusion from the commonwealth sets a limit for the

punishment determined by general deterrence. It is indisputable that exclusion fromthe commonwealth is mostly less painful than torture before execution. Nevertheless, inmost cases, exclusion from the commonwealth, which leads to a sure death “in thedesert,” remains a worse fate than the degree of punishment recommended by Fichte’stalion law.


milder than the lack of a legal punishment, that is to say, milder thanextralegal punishment or exclusion from the commonwealth. There-fore, whoever compares the legal punishment to the lack of a legalpunishment will come to see that the legal punishment does not implypunishment in the sense of the infliction of suffering due to thecrime, but instead a mitigation of extralegal punishment. Under thesecircumstances, Fichte should have argued for the weaker antithesisaccording to which legal punishment, despite its lesser deterrenteffect, does not lack every deterrent effect, as it would in the case inwhich the criminal were further to enjoy the status of a normal citizen,that is, impunity. With such an antithesis, the legal punishment rec-ommended by Fichte must exert a not insignificant deterrent effect.However, because this legal punishment in no way optimizes generaldeterrence, it is impossible for general deterrence to be punishment’sprimary end. If at all the case, general deterrence would rather be theprimary end of an extralegal punishment (in my opinion, it is noteven that, because exclusion from the commonwealth has a reason –the criminal is a potential danger for the commonwealth – rather thana general deterrent end). In this respect, Fichte is not able to make theclaim anymore that, since exclusion from the commonwealth is notas effective a deterrent as legal punishment, there is “in such cases . . .no reason to exclude the offender; but admittedly . . . there would alsobe no reason not to exclude him. The decision would be a matter offree choice.”37

After Fichte has in this way wrongly assumed the parity of legal andextralegal punishment out of consideration for general deterrence,he reaches for a subsidiary criterion in order that the aforementioned“decision” should not be left to “free choice.”

The subsidiary criterion could consist of the investigation into whatdirectly arises from the combination of the concept of right and theconcept of infringement of right. The result of this investigationshould actually not deliver merely a subsidiary criterion for theappraisal of competing notions of degrees of punishment, but insteadthe primary grounding of punishment itself. But, at first glance, thisoption is not to be found. Now, the logical consequence of the recip-rocal limitation of freedoms contained in the Fichtean concept ofright combined with the infringement of these limits would be thecomplete exemption from legal punishment, that is, the unadorned

37. GNR i/4 60. Foundations of natural right, p. 227.


exclusion from the commonwealth. It would primarily be concernedmerely with the consequence of the infringement of rights, that is,with the elimination of an obstruction of rights, and not be concernedwith goal-oriented measures, that is, with the prevention of obstruc-tion of rights that has not yet occurred.

Instead of this, Fichte adopts the reciprocal “usefulness” of thecriminal and of the state to one another as a subsidiary criterion. Itresults from this criterion that the criminal, despite the crime, con-tinues to be a member of the commonwealth. According to thesubsidiary nature of this criterion, general deterrence by threateningpunishment and by example remains, throughout the entire text of} 20 of the Foundations of natural right, the “primary end”;38 rehabili-tation constitutes only a secondary justification of punishment.

Now, Fichte introduced the subsidiary criterion only because of itsfalse premise that exclusion from the commonwealth – extralegalpunishment – and legal punishment out of consideration for deter-rence are equally efficient. Because this premise is false, Fichte musthave given up, de facto and tacitly, general deterrence as being theprimary end of penal law that would require exclusion from thecommonwealth. Admittedly, Fichte refers to the principle of generaldeterrence not only at the beginning, but also in the entire text of} 20. Nevertheless, he does not argue in favor of this principle, anddoes not derive any consequences from it regarding the degree ofpunishment. Therefore, we must draw the following conclusion: whatFichte introduces as being a subsidiary criterion constitutes in realitythe single true, primary justification of legal punishment in } 20. It isnow time to investigate this justification.

4.4. Specific deterrence and rehabilitationas justification of legal punishment

Fichte justifies only very briefly why he opts for an alternative to theexclusion of the criminal from the commonwealth. His entiregrounding can be summarized in two citations:

It is just as much in the state’s interest to preserve its citizens . . . as it isin each individual’s interest not to suffer the loss of all rights for everysingle offense.39



This expiation contract is useful for all (for the state as a whole) as well asfor each individual citizen. Under it, the whole obtains both the prospectof preserving citizens whose usefulness outweighs their harmfulness, aswell as the obligation to accept their expiation; the individual citizenobtains the perfect right to demand that some expiation be accepted inplace of the more severe punishment that he deserves.40

Whereas the interest of the citizens of a state to be legally instead ofextralegally punished is easily comprehensible, the usefulness of thecriminal to society deserves at least some explanation.

It would be a mistake to suspect that there is a utilitarian element inthis usefulness. It would be in equal measure wrong to believe that thisappeal to usefulness does not represent any legal rationale and thatthe legal concept of right remains unaffected in the exclusion of thecriminal from the commonwealth only for the reason that the com-monwealth persists amongst the other, innocent citizens of the state.Already in } 4, two instances of deduction of the concept of right attestto the fact that, for Fichte, exclusion from the commonwealth, owingto the denial of mutual recognition, cannot be a final and absoluteexclusion. Exclusion should rather remain temporary for the durationof the denial of mutual recognition. This holds true for all finite,rational beings. Let us look at the aforementioned two instances.

1. Fichte writes:

In each relation into which I enter with the individual C, I must refer tothe recognition that has occurred and must judge him in accordancewith it . . . Assuming that his action is indeed determined by the sensiblepredicates of his prior actions . . . but not determined by his havingrecognized me as a free being, i.e. assuming that, by means of his action,he robs me of the freedom that belongs to me and thus treats me as anobject . . . Thus, in this case, I am able, with perfect consistency (which ismy only law here), to treat him as a merely sensible being, until bothsensibility and rationality are once again united in the concept of hisaction.41

Fichte explains his position in the following way:

I . . . appeal to a law that is valid for us both, and apply that law to thepresent case . . . But, in so far as I appeal to that common law in myopposition to him, I invite him to be a judge along withme; and I demandthat in this case he must find my action against him consistent and must

40. GNR i/4 60. Foundations of natural right, pp. 227–8.41. GNR i/3 355–6. Foundations of natural right, pp. 45–6.


approve of it, compelled by the laws of thought . . . The source of thisobligation is certainly not the moral law: rather it is the law of thought;and what emerges here is the syllogism’s practical validity.42

In the lines that follow that (heading “C”), Fichte universalizes thisrequirement for consistent behavior: it holds true for relations toevery individual.

2. In } 4, Fichte in general attempts to show what his headingsalready clearly formulate: “(i) I can expect a particular rational being torecognize me as a rational being, only if I myself treat him as one.”43 “(ii) Butin every possible case, I must expect that all rational beings outside me recognizeme as a rational being.”44 “(III) . . . I must in all cases recognize the free beingoutside me as a free being, i.e. I must limit my freedom through the concept ofthe possibility of his freedom.”45 I will not address the reasons why,according to Fichte, the satisfaction of these requirements is a pre-requisite of the possibility of self-consciousness in finite, rationalbeings; even less will I treat the question here as to whether Fichtewas right about this issue. Here I only have to observe that, accordingto Fichte, the requirement that every finite, rational being shouldbelong to the commonwealth and be treated as a member of it is acomponent of the concept of right, so that the decision about whobelongs to the commonwealth is not even at the disposal of thecommonwealth itself. The “laws of thought” demand that every finite,rational being is recognized as a member of the commonwealth, solong as this being recognizes other rational beings and treats them assuch, or as soon as the being is again ready to do so after having deniedthis recognition for a time to others.

Thus Fichte makes it clear in } 4 that exclusion from the common-wealth cannot be final. It is merely a suspension that lasts as long as ittakes for the denier-of-recognition to recognize others. It becomesclear that Fichte’s use of the word “usefulness” is absolutely notapplied in the utilitarian sense. Fichte’s background is neither theScottish school nor the utilitarianism of the eighteenth century, butinstead modern natural right, which is inspired more from Cicerothan from Aristotle. According to Cicero, that same usefulness (utili-tas), which is not a merely apparent usefulness, is defined as being

42. GNR i/3 356. Foundations of natural right, p. 47.43. GNR i/3 352. Foundations of natural right, p. 42.44. GNR i/3 353. Foundations of natural right, p. 43.45. GNR i/3 358. Foundations of natural right, p. 49.


nothing other than that which contributes to right in the sense of ajust ordering of the community.46 Fichte’s following explanationshould be read with this definition of usefulness in mind:

This contract is useful for all (for the state as a whole) as well as for eachindividual citizen. Under it, the whole obtains . . . the prospect ofpreserving citizens whose usefulness outweighs their harmfulness.47

Usefulness for state and citizens consists of the punishment contri-buting to what is just, that is, contributing to the establishment of alegal system. If the state of law and the state of nature amount to adichotomy without a third option, which is the case with Kant and alsowith Fichte, then the crime poses a transition from a state of law to astate of nature between the criminal and other people. The conceptsof right and usefulness require, however, that all humans are part ofthe commonwealth. The usefulness of legal punishment (in the two-fold meaning of the word “usefulness”: in both the ancient and themodern senses of the term) therefore consists in reestablishingthe state of law between the criminal and the community, and, indeed,as quickly as possible. The concept of right requires simultaneouslythe suspension and the fastest possible rehabilitation of the criminalas a member of the commonwealth, provided that he or she is againjust as able as others to adhere to the laws.

That is why Fichte quickly abandons ius talionis, which he hadinitially applied for the degree of punishment, in favor of specificdeterrence and the theory of reform. At the beginning of } 20, Fichtestill speaks of “punishment . . . equal to the offense: poena talionis.”48

46. Cf., for example, Cicero in De officiis, ii.ii.9f: “What I am going to next address is thatwhich is labelled ‘beneficial.’ Custom has stumbled over this word and strayed fromthe path, gradually sinking to the point where she has severed honourableness frombenefit, decreeing that something can be honourable which is not beneficial, andbeneficial which is not honourable. Nothing more destructive than this custom couldhave been introduced into human life . . . For they [the philosophers] hold thatwhatever is just is also beneficial, and again, whatever is honourable is also just.Therefore it follows that whatever is honourable is also beneficial. Those who donot see this clearly often admire shrewd and crafty men and mistake wickedness forwisdom. Theirs is an error that must be uprooted; and their fancy must be whollyconverted to that hope which consists of the understanding that they will achieve whatthey want by honourable policies and just deeds, and not by deceit and wickedness.”Marcus Tullius Cicero, On duties, ed. M. T. Griffin and E. M. Atkins (Cambridge:Cambridge University Press, 1991), p. 66.



Four pages later, Fichte says of the criminal that, “He must forfeit hisfreedom until it is clear that he has reformed; or else he must beexcluded from the state without mercy.”49 Five more pages later,Fichte states more precisely:

But these institutions for reform must also be prudently arranged. First,they must be actually separated from society and established accordingto the spirit of the law [specific deterrence through incapacitation].The state has full responsibility for any damage caused by someone who,at the time, is being excluded from society. Therefore, these personshave lost all their freedom. However, if a person is to reform himself,and if his efforts at reform are to be subject to judgment, then he mustbe free. Therefore, a chief maxim is: such people must be free withinnecessary limits and must live in society among themselves.50

Therefore, what remains of this Thesis is the exclusion from thecommonwealth, which is, however, still only a provisional exclusion.What remains of the Antithesis is the criminal’s affiliation to thecommonwealth, which, however, is provisionally suspended and onlyfully enters into force after the criminal’s reform. The synthesis is the“expiation contract.” The expiation contract consists in offering achance to the convicted criminal to escape exclusion from the com-monwealth and from that status as a human being who, like a savageoutlaw, can be freely shot like a harmful, wild animal. Instead of beingexcluded, the criminal can decide to be punished by spending time ina prison or upon an island with other convicts. Depending on thecriminal’s behavior in the penal camp, after a specific period of timehe or she will then be either readmitted into the commonwealth orpermanently excluded from it.51 Fichte’s contract of expiation entailsa double originality: first, it is not a contract between the ordinarycitizens designed to legitimate the state’s authority to punish; and,secondly, it conceives of punishment not as it would of an evil, but,instead, as of an opportunity for the criminal to improve his or her fate.In these two regards, Fichte’s conception diverges, for example, fromthe “non-retributive Kantian approach to punishment” based on anappeal to a hypothetical social contract.52

49. GNR i/4 65. Foundations of natural right, p. 233.50. GNR i/4 70. Foundations of natural right, p. 239.51. GNR i/4 71. Foundations of natural right, p. 240.52. Michael Clark, “A non-retributive approach to punishment,” Ratio, 17, no. 1 (2004),

12–27.


What soon proves to be Fichte’s justification of penal law (specificdeterrence and rehabilitation) takes on two essential elements ofKant’s philosophy of right. First, the decision about who shall belongto the commonwealth is a decision that does not fall to the common-wealth, because every human should belong to it. Secondly, the stateof law should be reached by the shortest possible path; detoursare allowed – and dictated – only along this path (Kant’s “permissivelaw”).53

4.5. Is general deterrence really necessary fora justification of punishment?

Naturally, the objection can be raised against my interpretation of } 20that nowhere in this entire paragraph does Fichte explicitly abandonthe justification of penal law with that sort of deterrence that consistsof threatening punishment. Nevertheless, in the course of } 20, Fichteabandons general deterrence and talion law; in the end, the criminal’sreform – and not ius talionis – determines the date of release. This alsomeans that sometimes the punishment should be harsher than whatretaliation would require:

It would be very prudent if the criminal himself were allowed todetermine, in accordance with the degree of his depravity, the lengthof time of his reform – but with the proviso that he would later be free toextend it in accordance with a certain standard. But each criminal mustbe given a peremptory term for reform, in accordance with hisparticular crime.54

After this term expires, the criminal should either – in the case ofreform – be reassumed into the commonwealth, or – in the case thatreform did not occur – be permanently excluded from it.55

As a result of the occurrence of multiple justifications of penal lawby Fichte, which exclude each other from being primary justifications

53. Cf. Section 2.4 above and Reinhard Brandt, “Das Erlaubnisgesetz, oder: Vernunft undGeschichte in Kants Rechtslehre,” in Brandt (ed.), Rechtsphilosophie der Aufklarung(Berlin: De Gruyter, 1982), pp. 233–85.

54. GNR i/4 71. Foundations of natural right, p. 240.55. Here it appears to me that Fichte contradicts his premise and antithesis, according to

which every human should be offered legal punishment, provided that this persondoes not pose any danger – for instance, for the prison guards. Now, we cannotexclude the possibility that a criminal, after a certain period of time, still has notreformed, but without having posed an additional security problem ipso facto bycontinuing to remain in the “correctional facility.”


and also cannot be combined into a mixed theory because they lead todifferent notions of degree of punishment, one must either declareFichte’s penal law as being hopelessly contradictory or investigate theindividual arguments, as well as their role and relevance in Foundationsof natural right, and then reach an assessment. In view of this situation,my argument has shown two things. First, if we do not accept the falsepremises, according to which legal and extralegal punishments wouldhave an equally deterrent effect, then only Fichte’s argument for spe-cific deterrence by incapacitation would remain as a primary justifica-tion for penal law. Secondly, that argument is the only one that iscompatible with the degree of punishment, which Fichte conclusivelyadopts at the end of } 20, for this reason. That is why I amof the opinionthat Fichte – admittedly only after lengthy comings and goings –substantiates penal law with specific deterrence and rehabilitation.

Naturally, legal punishment exerts a deterrent effect by beingopenly disseminated (general deterrence by threatening punishment)and by the execution of punishment not being a secret (generaldeterrence by example). This effect certainly contributes to publicsecurity and, for that reason, is thoroughly welcomed by the common-wealth. Nevertheless, threatening punishment and public executionof punishment are only then allowed if the infliction of punishment isitself justified, because punishment exceeds the reciprocal limitationof freedoms contained in the concept of right. If general deterrenceby threatening punishment were 100 percent effective and the pun-ishment, ipso facto, never needed to be executed – as Fichte wishesin the chapter about the right of coercion – then the infliction ofpunishment would never be justified; then, however, the whole of } 20of the Foundations would be superfluous.

The chapter about the right of coercion clearly stipulated a 100percent effective general deterrence by threatening punishment: “thesecurity” of any two given persons

is not supposed to depend on a contingency, but on a near-mechanicalnecessity that excludes every possible exception. There can be suchsecurity only if the law of right is the inviolable law of both parties’ wills.56

It is appropriate that the law of right should be normatively inviolable.In contrast, human freedom, in the sense of the power of choice, ispotentially infinite; it is always able to transgress limits – and even

56. GNR i/3 424. Foundations of natural right, p. 124.


normatively inviolable limits as well. Solely the laws of nature areunable to be infringed. Coercion is surely an effective yet not 100percent effective means to obtain public security. As Hegel later wrote,“Only he who wills to be coerced can be coerced into anything.”57 Manya culprit accepts the risk of punishment, even if he or she later regretsit because of fear and may come to be aware of the fact that thepunishment actually was too high a price to pay. Therefore, not onlythreatening punishment but also inflicting punishment needs to bejustified. The threat is not first in line for justification.

Hegel attains specifications of penal law, similar to Fichte’s, withoutneeding to take a detour via general deterrence and via its shortcom-ings. Furthermore, Hegel proceeds under the assumption of thenecessary validity of right and, therefore, does not need the thoughtexperiment of exclusion from the commonwealth. In this respect,Hegel offers a systematically more stringent representation of thesame view about penal law, as I would like to show in the following.

An enlightening comparison to the conceptions of Fichte andHegel is offered by Karl Christian Friedrich Krause (1781–1832) inhis philosophy of law, which is based upon similar Kantian–idealistfoundations and was published in the period between the publicationof both of their works. Admittedly, only the first section (1803) andthe posthumous second section about cosmopolitan law (1890) ofKrause’s own Grundlage des Naturrechts (The foundation of natural right)have been published, whereas the planned section about public lawand a subsection about penal law within this section remained stuck inthe planning stages.58 But Krause’s Latin habilitation thesis of 1802,which he translated into German and upon which he made com-ments,59 provides an informative sketch of the philosophy of law.According to Krause, penal law should cause it to appear to the citizenthat adherence to the law is more useful than adherence to impulses;60

in a footnote – probably from 1805 – he states this more precisely:punishment aims, according to Krause, “especially toward civicly

57. GPhR } 91. G. W. F. Hegel, Elements of the philosophy of right, trans. H. B. Nisbet(Cambridge: Cambridge University Press, 1991), p. 120.

58. Cf. Wolfgang Forster, Karl Christian Friedrich Krauses fruhe Rechtsphilosophie und ihrgeistesgeschichtlicher Hintergrund (Ebelsbach: Aktiv Druck undVerlag, 2000), pp. 210–16.

59. K. C. F. Krause, Abhandlung uber die Idee und die Eintheilung der Philosophie und derMathematik und den innigen Zusammenhang beider, in Paul Hohlfield and AugustWunsche (eds.), Philosophische Abhandlungen aus dem handschriftlichen Nachlasse (Leipzig:Schulze, 1889), pp. 5–40.

60. Krause, Abhandlung uber die Idee, pp. 34–5.


reforming the criminal thoroughly and in a rational way, and towardalso ensuring that punishment does not affect his other rights.”61

Krause remarkably proceeds in the determination of punishmentfrom the assumption of the inalienable rights of the perpetrator,which, as a “formula,” indicates “at the same time the limits of punish-ment.”62 Krause clearly breaks away from retributivism. As the otherpublished parts of Grundlage des Naturrechts63 as well as unpublishedreferences in his letters show, Krause derives from that axiom,amongst others, the absolute inviolability of the body, and, along withthat, a proscription of death as well as of all corporal punishment.64

As in other areas,65 the visionary Krause ironically comes somewhatcloser in the consequences which he draws from his theory to theconclusions which one should expect from a liberal interpretation ofthe Kantian idealist concept of right than the more famous authorsFichte and Hegel.66

61. Krause, Abhandlung uber die Idee, pp. 34–5 (footnote).62. Krause, Abhandlung uber die Idee, p. 35.63. Part i in Krause, Abhandlung uber die Idee, pp. 237–8.64. Cf. Forster, Krauses fruhe Rechtsphilosophie, pp. 108–9, 303.65. For instance, in the theory of marriage, in K. C. F. Krause, Grundlage des Naturrechts oder

philosophischer Grundriß des Ideals des Rechts (Jena: Gabler, 1803), pp. 169ff.66. See the overview provided by Peter Landau, “Karl Christian Friedrich Krauses

Rechtsphilosophie,” in Klaus-Michael Kodalle (ed.), Karl Christian Friedrich Krause(1781–1832): Studien zu seiner Philosophie und zum Krausismo (Hamburg: Meiner,1985), pp. 80–92.


5

HEGEL’S “NEGATION OF CRIME”

5.1. The controversial classification of theHegelian theory of punishment

Unlike in the secondary literature about Kant’s theory of punishment,which for a long time served as the model for all retributivist orabsolutist theories of penal law, the divergence from the traditional– and often held as being self-evident1 – retributivist interpretationhas quite a long history with regard to the interpreters of Hegel’spenal law. Whereas the interpretation of Kant’s penal law as being amixed theory has only come about since the 1980s,2 Hegel’s penallaw was very early on regarded by Christian Reinhard Kostlin as amixed theory,3 by British Neo-Hegelianism (for instance by BernardBosanquet)4 as being a deterrent theory and by John Ellis McTaggart5

1. Cf. Part i, } 62 of Arthur Schopenhauer, The world as will and idea, trans. R. B. Haldaneand J. Kemp, seventh edition (4 vols., London: Kegan Paul, Trench, Trubner, 1909?),vol. 1, pp. 430–52. Nowadays, cf., for instance, Ulrich Klug, “Abschied von Kant undHegel,” in Jurgen Baumann (ed.), Programm fur ein neues Strafgesetzbuch: der Alternativ-Entwurf der Strafrechtslehrer (Frankfurt a.M.: Fischer, 1968), pp. 36–41; Claus Roxin,Strafrechtliche Grundlagenprobleme (Berlin: De Gruyter, 1973), p. 2; Winfried Hassemer,Einfuhrung in die Grundlagen des Strafrechts, second edition (Munich: C. H. Beck, 1990),pp. 283–4; Jean-Claude Wolf, Verhutung oder Vergeltung? Einfuhrung in ethischeStraftheorien (Freiburg i.Br.: Alber, 1992), p. 50; Otfried Hoffe, Gerechtigkeit: einepolitische Einfuhrung, second edition (Munich: C. H. Beck, 2004), p. 79.

2. See Section 2.1.3. Christian R. Kostlin, Neue Revision der Grundbegriffe des Kriminalrechts, reprint (Aalen:

Scientia, 1970). Cf. Andrei A. Piontkowski, Hegels Lehre uber Staat und Recht und seineStrafrechtstheorie, trans. Anna Neuland (Berlin: De Gruyter, 1960), p. 199.

4. Bernard Bosanquet, The philosophical theory of the state (London and New York:Macmillan, 1899).

5. John E. McTaggart, Studies in Hegelian cosmology (Cambridge: Cambridge UniversityPress, 1918), chapter 5.

107

as being a reform theory. Admittedly, this last interpretation has notprevailed. Even today, there is controversy about whether Hegel isreally a retributivist6 or whether he possibly also adopts general orspecific deterrent elements,7 or whether at the opposite extreme, he isonly a theorist of pure deterrence.8

Just as in Kantian studies, at this time mixed theories are alsopredominant in Hegelian studies. Some of the best interpreters –Allen Wood and Georg Mohr, for instance – combine the retributivistinterpretation with that of deterrence. According to Mohr, Hegel’stheory of punishment delivers “a type of grounding, which was new forthat time, for the theory of retribution, which was up to then ratherpoor in its arguments.” According to Mohr, Hegel is able to attain thisthrough allotting “to the punishment as its inherent goal the task of apositive general deterrence.”9 Likewise, Wood initially contends that

6. Cf. especially Mark Tunick, Hegel’s political philosophy: interpreting the practice of legalpunishment (Princeton: Princeton University Press, 1992); in addition, for instance,Ossip Kurt Flechtheim, “Die Funktion der Strafe in der Rechtstheorie Hegels,” inFlechtheim, Von Hegel zu Kelsen: rechtstheoretische Aufsatze (Berlin: Duncker andHumblot, 1963), pp. 9–20 (p. 13); Flechtheim, “Zur Kritik der HegelschenStraftrechtstheorie,” Archiv fur Rechts- und Sozialphilosophie, 54 (1968), 539–48 (p. 542);Peter G. Stillman, “Hegel’s idea of punishment,” Journal of the History of Philosophy, 14(1976), 169–82 (p. 173); Klaus Scala, “Hegels Begriff der Strafe und die moderneStrafvollzugsproblematik,” Hegel-Jahrbuch (1987), 164–70; Igor Primorac, “Punishment asthe criminal right,” Hegel-Studien, 15 (1980), 187–98 (p. 193); and Igor Primoratz (theseare two spellings of the same name), “Banquos Geist: Hegels Theorie der Strafe,” Hegel-Studien, supplement 29 (1986), 53; Kurt Seelmann, “Wechselseitige Anerkennung undUnrecht: Strafe als Postulat der Gerechtigkeit?”, Archiv fur Rechts- und Sozialphilosophie, 79,no. 2 (1993), 228–36; Seelmann, “Versuch einer Legitimation von Strafe durch dasArgument selbstwiederspruchlichen Verhaltens des Straftaters,” Jahrbuch fur Recht undEthik, 1 (1993), 315–26; Matthias Kaufmann, Rechtsphilosophie (Freiburg i.Br.: K. Alber,1996), p. 308; footnotes in G. W. F. Hegel, Principes de la philosophie du droit, trans. Jean-Francois Kervegan (Paris: Presses Universitaires de France, 1998), pp. 180–1; ArthurRipstein, Equality, responsibility and the law (Cambridge: Cambridge University Press,1999), p. 93; Otfried Hoffe,“Konigliche Volker”: zu Kants kosmopolitischer Rechts- undFriedenstheorie (Frankfurt a.M.: Suhrkamp, 2001), p. 79.

7. Cf. Section 1.6 of Ted Honderich, Punishment: the supposed justifications (New York:Harcourt, Brace and World, 1969); Wolfgang Schild, “Ende und Zukunft desStaatsrechts,” Archiv fur Rechts- und Sozialphilosophie, 70 (1984), 71–112; Allen W. Wood,Hegel’s ethical thought (Cambridge: Cambridge University Press, 1990), p. 110; MatthiasKaufmann, “Zwangsrecht (}} 13–16),” in Jean-Christophe Merle (ed.), Johann GottliebFichte : Grundlage des Naturrechts (Berlin: Akademie Verlag, 2001), pp. 125–37.

8. Cf., for instance, Stanley Benn, “An approach to the problems of punishment,” Philosophy,33 (1958), 321–41; Anthony M. Quinton, “On punishment,” in Harry Burrows Acton(ed.), The philosophy of punishment (London: Macmillan, 1969), pp. 55–64.

9. Georg Mohr, “Unrecht und Strafe,” in Ludwig Siep (ed.), G. W. F. Hegel: Grundlinien derPhilosophie des Rechts (Berlin: De Gruyter, 1997), pp. 95–124 (pp. 95, 105).


“Hegel is a genuine retributivist. He rejects as ‘superficial’ all theoriesthat try to justify punishment by the ‘good’ which is supposed to comeof it.” On the next page, Wood observes: “but the state’s intention toreassert the validity of right in the face of wrong looks like an intentionnot to do justice as such, but to promote a good end, namely the publicrecognition of the validity of right.” Wood asks:

Why is it important for the state to assert the validity of right, to express itsdisapproval of crime? Is there any reason for it to do this apart from itsdevotion to such consequentialist ends as preventing future crimes andreassuring people that their rights are being protected . . .?10

These very interpreters highlight in this way that Hegel arguesmore convincingly for retributivism than Kant. According to Mohr,“compared to Kantian retribution theory . . . Hegel’s theory of punish-ment is the argumentatively better executed foundation of this type oftheory.”11 Wood says of Hegel’s theory of punishment: “This positionagrees with Kant, whose commitment to retributivism is clear, butwhose defense of it remains at best embryonic.”12

Just as for Kant’s penal law, for Hegel’s penal law as well it is fair toassume a mixed theory of punishment that ascribes to every argumenta part of the task of justification. Thus, Mohr wishes to recognizeHegelian penal law’s merit of building up “a theoretic instrument”that “does profitable preliminary work underlying the current claimsto a mixed theory.”13 Just like the currently prevailing interpretationof Kant’s theory, Mohr also sees a mixed theory of punishment inHegel:

Taking all the relevant passages from the Elements as the basis for theinterpretation of Hegel’s theory of punishment yields the diversifiedpicture of a modern mixed theory. It differentiates between justice inpunishment as an institution rooted in retributive theory and justice indegree of punishment specified by deterrent theory.14

According to Mohr’s interpretation, the relationship between acertain element of retributive theory and a certain element of deter-rent theory is formulated by Hegel the reverse way of that by Kant.

10. Allen W. Wood, Hegel’s ethical thought (Cambridge: Cambridge University Press, 1990),pp. 109–10.

11. Mohr, “Unrecht und Strafe,” p. 121.12. Wood, Hegel’s ethical thought, p. 109.13. Mohr, “Unrecht und Strafe,” p. 122.14. Mohr, “Unrecht und Strafe,” p. 119.

hegel’s “negation of crime” 109

According to the interpretation of the mixed theory, in Kant deterrenttheory justifies punishment as an institution, and retributive theoryjustifies the degree of punishment. In contrast, Heiko H. Lesch sees inHegel the same relation between both elements of the mixed theoryas in Kant.15

The way in which Hegel’s penal law is interpreted as a mixedtheory, however, only indirectly correlates to the process of argumen-tation in Hegel’s text. Indeed, the connection made by such a mixedtheory between each of the arguments offered respectively by thetheories of deterrence and retributivism does not concern the rela-tionship of the argumentation based on “Abstract right” to the argu-mentation based on “Ethical life.”16 The internal relationship of theindividual arguments based on “Abstract right” to one another is notinquired into by the mixed theory.

5.2. The twofold justifications of punishment

It may astound one to see that the interpretations of Hegel’s theory ofpunishment can contradict one another so deeply that some inter-preters seem to vacillate between multiple interpretations. These dif-ferent interpretations, however, draw on similarly different passages inHegel ’s Elem ents of the philos ophy of right ( 1821 ).

Elements of the Hegelian theory of punishment can be foundnot only in the section “C. Coercion and crime” within “Abstract right,”17

but also in “Ethical life,” under the heading “B. The administrationof justice,” specifically in “b. The existence (Dasein) of the law” and“c. The court of law.”18 The deterrence theory interpretations ofHegel often base themselves on “Ethical life,” whereas the retributivistinterpretations either simply forget }} 218–20 or declare them tobe unimportant. The latter interpretations categorize these para-graphs mainly as being a mere indication that a punishment mustbe tailored to the respective societal conditions without the rationalefor punishment and the principle determining its degree beingaffected by }} 218–20.

15. Heiko Hartmut Lesch, Der Verbrechensbegriff: Grundlinien einer funktionalen Revision(Cologne: Heymanns, 1999), p. 97.

16. GPhR, Parts i and iii, respectively.17. GPhR }} 90–103.18. See GPhR }} 218 and 220, respectively.


5.2.1. An objective and subjective justification of punishment. The reverseof those interpretations that wish to ignore or neglect }} 218–20 is anoverestimation of the second part of the Elements of the philosophy ofright (“Morality”). In “Morality,” there are indeed elements that clearlyhave to do with crime, especially in the section entitled “Purpose andresponsibility.” Yet in the “Morality” section, the word “punishment”is found only three times. Moreover, it is not to be found either inthe main body of the text or in the Remarks but instead in theAdditions – and there it is only an aside. Punishment is only treatedin the part of right as the “universal” or right “in itself” and in the partof the “subjective disposition, but of that right which has being initself.”19 Punishment clearly belongs to the objective aspect of the ideaof right. Yet in the passage out of “Abstract right,” many interpretersattempt to demonstrate the presence of two arguments that are bothin favor of retributivism: an objective one and a subjective one. Theclearest presentation of this twofold argumentation is provided byOssip K. Flechtheim and by Igor Primorac, and in a divergent formby Kurt Seelmann.20

Flechtheim differentiates an “objective justification” from a “sub-jective justification,” since “the objective way to legitimate punishmentappears to Hegel . . . as still being insufficient. He wishes to demon-strate that the criminal, by the criminal deed, also subjectively con-sents to the punishment.”21

In the objective justification, there is an argument that draws “on thenature of the crime and on the law that was injured by it.”22

Flechtheim emphasizes that:

As early as in the theological writings of his youth, Hegel’s deductionhas been based on the axiom that the relation between deed andpunishment is inseparable; that the deed already contains the

19. GPhR } 141, Remarks. G. W. F. Hegel, Elements of the philosophy of right, ed. AllenW. Wood, trans. H. B. Nisbet (Cambridge: Cambridge University Press, 1991), p. 186.

20. Flechtheim, “Die Funktion der Strafe,” pp. 9–20; Flechtheim, “Zur Kritik derHegelschen Straftrechtstheorie,” 539–48; and Flechtheim, Hegels Straftheorie(Berlin: Duncker und Humblot, 1975). Primorac, “Punishment as the criminal right,”187–98; and Primoratz, “Banquos Geist”; Seelmann, “Wechselseitige Anerkennung,”228–36; Seelmann, “Versuch einer Legitimation,” 315–26. Further authors assertthe existence of a twofold argumentation in Hegel’s theory of punishment, forinstance, Bertrand Guillarme, Penser la peine (Paris: Presses Universitaires de France,2003), pp. 63–4.

21. Flechtheim, Von Hegel zu Kelsen, pp. 17–18.22. Primoratz, “Banquos Geist,” 39.


punishment in itself; and that the punishment of the deed necessarilyproceeds from the deed.23

I shall name this thesis the thesis of negation of the negation. According toFlechtheim, the thesis was formulated in } 93, where Hegel states:

Because coercion24 destroys itself in its concept, it has its real expression(Darstellung) in the fact that coercion is cancelled (aufgehoben) by coercion;it is therefore not only conditionally right but necessary – namely, asa second coercion which cancels an initial coercion.25

In } 99, Hegel equates “the cancellation (Aufheben) of the crime, whichwould otherwise be regarded as valid” with “the restoration of right.”26

Flechtheim identifies the “restoration” (Wiederherstellung) explicitlywith the classical thesis of retaliation (Wiedervergeltung), which isdiscussed in } 101.27

According to Flechtheim’s interpretation, the classical retributivistthesis forms the link between the objective and the subjective justifi-cation. That is because, unlike right in itself, the principle of retaliationdraws not on the total legal system, but instead on the individual andon his or her individual deed.

In the subjective justification, there is an argument that draws on therelation of punishment to the “empirical will of the criminal.”Flechtheim sees the subjective justification to be mainly located in} 100, which he interprets in the following manner:

Allegedly, it is inherent in the crime, which is considered as being theaction of a rational being, that the crime is something universal andthat it lays down a law. The lawbreaker recognizes through his or herdeed that this law applies to him- or herself, so that the lawbreaker issubsumable under that law, which is his or her own right. The criminalis honored as a rational being by his or her action being regarded injust such a way, that is, as if the action itself placed a claim on thepunishment by virtue of it being its own right.28

Flechtheim comments that:

From this conception of the legal offense, Hegel obtains the importancethat he attributes to the guilty plea and to the jury court. If the wrongdoer

23. Flechtheim, Von Hegel zu Kelsen, p. 12.24. “Coercion” can also be regarded as being “force.”25. GPhR } 93. Elements, ed. Wood, p. 120.26. GPhR } 99. Elements, ed. Wood, p. 124.27. Flechtheim, Von Hegel zu Kelsen, p. 13.28. Flechtheim, Von Hegel zu Kelsen, p. 17.


consents to his or her punishment, not only in so far as the wrongdoer isthe personification of an objective rational will but also as he or she is themanifestation of his or her subjective particular will, hence every courtand every legal proceeding becomes superfluous since the criminal canpass his own sentence him- or herself. But even Hegel does not dare toconstrue reality in terms of this “idea,” and contents himself with therequirement that “what the judges pronounce does not differ from whatis in the consciousness” from the subjective self-consciousness of thewrongdoer.29

What Flechtheim quotes without providing a page reference isderived, however, from the Additions in } 227, which is expresslyconcerned with confessing to the elements of a crime – therefore, itis not, as Flechtheim assumes, concerned with the criminal’s consentto penal law. Then, according to Flechtheim’s conception, } 101 isseen as being a qualification of } 100. According to Flechtheim, thecriminal should have recognized talion law. Meaning, he continues,that this criminal should recognize that he or she “deserves to bepunished” and that “what the criminal has done should also happen tohim.”30 This thesis is realized in the principle of “equality” betweencrime and punishment. In turn, this principle is specified not as beinga “specific equality,” that is, as being in natura, but instead as being anequality “in terms of its value.”31

It remains unclear, however, why punishment should require sucha twofold justification. The context in Flechtheim suggests that thistwofold justification is meant to bring about the criminal’s reconcili-ation with society. An obvious objection arises, though, against thisrationale for a twofold justification. Reconciliation was indeed Hegel’sgoal in the writings of his youth, but punishment was seen by him atthat time as something that is alien to the criminal as a human being.That is because it was not something subjective, but instead onlysomething objective. At the outset of the section with the title“C. Coercion and crime,” the concern is about something objective: thecancellation (Aufhebung) of crime and the restoration (Wiederherstellung)of right. Admittedly, Hegel writes: “The injury . . . which is inflictedon the criminal is not only just in itself [objective aspect] . . . it is alsoa right for the criminal himself, that is, a right posited in his existent

29. Flechtheim, Von Hegel zu Kelsen, p. 17. Elements, ed. Wood, p. 257.30. GPhR } 101, Remarks. Elements, ed. Wood, p. 127.31. GPhR } 101, Remarks. Elements, ed. Wood, p. 128.


will [subjective aspect].”32 Yet this implies neither that this subjectiveaspect would be necessary for the justification of the punishment as acomplement to the objective justification, nor that the criminal wouldhave to consent to penal law in order that the punishment might belegitimate. Hegel’s formulation in } 100 completely allows that, afterfirst being sentenced and then having served a punishment, thecriminal gains the insight that his or her deed must imply the punish-ment, even though the deed implied the punishment in itself alreadywhile he or she was committing the deed. In such a case, the criminal’sconsent would not be – unlike the case with the cancellation of thecrime and the restoration of right – a (normative) prerequisite, butinstead a result of the crime.

Naturally, this does not mean that the subjective insight of thecriminal into the justification of his or her punishment has no system-atic locus in the Elements of the philosophy of right. Obviously, the processby which the criminal gains insight into the necessity of penal lawmakes possible the “Transition from right to morality.” In the partabout morality that truly does constitute “a further advance in theinner conceptual determination of the will,” punishment is not atissue – the word “punishment” itself practically never appears.33 It isin the part about “Ethical life,” in which the opposition between thegeneral will in itself and the general will for itself is cancelled, thatpunishment is discussed again for the first time. There it is handledfrom the perspective that dangerousness is a criterion for punishmentin general as well as for legal punishment in particular, that is, for theopposite of private revenge. }} 218 and 220 clearly do not make theclaim, however, to contributing to the justification of the existence ofpunishment. To this extent, the justification of punishment restsexclusively on the development of the general will in itself; therefore,

32. GPhR } 100. Elements, ed. Wood, p. 126.33. Cf. GPhR } 104. Elements, ed. Wood, p. 131. The word Strafe (“punishment”) itself

appears only in Hegel’s handwritten marginalia in GPhR } 118 as well as in theRemark on } 132, though Allen W. Wood, for instance in } 218, renders theGerman Ahndung as the English “punishment,” though it could also be rendered as“revenge for a wrong.” In these passages, the word Strafe is used only in the followingcontexts: “Suffering in general – including punishment – as a consequence in general ofthe action,” “punishment, infringement,” and “unhappiness merely as such has noethical interest – just as injustice and punishment are not mere ills” (translationmine); “The sphere in which the above circumstances come into consideration asgrounds for relaxing the punishment is not the sphere of right, but the sphere ofclemency,” in Elements, ed. Wood, p. 161.


it rests on the objective part. My conclusion is that the subjectiveaspect of punishment provides no justification of punishment as beingthe justification of a legal institution, even if there happen to beelements of the concept of right in that justification.

In the subsection “Coercion and crime,” Hegel clearly formulatesthe affiliation of the justification of punishment to the objectiveaspect of the concept of right. In his Remark on } 99, Hegel speaksof “the objective consideration of justice, which is the primary andsubstantial point of view in relation to crime.”34 While the supposedsubjective justification comes under the heading “the criminal’s ownright,” its validity expressly remains conditional to the validity of theobjective justification: “The injury . . . which is inflicted on the crim-inal is not only just in itself (and since it is just [emphasis mine], it is atthe same time his will as it is in itself, an existence . . . of his freedom, hisright).”35 And while the alleged subjective justification appears in thecontext of classical retaliation theory, then it is again expressly condi-tional to the validity of the objective justification: “The cancellation . . .of crime is retribution (Wiedervergeltung) in so far [emphasis mine] as thelatter, by its concept, is an infringement of an infringement.”36 Hegelcriticizes the classical theory of retaliation by opposing it withthe argument of the nullity of crime: “Yet the concept itself mustalways contain the basic principle, even for the particular instance.”Hegel also terms this “the substance of the thing (Sache) itself.”37

Furthermore, no interpreter asks the question whether either theobjective justification alone or the subjective justification alone issufficient. If neither of the two would suffice then one would have toask one more question to find to what extent they reciprocally comple-ment each other and they together would be sufficient. What canbe found regarding this point in the secondary literature tends notreally to be convincing. Primoratz attempts to substantiate the twofoldjustification in the following way:

This aspiration of Hegel to justify punishment both in the objective andin the subjective sense is in no way contingent. Nor is it an expression ofhis doubts about the rightness and convincingness of the one or theother justification and about the consequent need for a complementingargument . . . His fundamental objection that he brings forward against

34. GPhR } 99, Remark. Elements, ed. Wood, p. 125.35. GPhR } 100. Elements, ed. Wood, p. 126.36. GPhR } 101. Elements, ed. Wood, p. 127.37. GPhR } 101. Elements, ed. Wood, p. 128.


Plato consists in Plato not recognizing the principle of subjectivity, thatis, there being no place in his republic for the individual.38

5.2.2. A legal argument and an argument based on recognition of the criminal’shumanity. It still remains for us to consider the alternative interpret-ation of the twofold justification of punishment given by Seelmann.He sees in Hegel “two different constructions for legitimating pun-ishment and penal law”: the “argument from law” and the “argumentfrom recognition.”39 Both link the objective with the subjectivedimension.

The argument from law is the thesis previously mentioned, that is, thethesis of “the criminal’s own right” or “subsumption” under the crim-inal’s own law.40 I will come back to this later. In what follows, I willaddress the question whether Hegel’s (objective) argument from lawis linked to the (subjective) argument from recognition in a twofoldjustification.

The argument from recognition relies on } 97 of the Elements of thephilosophy of right:

When an infringement of right as right occurs, it does have a positive

external existence . . . but this existence within itself is null and void. Themanifestation of its nullity is that the nullification of the infringementlikewise comes into existence; this is the actuality of right, as its necessitywhich mediates itself with itself through the cancellation . . . of itsinfringement.41

According to Seelmann this argument can be understood “only ifone takes as a point of departure Hegel’s understanding of right asbeing a universal relation of reciprocal recognition of free and equalpersons . . . an understanding that Hegel borrowed from Fichte, whileintroducing some modifications.”42 Seelmann himself formulates theargument along the following lines:

Since . . . recognition is a reciprocally conditioned relation [accordingto Hegel], the wrongdoer himself withdraws the recognition from him-or herself. Therefore, punishment is nothing other than the mere

38. Primoratz, “Banquos Geist,” 40.39. Seelmann, “Versuch einer Legitimation,” 319; cf. Seelmann, “Wechselseitige

Anerkennung,” 228.40. Cf. GPhR } 100. Elements, ed. Wood, pp. 126–7.41. GPhR } 97. Elements, ed. Wood, p. 123.42. Seelmann, “Wechselseitige Anerkennung,” 230.


manifestation of this situation that the wrongdoer him- or herself hasbrought about, whereby punishment lowers the physically observablelegal status of the wrongdoer in order to create a legal relationshipbased on equality.43

The argument from recognition is thus not to be understood as ifpunishment were justified by the crime being a refusal of recognitionon the part of the criminal, because based on this denial of recogni-tion one could just as well shoot the criminal as if he or she was adangerous animal that anyone should be allowed to kill without anyparticular procedure – that is, without a legal sentence, as is the case inFichte’s thought experiment, to which Hegel alludes in another pas-sage.44 Rather, the argument from recognition signifies that the aimof punishment is situated in compelling the recognition of the victimas an equal by the wrongdoer: the aim of punishment is the creationof (in Seelmann’s words) “a legal relationship based on equality.”45

This means that the humiliation of the criminal, which goes alongwith the punishment, should only occur temporarily. Therefore, thepunishment is justified not in itself, but merely by its aim. If retaliationis to play a role here, then it should only play a role as an instrumentof rehabilitation.

Before one adopts the argument from recognition that has beenreconstructed in such a way, one must note that it lacks any basis in thisparagraph’s text; the relation of the argument to } 97 is not substanti-ated by Seelmann.46The argument from recognition sees disruption of“a reciprocally conditioned relation” in the nullity of the infringementof right.47 At least for right in the state as defined in the “Ethical life”section, this cannot be the case. In } 218 Hegel writes:

Since property and personality have legal recognition and validity incivil society, crime is no longer an injury . . . merely to a subjective infinite,but to the universal cause . . . whose existence . . . is inherently . . . stableand strong.48

43. Seelmann, “Versuch einer Legitimation,” 320.44. Fichte’s thought experiment is found in GNR i/4 59. Fichte, Foundations of natural

right, p. 226. For more on Fichte’s thought experiment, see Lazzari, “Eine Fessel,”pp. 173–86. Hegel alludes to this in GPhR } 100. Elements, ed. Wood, p. 126.

45. Seelmann, “Versuch einer Legitimation,” 320.46. It is substantiated neither in “Versuch einer Legitimation” nor in “Wechselseitige

Anerkennung.”47. Seelmann, “Versuch einer Legitimation,” 320.48. GPhR } 218. Elements, ed. Wood, p. 250.


In the state, right, getting legal validity, is no longer dependent onrecognition on the part of individuals. If the crime would not becancelled by a “second coercion” then, indeed, it would constitute aprecedent that might call into question the validity of right.49 In sucha case, it would be not only the criminal’s legal status – as in Fichte –that would be at risk, but also – as in Fichte – the legal relationship ofall persons with one another. Moreover, the crime does not necessar-ily imply the non-recognition of a legal person by the criminal. In hisRemark on } 95, Hegel draws attention to the negation of the legalcapacity of a person as only representing “shapes” of the crime. Hegelmakes mention of “further development . . . and . . . further shapes,”for which he provides the examples of “perjury” and “counterfeiting,forgery, etc.,” and thus of crimes that are not directed against a singleperson but rather against the state. Furthermore:

The substantial element within these forms is the universal, whichremains the same in its further development and in the further shapesit assumes; thus its infringement, i.e. crime, also remains the same, inconformity with its concept.50

First and foremost, the crime’s “nullity” can be explained in anotherway. According to Hegel, “force or coercion immediately destroys itselfin its concept, since it is the expression of a will which cancels . . . theexpression or existence of a will.”51 The crime’s nullity relies on thewill itself being destroyed in the crime. Thus, according to Hegel,the crime consists in a self-contradiction rather than in an infringe-ment upon reciprocity.

5.2.3. A mixed theory of punishment. Another twofold objective andsubjective justification of punishment is also presented on the modelof Kant’s penal law: a mixed theory (see Section 2.1). Such a mixedtheory of punishment consists of ascribing to retributivism and to thetheory of general deterrence each a part of the task of justification. Inthis way, Mohr wishes to recognize “a theoretic instrument [that] doesprofitable preliminary work underlying the current claims to a mixedtheory.”52 According to Mohr, the Elements combine a retributivist

49. GPhR } 93. Elements, ed. Wood, p. 120.50. GPhR } 95, Remark. Elements, ed. Wood, p. 122.51. GPhR } 92, Remark. Elements, ed. Wood, p. 120.52. Mohr, “Unrecht und Strafe,” p. 122.


rationale of punishment as an institution with a general deterrentspecification of the degree of punishment.53

Such an interpretation sees the element of deterrence theory in“Ethical life” (especially in } 218), whereas it locates the retributivistelement in “Abstract right.” Therefore, it inverts that interpretationof Kant’s penal law as a mixed theory according to which deterrencetheory justifies punishment as being an institution and retaliationtheory justifies the degree of punishment. If Kant and Hegel are reallymixed theorists, then Hegel is thus the only one of the two whojustifies retributivistically the existence of the punishment itself.

A mixed interpretation has to ask itself the question, however,about the compatibility of both of its components. Let us call to mindagain the results of Chapter 2: with Kant’s conception, as it is generallyknown, one cannot avoid a collision between general deterrencein the execution of punishment and the categorical imperative’sapplication dictates that the humanity in the criminal should notbe treated as a means, but always only as an end. General deterrence,which is mixed with retributivism, does not solely concern the threatof punishment. Kant’s (controversial) implicit assumption presup-poses that the mere threat of punishment treats no one as a meremeans, since, merely by threat, no harm is inflicted on anyone so longas this person does not commit a crime. Whoever commits a crimewould then not be punished according to the degree dictatedby general deterrence but rather according to the degree dictatedby retributivism. Hence only the threat of punishment – not thepunishment itself – is justified using general deterrence. Thus, boththe existence of punishment and the degree of punishment remaingrounded in a purely retributive way. That is due to Kant not allowingthe degree of punishment to be defined other than retributively.In Kant, the threat of a non-retributivist degree of punishment isnot allowed to lead to a non-retributivist degree of punishmentduring the execution of the punishment, which would contradictthe categorical imperative. Mixed theory’s interpretation, hence, isonly in appearance a mixed theory – in reality it remains a retributivistjustification of punishment.

Now how does all of this look in Hegel’s alleged mixed theory? ToMohr, } 218, with its character of general deterrence, is “a furtherrelevant angle for the apportionment of punishment appended by

53. Mohr, “Unrecht und Strafe,” p. 122; see also Section 5.1.


Hegel to his conception of punishment first in the theory of civilsociety.”54 The possibility of a mixed theory presupposes that thisrelevant angle is compatible with the alleged retaliation from“Abstract right.” In regard to retaliation in the way it is meant by Kant,this would obviously not be the case.

Hegel defines danger in relation to the “inherently . . . stable andstrong” “existence” that the law should have, that is, in relation tocompliance with the law.55 The more the crime endangers generalcompliance with the law, the more strictly the crime must be pun-ished. If no state exists, the danger is, by definition, “infinite” and therevenge must then be in equal measure to the danger.56 If “the powerof society has now become sure of itself” then the punishments could,and should, become much more lenient. Such a definition of punish-ment prescribes deterring further crimes as the aim of punishment.One cannot see, in Hegel’s view, any retributivism unless one wereto conceive so broadly of the definition of retributivism that theproportionality, or the parity, between crime and punishment wouldend up amounting to the parity between danger and punishment.By so doing, though, the retributivist thesis actually would becomeindistinguishable from the deterrent thesis.

Conclusion: a Hegelian mixed theory of punishment, in reality,is just as improbably a true mixed theory as can be the supposedKantian mixed theory. A Kantian mixed theory remains fundamen-tally retributive; a Hegelian mixed theory ultimately remains a theoryof deterrence.

5.3. Objections to the retributivist interpretations

If one can find in Hegel’s theory of punishment neither a twofoldjustification nor a mixed theory, but instead only an objective inter-pretation, which draws either on retributivism or on a deterrenttheory, then the retributivist reading seems likely. Therefore, it isthe traditional and most supported reading. Now, however, compati-bility of the classical principle of retaliation with the Hegelian conceptof right proves itself to be problematic in at least two ways.

54. Mohr, “Unrecht und Strafe,” p. 118.55. GPhR } 218. Elements, ed. Wood, p. 250.56. GPhR } 218, Remarks. Elements, ed. Wood, pp. 250f.


5.3.1. Hegel’s critique of talion law concerns more than its literal application.In the first passage in which Hegel refers to the theory of retaliation,this occurs solely with reference to the argument of the negation ofthe negation57 as well as in relation to taking into account the crime’squantity and quality in the determination of the negation of thenegation:58

The cancellation . . . of crime is retribution in so far as the latter, by itsconcept, is an infringement of an infringement, and in so far as crime,by its existence . . . has a determinate qualitative and quantitativemagnitude, so that its negation, as existent, also has a determinatemagnitude.

Immediately thereafter, Hegel expressly rejects classical retaliation asbeing “equality in the specific character of the infringement.” Hegelwrites: “this identity [of crime and retribution], which is based on theconcept, is not an equality in the specific character of the infringe-ment, but in its character in itself – i.e. in terms of its value.”59 What“this identity . . . which is based on the concept” or this “equality in thespecific character of the infringement” can possibly mean is eluci-dated by Hegel in the Remarks:

It is only in terms of this specific shape [specific equality] that theftand robbery . . . and fines and imprisonment etc. . . . are completelyunequal, whereas in terms of their value, i.e. their universal character asinjuries . . . they are comparable.60

Therefore, the only “retaliation” (Wiedervergeltung) that Hegelannounces himself as supporting is nothing other than the infringe-ment of the infringement.

Some interpreters are of the opinion that Hegel’s critique isdirected only against a literal application of the classical theory ofretaliation. According to these interpreters, Hegel’s critique can bereduced to these lines: “the sphere of externality, in which no absolutedetermination is in any case possible”; “in the realm of finite things, theabsolute determination remains only a requirement, on which theunderstanding must impose increasing restrictions . . . [which admits]only an approximate fulfilment.” This aspect of Hegel’s critique surelyexplains the “insuperable difficulty” for the thesis of “specific equality”

57. GPhR }} 93 and 97. Elements, ed. Wood, pp. 120–1 and p. 123.58. GPhR } 96. Elements, ed. Wood, pp. 122–3.59. GPhR } 101. Elements, ed. Wood, p. 127.60. GPhR } 101, Remarks. Elements, ed. Wood, p. 129.


to determine punishments (especially if psychology also invokesthe strength of sensuous motives . . . and, as a corollary, either thecorrespondingly greater strength of the evil will or – if we prefer – thecorrespondingly lesser strength and freedom of the will in general).61

If, however, one reads Hegel’s critique of classical retaliation in anunabridged form then one must particularly take into account thesecond half of Hegel’s argument:

If we not only overlook this nature of the finite realm but [emphasismine] also proceed no further than abstract and specific equality, aninsuperable difficulty arises when we come to determine punishments. . . Furthermore [emphasis mine], it is very easy to portray the retributiveaspect (Wiedervergeltung) of punishment as an absurdity (theft asretribution for theft, robbery for robbery, an eye for an eye, and atooth for a tooth, so that one can even imagine the miscreant asone-eyed or toothless); but the concept has nothing to do with thisabsurdity, for which the introduction of that [idea of] specific equality

is alone to blame.62

Thus the absurdity lies not in an erroneous, literal determination, butrather in the axiom that “what the criminal has done should also happen tohim,” and therefore in the specific equality or in talion law in contrastto the Hegelian principle of the negation of the negation.63

It should be noticed that in this context Hegel, from the outset,does not ascribe to this axiom of classical retaliation the status of aphilosophical determination of the concept of right. He introducesthe remarks on } 101 as follows:

It is usual in science for a determination – in this case, that ofpunishment – to be defined in terms of the universal representations . . .of conscious psychological experience. In the present case, thisexperience would indicate that the universal feeling of peoples andindividuals towards crime is, and always has been, that it deserves to bepunished, and that what the criminal has done should also happen to him.64

Primoratz maintains that this “attitude” of “reciprocity” is the veryreciprocity of a “man towards punishment.” He annotates Hegel: “Onthis point, commonsense is correct, Hegel believes, and philosophy has

61. GPhR } 101, Remarks. Elements, ed. Wood, p. 128.62. GPhR } 101, Remarks. Elements, ed. Wood, p. 128.63. GPhR } 101, Remarks. Elements, ed. Wood, p. 127.64. GPhR } 101, Remarks. Elements, ed. Wood, p. 127.


to abide by the judgment of common sense.”65 This is contradicted notonly by the fact that Hegel has time and again either strictly differenti-ated between common sense philosophy and the mere science ofunderstanding or expressly wanted to differentiate between the two.One of themany points of Hegel’s critique of common sense and of thescience of understanding is that they become entangled in contradic-tions from which they attempt in vain to loosen themselves, for theycannot cancel these very contradictions philosophically. Again, in thepassage about crime Hegel continues:

It is incomprehensible how those sciences which derive theirdeterminations from universal representations . . . should on otheroccasions accept propositions which contradict such so-calleduniversal facts of consciousness.66

Hegel presumably means by this those “various theories of punish-ment as prevention, as a deterrent, a threat, a corrective, etc.” whichpresuppose, in the punishment’s rationale, the “superficial characterof an evil.”67 Hegel sharply criticizes them:

As a result of these superficial points of view, however, the objectiveconsideration of justice, which is the primary and substantial point ofview in relation to crime, is set aside; it automatically follows that theessential consideration is now the moral point of view, i.e. the subjectiveaspect of crime, intermixed with trivial psychological ideas.68

I proceed under the assumption that the contradictory “proposi-tions” that Hegel criticizes in the Remark on } 101 are those verytheories I have described above, because both these and retributivismwere already then the two dominant competing theories of punish-ment: punishment was imposed either because the crime was committed(retributivism), or in order that something occur (rehabilitation, eithergeneral or specific deterrence; all following a consequentialist intent).Sometimes, these are referred to as “retrospectively” or “prospectively”directed theories.

Hegel raises the same objection against retributivism as he raisesagainst the various consequentialist theories of punishment, that is,that they rely on common sense, that they are subjective rationalesand that they contradict one another.

65. Primoratz, “Banquos Geist,” 37.66. GPhR } 101, Remarks. Elements, ed. Wood, p. 127.67. GPhR } 99, Remarks. Elements, ed. Wood, p. 124–5.68. GPhR } 99, Remarks. Elements, ed. Wood, p. 125.


It is well known that, whenever the theories of understandingcontradict one another and the dispute seems to be irresolvable,Hegel’s “science of reason” consists in canceling (aufheben) thesecontradictions.69 In this cancellation, the opposing theories provethemselves to be merely partial insights into the relevant concept.In the question of penal law as well, Hegel proceeds exactly accordingto this method in the Elements of the philosophy of right. Both retributi-vism and consequentialism are categorized expressly as partialinsights into the concept of punishment by Hegel so that both, bytheir claim that they provide a rationale of punishment, distortthat very concept of punishment. Hegel says of the consequentialisttheories of punishment that

the various considerations which are relevant to punishment as aphenomenon . . . and to its relation . . . to the particular consciousness,and which concern its effect on representational thought (as a deterrent,corrective, etc.), are of essential significance in their proper context,though primarily only in connection with the modality of punishment.But they take it for granted that punishment in and for itself is just. Inthe present discussion, we are solely concerned with the need to cancel(aufheben) crime – not as a source of evil, but as an infringement of rightas right – and also with the kind of existence which crime possesses, whichmust also be cancelled.70

About retributivism, Hegel states: “It is this inner identity which, for theunderstanding, is reflected in external existence (Dasein) as equality.”71

By “inner identity” is meant the infringement of the infringement; by“equality” is meant classical retaliation.

Before I move on to the Hegelian cancellation of the oppositionbetween both of these theories in the concept of punishment, there isa point that contradicts the assumption of some interpreters, statingthat the classical theory of retaliation and the Hegelian penal theoryare compatible, to which I have yet to attend.

5.3.2. The negation of the negation is not retaliation. The aforementioned“negation of the negation” is not to be understood as being thenegation of an opposing element, as in the case that A is the negationof :A, and therefore it is ::A. For Hegel, the negation of the negationis rather the cancellation of the negation, that is, the cancellation of

69. Cf., for example, Hegel’s Encyclopedia of the philosophical sciences, } 36, Addition.70. GPhR } 99, Remarks. Elements, ed. Wood, p. 125.71. GPhR } 101, Remarks. Elements, ed. Wood, p. 128.


coercion. It is not ::A, but rather A. Yet it is not the absolute A, butinstead an A having validity by the cancellation of the negation.

Naturally, by the negation of the negation, :A becomes negated.That :A becomes negated does not mean, however, that the negationis on the whole negated. Now, right requires that the negation itselfbecomes negated, and therefore that no violence results from thepunishment. The punishment is the special negation actually bringingto termination the very act of negating in general.

Now, how does this relate to classical retaliation, according to which“what the criminal has done should also happen to him”?72 The criminal hasnegated the personality (A) of his or her victim: :A. Happening to thecriminal what he or she him- or herself has done to the victim meansthe negation of the personality (B) of the criminal: :B. The relation-ship between :A and :B is naturally that B executed the action :A.Through :B, :A was indeed negated but B was also negated. There-fore, :B must be negated. In other words: whoever is responsible formaking sure that what the criminal has done happens to the criminalhim- or herself does exactly what the criminal has done. Therefore,what the criminal has done ought also to happen to him or her(whoever is responsible, that is). Retaliation reproduces the negationinstead of canceling it. Hegel expresses this in the following way:

In the sphere of the immediacy of right, the cancellation . . . of crime isprimarily revenge, and its content is just so far as it constitutes retribution. . . Thus revenge, as the positive action of a particular will, becomesa new infringement; because of this contradiction, it becomes part of aninfinite progression and is inherited indefinitely from generation togeneration.73

Hegel clearly draws the consequence from this opposition betweenthe negation of the negation and the classical theory of retaliation.

To require that this contradiction [between the negation of the crime andits multiplication] . . . should be resolved . . . is to require a . . . punitiverather than an avenging justice. Primarily, this constitutes a requirementfor a will which, as a particular and subjective will, also wills the universalas such.74

72. GPhR } 101, Remarks. Elements, ed. Wood, p. 127.73. GPhR } 102. Elements, ed. Wood, p. 130. This progression of revenge is elucidated by

the Addition to } 102 with the censure of “undying” “revenge” among “uncivilizedpeoples” that propagates itself as revenge and then revenge for that revenge fromgeneration to generation.

74. GPhR } 103. Elements, ed. Wood, p. 131.


A common way of interpreting this passage involves reading it asmerely being a requirement for impartiality of the judge against therisk of an interested – and thus partial – private justice. However, thisimplies an inadmissible abridgment. For even if this private justicecommitted itself to impartiality, the problem that Hegel addressedwould still exist. The problem consists namely in that the victim “place[s][his or her] infinity in any infringement . . . which occurs, and whosejustice is therefore altogether contingent.” In this sentence, Hegeldoes not deny that the revenger acts justly. The “content” of revengeis, he writes, “just so far as it constitutes retribution,” and Hegel speaksof “[its] justice.”75 Therefore, this sentence is not concerned with therisk of the misuse of punishment. It is concerned with the right that isto be restored.

Whereas in revenge two persons stand opposed to one another,who are both to be seen as being the existence (Dasein) of the will andtherefore are allowed to make a claim on the right of personality, inthe state every person is subject to the law, in which inheres theobjective right of universality. Before the law, the person has onlythe subjective right of personality, which implies, for instance, thatthe state is not allowed to handle him or her arbitrarily. The conse-quence is that that which is to be restored as right is, in the case ofrevenge, the personality of the criminal but in the case of punishmentthe validity of the law.76

Now the victim’s personality does not directly contain the person-ality of the criminal, whereas the law, by definition, recognizes thepersonality of all members of the commonwealth. Hegel says of thepunitive justice of the state: “Instead of the injured party, the injureduniversal now makes its appearance.” Hegel places the subjective andcontingent revenge in contrast with the punishment:

Objectively, this reconciliation applies to the law, which restores andthereby actualizes itself as valid through the cancellation (Aufheben) of thecrime; and subjectively, it applies to the criminal in that his law, . . . whichis known by him and is valid for him and for his protection, is enforced uponhim in such a way that he himself finds in it the satisfaction of justiceand merely the enactment of what is proper to him.77

75. GPhR } 103. Elements, ed. Wood, p. 130.76. This differentiation is lacking, in my view, in Axel Honneth, The struggle for recognition:

the moral grammar of social conflicts, trans. Joel Anderson (Cambridge, Mass.: MIT Press,1996), p. 21.

77. GPhR } 220. Elements, ed. Wood, p. 252.


What Hegel calls “retribution” (Wiedervergeltung), which is “an infringe-ment of an infringement,” should not be confused, therefore, with theretaliation (Wiedervergeltung) of classical retributivism.78

In the degree of punishment as well, a difference now arisesbetween the restoration of the personality of the individual – or of agroup of individuals – and the restoration of the law. Restoring thelaw, that is, to reaffirm its validity, is completely feasible. Restoring theexistence (Dasein) of the will in an individual person in which it wasfound before the crime is, on the contrary, only possible in somecases. These cases are treated in } 98 of the Elements of the philosophyof right: “The cancellation . . . of the infringement, where the latter hascaused damage, is civil satisfaction in the form of compensation (in sofar as any compensation is possible) [second emphasis mine].”79 In mostof the cases, though, there is no compensation to be found. And inthose cases as well, wherein compensation is possible, the “positiveexistence of the injury,” that is, “the particular will of the criminal,” remainsunaffected: the criminal can provide the whole compensation whilekeeping his or her criminal intentions.80

Some authors attempt to justify the classical principle of retaliationnot merely by appeal to reciprocity in general, but also in other ways.Jean Hampton mentions, for example, the explanation that talion lawhas the aim of destroying the criminal’s claim permanently to acquiresupremacy over his or her victim. Poking out the criminal’s eye is,according to this conception, a way for the victim, who was mademonocular by the criminal, to revoke that superiority unjustly gainedby the criminal. The question of the rightness of this and of similarexplanations remains to be addressed (Hampton herself rejects it).Bear in mind, though, that Hegel does not enter into such interpret-ations of classical retaliation theory.81 In its place, Hegel adopts asymbolic interpretation: “It is this inner identity which, for the under-standing, is reflected in external existence . . . as equality.” In revenge,the restoration of right cannot really occur: therefore, it occurs sym-bolically – that is, as talion law, as a specific equality: “robbery forrobbery, an eye for an eye, and a tooth for a tooth.”82

78. GPhR } 101. Elements, ed. Wood, p. 127.79. GPhR } 98. Elements, ed. Wood, p. 124.80. GPhR } 99. Elements, ed. Wood, p. 124.81. Cf. John Kleinig, “Punishment and moral seriousness,” Israel Law Review, 25, no. 3

(1991), 401–21 (p. 416).82. GPhR } 101, Remarks. Elements, ed. Wood, p. 128.


The “absurdity” of such “punishments” consists of their havingnothing to do with the restoration of right, with the termination ofthe coercion and with reconciliation – a reconciliation with the crim-inal as well. On the contrary, such punishments do not treat thecriminal as an existence of freedom. Only in so far as the criminalis a member of a commonwealth is the preservation of the will’sexistence in the criminal’s person necessary and required for therestoration of a commonwealth to which he or she belongs and bywhich his or her rights are protected as well. The criminal’s forcedamputation is not necessary for such a restoration of the validityof law. Rather, amputations contradict the idea of a commonwealthof free persons. For that reason, amputations are “absurdit[ies].”83

5.4. Reconstruction of the Hegelian legaljustification of punishment

The grounding of punishment in Hegel consists of the verificationthat the punishment is a necessity of right. The keystone of thenormative construction of punishment, namely the “second coercion”or that which “cancels an initial coercion” (we will see later how thismight be understood), is “not only conditionally right but necessary”for right.84 Punishment, therefore, is not only in compliance withright by itself; rather, without it there would be no right. In contrastto that, the initial coercion is “contrary to right.” Hegel’s grounding ofpunishment does not separate the authority to punish from the dutyto punish; initially, it does not attempt to demonstrate first thatpunishment may occur and then, after that, that punishment shouldoccur, but instead it fulfills both tasks together.

Hegel substantiates this necessity in the following way: “Becausecoercion destroys itself in its concept, it has its real expression . . . inthe fact that coercion is cancelled . . . by coercion.”85 This thesis of theconceptual self-destruction of coercion is not to be taken as a physicalthesis. It does not concern, for instance, the physical self-destructionof coercion; instead, what matters is that this coercion is “contrary toright.”86 Contrariness to right is the same as the concept of legal

83. GPhR } 101. Remarks. Elements, ed. Wood, p. 128.84. GPhR } 93. Remarks. Elements, ed. Wood, p. 120.85. GPhR } 93. Remarks. Elements, ed. Wood, p. 120.86. GPhR } 92. Remarks. Elements, ed. Wood, p. 120.


impossibility.87 The coercion destroying itself means that coercion is“rightly” impossible or incompatible with the concept of right. Inthe self-destruction two elements must be distinguished: (1) coercionis destroyed; (2) it itself carries this destruction through: it is a self-destruction. The first coercion is canceled by the second coercion;important here is to realize that it is canceled by nothing else other thancoercion.

In the second regard, coercion contrary to right differentiates itselffrom the infringements it creates, that is, from the “evil” or the“damage,” because “when an infringement of right occurs, it doeshave a positive external existence . . . but this existence within itself isnull and void.”88 Even if the infringement being a consequence ofthe coercion has a physical existence – that is, as damage – it actuallydoes not have any existence in right at all. This infringement is notdestroyed by itself.

The difference between the self-destruction of coercion and theextrinsic destruction of damage comes to expression in the form ofthe differentiation between the “cancellation” (in right) of coercionand the “nullity” (in right) of damage.89 As is generally known, theAufhebung (“cancellation”) of contradictions, which is a central con-cept in the Hegelian dialectic, means not only destroying whatevermay contain a contradiction, but also simultaneously preserving some-thing of it. Therefore, something is preserved from coercion by itsdestruction. We must take note of this point for later.

Many interpreters do not know how to make sense of self-destruction of coercion or regard it as being nonsensical. This incom-prehension often consists of overlooking the premises upon whichthe thesis of self-destruction rests. I shall provide the original textfrom Hegel’s Elements in order to make reference to the grammaticalstructures in the German original, which I also emphasize:

[Because] the will is Idea or [is] actually free only in so far as it hasexistence . . . and the existence in which it has embodied itself is thebeing of freedom, [then] force or coercion immediately destroys itselfin its concept.

87. There is more at stake here than either the credibility of the legal system or the honestyof the civil authorities. David Cooper believes this to be Hegel’s core argument. DavidCooper, “Hegel’s theory of punishment,” in Zbigniew A. Pelczynski (ed.),Hegel’s politicalphilosophy (Cambridge: Cambridge University Press, 1971), pp. 151–67.

88. GPhR } 97. Remarks. Elements, ed. Wood, p. 123.89. GPhR }} 93 and 97, respectively. Elements, ed. Wood, pp. 120 and 123.


Weil der Wille, nur insofern er Dasein hat, Idee oder wirklich frei unddas Dasein, in welches er sich gelegt hat, Sein der Freiheit ist, so zerstortGewalt oder Zwang in ihrem Begriff sich unmittelbar selbst.90

The fundamental premise here – as in the Elements of the philosophy ofright on the whole, actually – is that free will must develop; and thisdevelopment is what Hegel terms “right.” I shall here content myselfwith calling to mind a few central statements from the introductoryparagraphs of the Elements: “The subject-matter of the philosophical scienceof right is the Idea of right – the concept of right and its actualization.”91

The basis . . . of right is the realm of spirit in general and its preciselocation and point of departure is the will; the will is free, so thatfreedom constitutes its substance and destiny . . . and the system ofright is the realm of actualized freedom.92

And: “Right is any existence . . . in general which is the existence of thefree will. Right is therefore in general freedom, as Idea.”93 Right havingto be plays the role for Hegel that is played by the categorical impera-tive for Kant: the establishment of a state of law is a necessity.

The existence of freedom required by the idea of right consists inthe “abstract right” of the individual body and life as well as ofpersonal property. Hegel notes that

free will in and for itself cannot be coerced . . . except in so far as it fails towithdraw itself from the external dimension in which it is caught up, or fromits idea . . . of the latter . . . Only he who wills to be coerced can be coercedinto anything.94

It should be noted that “free will” must want to allow itself to becoerced by the criminal because, according to the idea of right, it musthave an existence: in terms of right it cannot withdraw from external-ity. Right forbids the victim from a stoic retreat into internality inorder to escape the criminal’s coercion.

In “Ethical life” it is not the individual but the state that displays thehighest existence of free will. However, the grounding of penal lawin abstract right holds for those criminals whose direct victims are

90. The emphases are mine. GPhR } 92. The original text can be found in Hegel,Grundlinien der Philosophie des Rechts. Elements, ed. Wood, p. 120.

91. GPhR } 1. Elements, ed. Wood, p. 25.92. GPhR } 4. Elements, ed. Wood, p. 35.93. GPhR } 29. Elements, ed. Wood, p. 58.94. GPhR } 91. Elements, ed. Wood, pp. 119–20. Regarding property, cf. GPhR } 90

(“property”) (p. 119) and } 91 (“his physical side and other external attributes”).


individuals, as well as for those whose direct victim is the state.95 Now,body, life and property of the individuals are all an existence ofthe free will. Therefore, the body, life and property of the criminalare just as much the existence of the free will as the body and life ofthe victim. Therefore, self-destruction in } 92 ought to be understoodin the following ways:

1 One must understand self-destruction as the thesis of the necessarydestruction of that which contradicts right. Since the realization ofright – and along with it as well, the existence of the free will in thebody, in the life and in the property of the victim – is a necessity, thecrime, being an attempt to infringe on this existence, can only failand lead to the exercising of coercion against the criminal.

2 Self-destruction must also be understood as the thesis of the identityof the criminal’s free will with the victim’s free will. The criminalcoerces or destroys the existence of free will in the person of thevictim and thereby coerces or destroys him- or herself, since thecriminal’s own existence is also the existence of free will.

The two theses are linked to each other, since right must have itsexistence in every person, so that, for right, the personality of thecriminal and that of the victim are equal. In Hegel, the refusal torecognize the personality of all the other humans implies a discrep-ancy on the part of that person who denies the recognition: on theone hand, the universal element in the person’s will is one with theuniversal element in the wills of the other persons; on the other hand,this universality is alien to it. Its particular will and its general willdiverge. This implies that the human who does not recognize thestatus of the personality in other humans lacks the foundation uponwhich he or she can base his or her own status as a person. Accordingto Hegel, those societies in which the personality of only somehumans is recognized (for instance, Egypt or the Greece of antiquity)are societies in which the personality is something contingent.

Therefore, the necessity of right requires the self-destruction ofcoercion. Now, this requirement is still “taken in the abstract” (thatis, abstractly understood) as long as it has not come to realization:“force or coercion immediately destroys itself in its concept [emphasis

95. Cf. GPhR } 95, Remarks. Elements, ed. Wood, p. 122: “the determination which will beconsidered in the following paragraph also applies to the particular and furtherdetermined content [of crime], e.g. in perjury, treason, counterfeiting, forgery, etc.”


mine].” According to Hegel, punishment is the “real expression” ofthe conceptual destruction of coercion.96 The necessity of right –meaning, the requirement that right realizes itself because it mustrealize itself – demands “that coercion is cancelled . . . by coercion.”97 Hegelalso terms the realization or the implementation of the requirementof right the “manifestation” of this requirement; “the punishment is. . . the manifestation of the crime.”98

How can coercion cancel coercion? Hegel gives the following defin-ition of coercion: in coercion’s externality, the will “may either experi-ence force in general, or it may be forced to sacrifice or do somethingas a condition of retaining some possession or positive being, therebysuffering coercion.”99 There are three aspects to coercion. Coercion is:

1 the exercise of force upon a person,2 in order to make the person act differently or sacrifice something,3 bymaking this action or this sacrifice a precondition for keeping his

or her body, life or property.

Whereas the thesis of self-destruction in } 92 relates both to coercionand to mere force, and “initial” coercion, that is, the coercion thatis contrary to right (the crime), can either be coercion or force,100

Hegel in } 93 merely mentions the cancellation of coercion bycoercion, and, with the mention of coercion in the Remarks to } 93,the cancellation of force by coercion. Nowhere does Hegel speak ofthe cancellation of either coercion or force by force. In the light of thedefinition of force as mere suffering, cancellation by force is explicitlyexcluded by Hegel by his criticism of Ernst Ferdinand Klein’s concep-tion of punishment as being an arbitrary cancellation of an evilthrough an evil:

If the crime and its cancellation . . . which is further determined aspunishment, are regarded only as evils . . . in general, one may wellconsider it unreasonable to will an evil merely because another evil is

already present.101

96. GPhR } 93. Elements, ed. Wood, p. 120.97. GPhR } 92. Elements, ed. Wood, p. 120.98. GPhR } 97. Elements, ed. Wood, p. 123; and } 101, Addition. Elements, p. 129.99. GPhR } 90. Elements, ed. Wood, p. 119.100. Cf. GPhR } 93, Remarks. Elements, ed. Wood, p. 120: “initial coercion, or at least force.”101. GPhR } 99. Elements, ed. Wood, p. 124. Regarding Klein: Hegel is referring to Ernst

Ferdinand Klein, Grundlinien des gemeinen deutschen und preußischen peinlichen Rechts(Halle: Hemmerde und Schwetschke, 1796), }} 9–10.


At least three consequences result from this.

1 It begs the question of the applicability of the cancellation ofcoercion by a coercion exercised upon the criminal, if the criminaldoes not allow him- or herself to be forced. We will come back tothis question later.

2 The death penalty obviously does not “coerce” the criminal, that is,it does not move the criminal to perform an action or a sacrifice butexercises mere force against him or her. Hegel’s position onthe question of the death penalty appears to be inconsistent. Onthe one hand, he maintains that murder “necessarily incurs thedeath penalty.”102 Admittedly he writes that it is to Beccaria’s creditthat the death penalty was repealed for most crimes. Hegel himselfwants to keep the death penalty for special cases, especially withmurder in mind:

Even if neither Joseph II nor the French have ever managed to secureits complete abolition, people have begun to appreciate which crimesdeserve the death penalty and which do not. The death penalty hasconsequently become less frequent, as indeed this ultimate form ofpunishment deserves to be.103

On the other hand, Hegel really sees alternatives to the death pen-alty. Because punishment must coerce the criminal, whereas thedeath penalty does perfectly fulfill the personal wishes of some tobe sentenced to death, then Hegel must actually allow some alterna-tive punishments, even for murder. A handwritten remark in his copyof the Elements on } 99 notes that “punishment has to be sensitive . . .It even has occurred that murders [were committed] in order to lose[one’s] life, – therefore, [they are] insensitive to the death penalty; –then it came to be, that it was changed to imprisonment.” In theLecture on the philosophy of right (1824/5) he says in a somewhat moredetailed way:

There have been cases in which a murder occurred in order [forone] to be sentenced to death. The murderer did it out of worldweariness, contempt for life, though especially in a religious sense . . .Thus the death penalty does not affect him – he already wills to leavethis life – consequently, a sentence of life has been replaced withprison in order to affect the will of the criminal.104

102. GPhR } 101, Addition. Elements, ed. Wood, p. 129.103. GPhR } 100, Addition. Elements, ed. Wood, p. 127.104. PhR 285.


In the Elements of the philosophy of right, there is no argument to befound against substituting alternative punishments for the deathpenalty for all crimes. On the contrary, an adherence to the deathpenalty would lead to another definition of punishment, namely tocancellation of coercion or (for example, in the case of murder) offorce by (mere) force. Moreover, this new definition would contra-dict Hegel’s theory in some ways. I have already mentioned hiscriticism of Klein. Also there is his criticism of a penal theory thatFichte once took into consideration. A cancellation of coercion orof force by force, according to Hegel, amounts to the conception ofpunishment in which a criminal “is regarded simply as a harmfulanimal which must be rendered harmless.”105 This theory of pun-ishment that Fichte deliberated upon – and that he rejected –would not treat the criminal as a rational being, according to Hegel.For this reason, he discards it.

3 The conceptual self-destruction of coercion and the real cancella-tion of coercion by coercion seem to imply the same conceptual self-destruction and real self-cancellation. It is an obvious temptation toconclude that there is a self-cancellation of coercion, whichcould possibly be an argument in favor of retributivism. But Hegeldoes not actually do this; instead, he refers to a conceptual self-destruction of force.106 Apart from this, we have just seen that a self-cancellation of mere force by mere force is out of the question, so thatonly a cancellation of mere force by coercion is conceivable.

Notwithstanding the question whether the crime was mere force orcoercion, what matters is that the cancellation of the crime happensthrough coercion. As we have seen, coercion is (1) the exercise offorce upon a person, (2) in order to make the person act differently orsacrifice something, (3) by making this action or this sacrifice a pre-condition for keeping his or her body, life or property. If this defin-ition is applied to the “second coercion” – to the coercion against thecriminal – then from these three elements result three implications.The cancellation of the crime is: (1) an evil – or force – that isexercised against the criminal, (2) so that the criminal acts differentlyor sacrifices something, (3) by making this altered deed or this sacrificeto be a precondition for keeping his or her body, life or property.

105. GPhR } 100, Remarks. Elements, ed. Wood, p. 126.106. Cf. GPhR } 92. Elements, ed. Wood, p. 129: “Consequently, force or coercion

immediately destroys itself in its concept.”


5.5. Integration of the other penal theories

A threefold objection can be made against such an interpretation:first, Hegel’s criticism of Klein’s conception of punishment as beingthe infliction of an evil; secondly, Hegel’s criticism of the penal reformtheory;107 thirdly, the fact that Hegel nowhere – at least nowhere in“Abstract right” – makes the criminal’s rehabilitation an issue.

The first two objections misunderstand Hegel’s criticism of Kleinand of the reform theory. Hegel’s criticism is directed exclusivelyagainst absolutizing evils and reform. Hegel writes:

If the crime and its cancellation, which is further determined aspunishment, are regarded only as evils in general, one may wellconsider it unreasonable to will an evil merely [emphasis mine] because

another evil is already present.108

Hegel’s critique only maintains that the aim of punishment, whichserves to ground punishment, cannot be to inflict an evil on thecriminal, although the criminal has actually inflicted an evil on hisor her victim. It is for this reason that Hegel speaks of a “superficialcharacter of an evil”: the evil is not a primary but – at most – aderivative aim of punishment. As regards the reform theory, it isequally discarded only as an absolutization of reform. Hegel writes:

The various considerations which are relevant to punishment as aphenomenon . . . and to its relation . . . to the particular consciousness,and which concern its effect on representational thought (as a deterrent,corrective, etc.), are of essential significance in their proper context,though primarily only [emphasis mine] in connection with the modality ofpunishment. But they take it for granted that punishment in and foritself is just.109

As regards the theory of rehabilitation, it is expressly found in “Ethicallife.” “Retribution through revenge” is contrasted by Hegel to “genu-ine reconciliation of right”: “subjectively, it applies to the criminal inthat his law, which is known to him and is valid for him and for hisprotection.”110 As early as in “Abstract right,” Hegel is saying of thecancellation of crime: “it is at the same time his will as it is in itself,

107. For both points one and two cf. the Remarks in GPhR }} 99 and 100.108. GPhR } 99, Remarks. Elements, ed. Wood, p. 124.109. GPhR } 99, Remarks. Elements, ed. Wood, p. 125.110. GPhR } 220. Elements, ed. Wood, p. 252.


an existence of his freedom, his right.”111 The punishment, being as itis just, guarantees the right, which is the existence (Dasein) of thecriminal’s free will; thus, punishment is the precondition of the fullmembership of the criminal. It should be noted once more thatthe criminal’s rehabilitation as a full member of the commonwealthpresupposes that punishment is just. Therefore, the rehabilitationaspect is also a secondary, derivative aim of punishment.

Hegel’s theory is not absolutist, that is, he does not support thethesis that punishment is in itself good and already justified becausethe criminal has deserved the punishment through his or her crime.In an absolutist theory of punishment, punishment is only groundedupon the criminal and what the criminal has merited; punishment isthen only seen as being a merited ill. We find good evidence in Kant’sexample of the island in which the sole justification of punishmentcited is: “so that each has done to him what his deeds deserve.”112 Therole of this example in the Kantian theory of right will not be dis-cussed here.113 Here I content myself with drawing attention to thefact that one searches in vain for a similar example in Hegel.

Hegel’s theory of punishment thoroughly assigns an aim to thepunishment: the cancellation of the crime. The cancellation ofthe crime by coercion has two unequal dimensions: the past and thefuture.

The cancellation of the crime’s past dimension clearly does not meanthat the crime would be made undone, that is, that everything were asif nothing had happened. At the most, some of the consequences ofthe crime can be eliminated by “civil satisfaction in the form ofcompensation (in so far as any compensation is possible).” In so far as“the damage amounts to destruction and is altogether irreparable,”“the universal character of the damage, as value, must in any case takethe place of its specific qualitative character.”114 To the extent towhich the criminal is thus coerced to make civil amends, the amendscan be placed in the category of cancellation of coercion by coercion.Hegel, however, finds this aspect of cancellation of crime to be ofmarginal importance. Civil amends is non-specific to the crime. In“collisions of right” or in the case of legal disputes that arise in civil law –in cases without any penal dimension – there is also often

111. GPhR } 220. Elements, ed. Wood, p. 126.112. RL Ak vi:333. Practical philosophy, ed. Gregor, p. 474.113. Cf., though, Section 1.3 above.114. GPhR } 98 and Remarks. Elements, ed. Wood, p. 124.


compensation as well.115 The criminal aspect is not situated in the evilor the harm. As I said before, Hegel indeed ascertains a “positive,external existence” in harm; it is, however, “within itself . . . null andvoid,” for the positive existence comprises not an existence of will but,instead, a “mere product” thereof.116 The criminal did not primarilyintend harm. The criminal primarily intended the realization of his orher freedom in a specific existence – in an alien good, in a largerfreedom through the elimination of other humans, etc. – by means ofthe exercise of a coercion on another will. At most, the criminalrecklessly disregarded the harm it might cause. The compensationhere is not to be confused with, for instance, the restitution of stolengoods to their owner. The question arises here of the appraisalof vandalism and of other crimes whose exclusive intent is destruction.A possible answer to this objection would be to see in these cases thecriminal’s disappointed will that does not receive the existence itdesigned. No matter how it stands with this objection, in any case forHegel the crime definitely belongs to the past: the harm correspondsneither to something in the will of the victim who ceased to becoerced by the criminal nor to something in the criminal who hadanother primary intention.

The future dimension of the cancellation of the crime is the specificdimension for the crime. Hegel observes: “The positive existence of theinjury consists solely in the particular will of the criminal.”117 The cancel-lation of the crime by coercion can now occur either during thecommission of the crime or after its commission.

The cancellation of the crime during its commission pertains to theright of coercion belonging to the classical ius strictum (the “abstractright – or right in the strict sense”).118 First and foremost, self-defensepertains to this immediate cancellation of the crime during its com-mission. By self-defense the criminal is coerced. Hegel mentions thistype of coercion – admittedly only briefly – which, in the strict sense, isneither a punishment (enacted by the state) nor (private) revenge.119

It is also nearly uncontroversial.This existence of the crime is to be seen not only as being the

purpose or intent of the crime during the crime’s commission, but also

115. GPhR }} 84–6. Elements, ed. Wood, p. 117.116. GPhR }} 97 and 99. Elements, ed. Wood, pp. 123, 124.117. GPhR } 99. Elements, ed. Wood, p. 124.118. GPhR } 94, Remarks. Elements, ed. Wood, p. 121.119. Cf. GPhR } 94. Elements, ed. Wood, p. 121.


as the present and past will of the criminal. Hegel’s presumption is thatthe “particular will of the criminal” continues to be a criminal willshould it remain unpunished. In his lecture Natural right and the scienceof state (Naturrecht und Staatswissenschaft) from 1818/19, Hegelexpressly says:

What the will does is, according to appearance, a single, but also auniversal – something temporal and something intemporal. – Thus theconsequences of the deed remain even if the deed is itself hidden.Thus whoever steals remains for ever a thief, not merely in memory butactually.120

And a little later he adds: “the deed of the criminal is not merelysomething transient, single, but instead something at the same timeuniversal, a law stating that infringement against somebody is allowed.”121

Therefore, if the punishment does not occur then the criminal againcommits a crime. According to this, in order that right should not beharmed again, the criminal must be coerced:

The positive existence of the injury consists solely in the particular will of the

criminal. Thus, an injury to the latter as an existent will is thecancellation . . . of the crime, which would otherwise be regarded as valid,and the restoration of right.122

Specific deterrence is therefore an aim of punishment that necessar-ily belongs to the primary aim of the cancellation of the crime. Generaldeterrence, however, is not only an indirect side effect of punishmentgrounded in such a way, but also necessarily itself an aim of punish-ment. It means: “If society is still inwardly unstable, punishments mustbe made to set an example, for punishment is itself a counter-exampleto the example of crime.”123

This general deterrent aim of the Hegelian theory of punishment isalso indirectly derived from the main aim of the cancellation of thecrime. With the deed, the criminal accepts all consequences that he orshe can foresee. These consequences that are internal to the criminalaction, under which there is the risk that the deed moves othercitizens to commit a crime, pertain to his or her criminal intent. Therelationship between the criminal’s will and these consequences,

120. NRSW 275–6.121. NRSW 276–7.122. GPhR } 99. Elements, ed. Wood, p. 124.123. GPhR } 218, Addition. Elements, ed. Wood, p. 251.


therefore, is closer than the relationship between the criminal’s willand the harm for which he or she is held responsible. The harm isnot necessarily foreseeable by the criminal: it partly pertains to theconsequences that are external to the criminal action. It follows fromthe criminal not only being held liable for the harm caused, but alsobeing punished for consequences internal to the criminal action, thatone must coerce the criminal, to whom belong exclusively the conse-quences internal to the criminal action and not those external to thecriminal action. The consequences internal to the deed must notnecessarily occur; they do, however, pose a risk (Hegel speaks of“danger”)124 that the criminal accepts. Hegel explains this throughan example in his lecture Philosophy of right:

A crime is dangerous, that is, it is primarily a crime, but it has a furthereffect, a further context. The single action is in itself so more farreaching, more universal than it appears according to its externalreality. If I hold a light to a piece of wood then the flame only touchesa small part of the surface. The arsonist does the same and leaves, thus,his crime is of a lesser extent; but, the wood is in relation to other woodthat together constitutes a house – this one with other houses thatconstitute a city, and thus the fire can destroy a city. This universalityis the danger of crime.125

This danger internal to the criminal action must be strictly differentiatedfrom danger external to the criminal action, if one wishes to avoid aconsequentialist misinterpretation of Hegel:

Various qualitative determinations [of crime], such as danger to public

security, have their basis in more precisely determined circumstances,but they are often apprehended only indirectly in the light of otherconsequences rather than in terms of the concept of the thing . . . Thus,the crime which is more dangerous in itself . . . in its immediatecharacter, is a more serious infringement in its extent or quality.126

Belonging to the particular will of the criminal is the risk that thecrime comes to be an example, therefore a precedent for a third party.The coercion exercised by punishment of the will of the criminal alsocancels this precedent, which has a general deterrent effect, that is,though only a secondary one.

124. GPhR } 218, Remarks. Elements, ed. Wood, p. 251. Also, cf. Ripstein, Equality,responsibility and the law, pp. 228–9.

125. PhR 279.126. GPhR } 96, Remarks. Elements, ed. Wood, pp. 122–3.


The various amounts of danger the criminal recklessly disregardsexplain the various severities of punishment, depending on the era inwhich the crime was committed. The severity of punishment conformsto how secure the commonwealth is or what the risk is for destabiliza-tion of the commonwealth by imitation of the crime. A prestate soci-ety, in which the punishment of crimes is still a private matter, or aweak commonwealth must both punish severely in order to preservethemselves; a secure commonwealth does not have to: it shouldpunish more mildly. Of one thing Hegel is sure, however: the com-monwealth can never be so sure of itself that it can allow itself to leavea crime unpunished. Hegel writes:

But whereas it would be impossible for society to leave a crimeunpunished – since the crime would then be posited as right – thefact that society is sure of itself means that crime, in comparison, isalways of a purely individual character, an unstable and isolatedphenomenon.127

That is because even if no other citizen imitates the criminal, thecriminal still will commit further crimes for as long as he or she isnot coerced to do otherwise.

Let attention be drawn to the fact that the aim of specific deter-rence is not itself some undetermined sort of specific deterrence. Hegel’sspecific deterrence is always formulated negatively: “cancellation . . . ofthe crime, which would otherwise be regarded as valid”; “it would beimpossible for society to leave a crime unpunished – since the crimewould then be posited as right.”128 Another conception of specificdeterrence would consist in first assuming that the more severe apunishment is, the more deterrent it is, and thus in demanding theharshest punishment possible, in order to maximize public safety.This conception is contradicted by Hegel’s specific deterrence.On the contrary, the criminal should be sentenced to a milder pun-ishment; punishments must simply be precisely so severe as notto contradict the cohesion of the commonwealth or directly to endan-ger it. In this respect, Hegel’s theory of punishment is, in my view,

127. GPhR } 218, Addition. Elements, ed. Wood, p. 251.128. GPhR }} 99 and 218, Addition. Elements, ed. Wood, pp. 124, 251. A comment made

by Hegel in his manuscript explains a “cancellation . . . of the crime, which wouldotherwise be regarded as valid” as follows: “that is, would have general existence, becausehere a single Being is universal – for everyone.” (Translation mine.)


primarily a minimal specific deterrence and secondarily an equallyminimal general deterrence.

It is telling that Hegel criticizes various theories as being superficial,though not specific deterrence: “prevention, deterrent[s] . . . threat[s],corrective[s], etc.,” “specific equalit[ies],” and so on, are mentioned –the only thing absent is specific deterrence.129

It is characteristic of specific deterrence that the criminal is thesole, essential addressee of the punishment, whereas, in the varioustheories of general deterrence that operate either by threateningpunishment or by making an example of the criminal, all of societyis the addressee, including the criminal him- or herself, but only asone among many. This reflexive structure of specific deterrenceexplains why Hegel in } 101 speaks of “retribution” (Wiedervergeltung)(more properly, though, of retaliation).Wehave seen above in Section 5.3,however, that this retribution should not be confused with the theoryof retaliation, that is, with retributivism, which is sharply criticizedby Hegel.

The question arises, though, how } 100, which is considered in theretributivist interpretations to be a subjective justification of retributi-vism and which is called an “argument from law” by Seelmann, iscompatible with the interpretation that I have suggested here.130

Hegel’s } 100 is actually occupied with the subjective aspect ofcrime. This can be easily explained, however, through a specificdeterrence interpretation. The lecture Philosophy of right (1824/5)deduces the subjective perspective expressly from the cancellation ofcrime in the specific will of the criminal:

This existent will of the criminal . . . must be affected . . . This point islinked to the fact that a punishment must be impressive on the criminal.If punishment is not impressive then his existent will is not infringedupon by it.131

Therefore, the subjective aspect does not occur, for instance as a sub-jective justification without which punishment would lack a rationale,but instead the justification is deduced from the punishment’s aim ofspecific deterrence, which is grounded in the cancellation of a wrong.Coercion must be – by definition – sensitive.

129. Remarks on GPhR }} 99 and 101. Elements, ed. Wood, pp. 124, 128.130. See below, Sections 5.2.1, 5.2.2 and Seelmann, “Wechselseitige Anerkennung” and

“Versuch einer Legitimation.”131. PhR 285.


Unlike Anselm Feuerbach, Hegel understands sensitivity to be notonly that which moves, for instance, the appetites and fear, but alsothe criminal’s moral judgments.

Now, talion law is, according to Hegel, what everyone, and there-fore the criminal as well, can sense the most: “Jus talionis is a naturalconcept in perception and was seen from time immemorial as beingan element of punishment.”132 This aspect of crime that belongs to asubjective moral validity is not to be confused with either a psychologyof the appetites or Hegel’s justification of the right of punishment.The remark made by Hegel on } 99makes this clear. In it Hegel clearlyspeaks of three aspects:

1 Coercion exercised against the appetites by an “evil” or by a “sourceof evil” (this correlates to Klein’s theory of punishment that wascriticized by Hegel; in the Additions to } 99, Hegel also rejectsFeuerbach’s rationale of punishment by the “threat” of an evil).

2 The justification of punishment (“a matter of wrong and of justice”;“the objective consideration of justice, which is the primary andsubstantial point of view in relation to crime”; “[the rationale that]punishment in and for itself is just”; and “we are solely concernedwith the need to cancel . . . crime – not as a source of evil, but as aninfringement of right”).

3 The moral feeling of the criminal (“the moral point of view, i.e. thesubjective aspect of crime”).

Hegel differentiates (1) from (3) by criticizing the consequences ofKlein’s theory of punishment, which considers punishment as an evil:

it automatically follows that the essential consideration is now the moralpoint of view, i.e. the subjective aspect of crime, intermixed with trivialpsychological ideas . . . of stimuli and the strength of sensuous motives . . .as opposed to reason, of psychological coercion and of psychologicalinfluences on representational thought . . . (as if such influences werenot themselves reduced by freedom to something purely contingent).133

The aspect of punishment that belongs to subjective morality isdefined through reason and freedom. Precisely, it is both of the latterthat make up both of the foundational elements of } 100. Thisexplains why Hegel writes:

132. NRSW 277.133. GPhR } 99, Remarks. Elements, ed. Wood, p. 125.


But it is neither a question merely of an evil [which should move theappetites] nor of this or that good; on the contrary, it is definitely . . .a matter of wrong and of justice [that is, of the cancellation of the crime].As a result of these superficial points of view, however, the objectiveconsideration of justice, which is the primary and substantial point ofview in relation to crime, is set aside; it automatically follows that theessential consideration is now the moral point of view, i.e. the subjectiveaspect of crime, intermixed with trivial psychological ideas.134

Hegel does indeed not recognize the retributivist dimension as beingthe “primary and substantial point of view,” in other words as the mainjustification of punishment, but he sees in it a necessary means to theend of punishment: this dimension is the “subjective aspect of crime”and of punishment though not a subjective justification.

Even if the objective justification would require a subjective justifi-cation, this justification surely could not be found in the principle ofretaliation (Wiedervergeltung). We have already seen how Hegel attri-butes it to common sense that is entangled in contradictions: “that[the crime] deserves to be punished, and that what the criminal has doneshould also happen to him.”135 This applies not only to abstract right:morality too considers the principle of retaliation as immediate or“superficial” and contradictory.136 Thus Hegel writes in conjunctionwith the principle of retaliation in “Abstract right”:

It is incomprehensible how those sciences which derive their determina-tions from universal representations . . . should on other occasionsaccept propositions which contradict such so-called universal facts ofconsciousness.137

We find an example of that straight away in the first paragraph of“Morality,” “Purpose and responsibility,” which leads to the differenti-ation between the deed or the “external event,” and the action or the“purpose and knowledge of the circumstances.” According to Hegel,the consequences internal to the action can be distinguished fromthe consequences external to the action only “indeterminate[ly],”because the “inner necessity [consequences internal to the action]

134. GPhR } 99, Remarks. Elements, ed. Wood, p. 125.135. GPhR } 101, Remarks. Elements, ed. Wood, p. 127.136. Regarding morality: GPhR }} 105ff. Elements, ed. Wood, pp. 135ff. Cf. Hegel’s

handwritten comment on GPhR } 118: “Superficial[:] it should go well for thegood – ill for the evil.” (Translation mine.)

137. GPhR } 101, Remarks. Elements, ed. Wood, p. 127.


comes into existence in the finite realm as external necessity,”138 thatis, inner necessity appears as consequences external to the action.Hegel writes:

The maxim . . . which enjoins us to disregard the consequences of ouractions, and the other which enjoins us to judge actions by theirconsequences and make the latter the yardstick of what is right andgood, are in equal measure [products of the] abstract understanding.139

His conclusion is:

It follows from this that the criminal stands to benefit if his action hasless adverse consequences, just as the good action must accept that itmay have no consequences or relatively few; and it also follows that,once the consequences of the crime have developed more fully, thecrime itself is made responsible for them.140

If one separates deed and action in an abstract way then this leads oneto break down (Zersplitterung) “the consequences [for penal law]minutely.”141 With this, though, the possibility of retaliation for thecriminal action disappears in the judgment of common sense, andonly the deed can be retaliated for: “Guilt or innocence – in relation toevils – and of evils to guilt. – Suffering in general, – punishment amongstit as well, – on the whole as a consequence of the action – or also not.”142

If the principle of retaliation only relates to the deed then it cannotperceive the subjective aspect and abstracts the criminal will, withoutwhich there is no crime, but instead only an evil.

The principle of retaliation cannot be regarded, therefore, as beingjustification of punishment, but – as in Fichte143 – is rather justifiedonly as a means to the end of punishment. As an external, superficialprinciple, the principle of retaliation is especially suited, though, to beunderstandable to the criminal: retaliation places an external, imme-diate identity between the crime and the punishment; this identity isimmediate because it does not operate through the concepts of rightand of law as well as through their necessary validity. Instead, thisimmediate identity only relates to the material consequences of thecriminal’s deed for the victim and to the material consequences for

138. GPhR } 118, Remarks.139. GPhR } 118, Remarks.140. GPhR } 118, Remarks.141. GPhR } 118, Remarks. Elements, ed. Wood, pp. 145–6.142. Handwritten comments on } GPhR 118. (Translation mine.)143. GNR i/4 61. Fichte, Foundations of natural right, p. 228. Cf. above, Section 4.3.


the criminal – in short, to the evil. Nonetheless, the principle ofretaliation suggests to the criminal a regular, non-physical identitybetween his or her deed and punishment. Thereby, a first stage ofthat conscious process is reached that will lead to rehabilitation.

Hegel’s rationale of punishment cannot be reduced to a meretheory of deterrence, because he justifies penal law primarily neitherwith specific deterrence nor with general deterrence nor with rehabili-tation nor with the principle of retaliation, but instead with thenecessary self-destruction of the coercion or from its self-cancellation.However, if one thinks in the usual categories, one must still observethat Hegel regards specific deterrence together with rehabilitation asthe primary aim and as the primary justification of punishment,whereas general deterrence represents a secondary aim, though anonetheless important one. Retributivism, on the contrary, deliversneither an aim nor a justification of punishment.

In conclusion of this investigation of the Hegelian justification ofpunishment, I must observe, as in my investigation of both the Kantianand the Fichtean justifications, that if one proceeds not from thedeparture point of the criminal’s intention or deed, but instead fromthe departure point of the concept of right, which all three authorshave in common, retributivism cannot then be obtained. Since theconcept of right in all three authors (and correctly so) stands on ahigher level than their penal law, any exponent of this concept ofright will have to renounce retributivism.


PART III

RETRIBUTIVIST INHUMANITY

The theory of retaliation is not only incompatible with the concept ofright of Kantian philosophy (see Part i) and of German idealism (seePart ii). Contrary to Kant’s claims, it even proves itself to be a theorythat respects the dignity in the criminal’s person less than the alterna-tives of deterrence. Most radically, Nietzsche excludes the concernfor human dignity both in the citizen’s and in the criminal’s personfrom being a motive and a plausible aim of retributive punishment.According to Nietzsche, the idea of retaliation allows no room for thenecessary internalization and reflection out of which alone bad con-science and remorse in the criminal as well as humane interactionwith society could arise (see Chapter 6).

In a deliberate move away from retributivism, I have made a plea inPart ii from the standpoint of a combination of specific deterrenceand rehabilitation for a type of interaction that respects human dig-nity even with the most hardened and inhumane of criminals, that is,with those who have perpetrated crimes against humanity. ClausRoxin believes he raises a decisive objection against the “conceptionof specific deterrence” by accusing it of not wishing to accept its finalconsequences:

The decisive example is at the moment the concentration campmurderer by whom some innumerable number of innocent humanbeings were killed due to sadistic motives. These murderers are livingtoday mostly inconspicuously and socially integrated, so they are not inneed of “rehabilitation.” Also, the danger of recidivism from which theymust be deterred and from which we must be secured, does not existwith them. Should they really remain unpunished because of that?1

1. Claus Roxin, Strafrechtliche Grundlagenprobleme (Berlin: De Gruyter, 1973), p. 7.

147

Roxin is implicitly operating under the assumption that thereis nobody who would answer “yes” to this question. But preciselythis answer appears to me to be the right answer, as I will attemptto show by way of the punishment of crimes against humanity(see Chapter 7).

148 retributivist inhumanity

6

NIETZSCHE AND PUNISHMENTWITHOUT REMORSE

Kant does not consider that the aim of punishment is to arouse badconscience, and thus the criminal’s remorse as well. If punishmentwere to be imposed in the service of such an aim, it would not occur –as Kant requires – merely because the criminal broke the law, butinstead in order that a certain situation might come about. Since thearousing of bad conscience or remorse seems to be a minimalistic aim,which has moreover a clear reference to the criminal’s dignity, someinterpreters of Kant develop a conception of retributive punishmentdirected toward this aim.1 If one proceeds under the assumption thatretributive punishment especially respects the criminal’s human dig-nity and that it can generate – as Hegel points out (see Section 5.5) –even in the simplest mind a connection between crime and punish-ment, then one can expect the criminal’s remorse from retributivepunishment.

It is Nietzsche who best shows us that such a justification is notself-evident but instead very problematic. Nietzsche’s “polemic”writing On the genealogy of morality (1887) opposes such a moralizingconception of retaliation and of punishment. From Nietzsche’s per-spective, retaliation does not stem from concern for human dignityin the criminal’s person but instead in the victim’s and other humanbeings’ active Schadenfreude. Also, the expiation of punishment cannotpossibly be an inducement to moral conversion, but ratherstrengthens the criminal’s felonious will. It is punishment that isdirected toward deterrence which proves to be the more humaneoption of penal law.

1. Cf. chapters 9 and 11 of Thomas E. Hill, Human welfare and moral worth: Kantianperspectives (Oxford: Clarendon Press, 2002).

149

In Nietzsche’s On the genealogy of morality, Essays ii 8–15 belong tothe context of the second tractate whose aim is an inquiry into thegenealogy of the feeling of guilt. Essay ii 8 begins with taking stockof the paragraphs regarding the genesis of promises in GdM ii 1–3as well as about the history of the genesis of Schuld (that is, guiltor debt)2 and its acquitment in the form of making the insolventdebtor suffer:

The feeling of guilt, of personal obligation, to pursue our train ofinquiry again, originated, as we saw, in the oldest and most primitivepersonal relationship there is, in the relationship of buyer and seller,creditor and debtor: here person met property for the first time, andmeasured himself person against person.3

The further course of the investigation throughout these aphorismsconsists of three steps. First, Nietzsche gives a description of thegenesis of public punishment according to the “canon of moralsrelating to justice.”4 Secondly, he then delineates the genesis of publicpunishment in his rejection of Karl Eugen Duhring’s conception ofjustice and punishment as either ressentiment or revenge.5 Thirdly andfinally, Nietzsche takes it a step further: he radically separates publicpunishment, whose genesis he has hitherto explained without refer-ence to the criminal’s feeling of guilt, from this alleged feelingof guilt.6 In the following I will mainly devote myself to the first part,that is, aphorisms 8–10, because these sections depict the core ofNietzsche’s conception of punishment.

6.1. The genealogy of public punishment

Out of the oldest relationship between human beings (accordingto Nietzsche), that is, out of the relationship between buyer andseller, arises “the habit of comparing power with power, of measuring,of calculating.”7 According to a physiological process, which was

2. In the German language, the word guilt (Schuld) is sometimes ambiguous. It can meaneither a financial debt or moral guilt. Nietzsche is, of course, keenly aware of itspolysemous nature.

3. GdM ii 8. Friedrich Nietzsche, On the genealogy of morality, ed. Keith Ansell-Pearson,trans. Carol Diethe (Cambridge: Cambridge University Press, 2007), p. 49.

4. GdM ii 8–10, cited here is section 8. Genealogy of morality, ed. Ansell-Pearson, p. 50.5. GdM ii 11–13. Genealogy of morality, ed. Ansell-Pearson, pp. 52–8.6. GdM ii 14–15. Genealogy of morality, ed. Ansell-Pearson, pp. 58–60.7. GdM ii 8. Genealogy of morality, ed. Ansell-Pearson, p. 49.


explained by Nietzsche in The philosophers’ book,8 the habit becomes agenerality. In this way, the “universalization” arises, which constitutesthe “first level” of justice or its “oldest, most naıve canon of morals”:“Every thing has its price: everything can be compensated for.”9

One reason why this canon of morals is “naıve” is because it is still notinternalized, “intellectualized” (vergeistigt) – to make use of Nietzsche’sown vocabulary. In this passage, which is based on the aforementionedprinciple of justice, guilt and punishment, moreover, are still investi-gated not in their internalized dimension, but instead only in theexternal dimension of a harm and of punishing the one who caused it.Not only the feeling of guilt, but also the internal guilt remains uncon-sidered in this explanation of the establishment of public punishment.

Because this canon of morals is still not internalized, it is also naıvein the sense that it openly compares “power with power” and definesitself by power relations:

Justice at the first level is the good will, between those who are roughlyequal, to come to terms with each other, to “come to an understanding”again by means of a settlement – and, in connection with those who areless powerful, to force them to reach a settlement amongst themselves.10

In short: justice should be comprehended merely in a positivisticsense. It is not the just exchange that is defined through the equiva-lence of exchanged goods, but instead the equivalence that is definedthrough the actual exchange of goods, which is decided, in turn,through apower relation.11Nietzsche, regarding this point, is followingHobbes, who raises the following objection against those authors whoregard fair exchanges as an arithmetical ratio:

As if it were Injustice to sell dearer than we buy; or to give more to a manthan he merits. The value of all things contracted for, is measured bythe Appetite of the Contractors: and therefore the just value, is thatwhich they be contented to give.12

The conclusion from this, we will have to remember later on, is thatimpartiality is impossible.

8. WuL i, 880. Daniel Breazeale, Philosophy and truth: selections from Nietzsche’s Notebooks ofthe early 1870s (Atlantic Highlands, NJ: Humanities Press, 1979), pp. 79–97.

9. GdM ii 8. Genealogy of morality, ed. Ansell-Pearson, p. 50.10. GdM ii 8. Genealogy of morality, ed. Ansell-Pearson, p. 50.11. Cf. HenningOttmann,Philosophie undPolitik beiNietzsche (Berlin:DeGruyter,1987), p.131.12. Part 1 , chapter 15 of Thomas Hobbes, Leviathan, ed. A. R. Walter (Cambridge:

Cambridge University Press, 1904), p. 102.

nietzsche and punishment without remorse 151

Just as in Hobbes’s Leviathan justice is defined by the Leviathanalone, with Nietzsche the commonwealth is limited to the tasks ofprotection, of peace and of ensuring social trust. According to this, thecriminal is an attacker (he “also actually assaults the creditor”), a “manwithout peace” and “somebody who has broken his contract and hisword” – therefore, someone who is not allowed to promise anymore.13

Nietzsche’s explanation of the genesis of punishment in Essay ii9–10 of his Genealogy of morality shows us prima facie public punish-ment’s four periods:

1 punishment as a sort of war14

2 punishment as a “compromise with . . . anger” at those times when a“community grows in power”

3 the mitigation of punishment (“As the power and self-confidence ofa community grows, its penal law becomes more lenient”)

4 impunity in times that are “not impossible to imagine” whensociety’s “conscious[ness] of its power” is especially developed.15

When we examine these four periods more closely then it becomesapparent not only that a public punishment exists solely in the secondand third periods, but also that a legal and civil order can only exist inthose two periods. In the following, I will attempt to show why this is soand to what function this fact gives rise for public punishment.

Essay ii 9, in a first period, takes the logical consequence out of thedetermination that the criminal is an attacker, a person without peaceand somebody who has broken his or her word and contract. Pun-ished the least is the “immediate danger”; instead it is, “quite apartfrom this,” the word-breaking that is punished, the logical result ofwhich ought to be that the criminal should not be allowed to promise.The criminal is namely more than merely an insolvent debtor: he orshe attacks the creditor, thus not recognizing what is owed. Therefore,the question poses itself whether the creditor can reach compensationin spite of this with him or her, as is the case with the simply insolventdebtor, upon whom the creditor can inflict suffering because thisgives him or her pleasure. I will return again to this question.

Nietzsche prefers, however, prima facie another manner of think-ing. According to Nietzsche, every member of the community is adebtor who still has to acquit the debt: he or she must contribute to

13. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 50.14. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, pp. 50–1.15. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.


domestic peace, to mutual protection and trust. Because the criminalcannot do this, he or she cannot be a member of the community anylonger. The criminal is “elend,” that is, expelled, exiled and outlawed.The same provisory conclusion was already reached, for example – asI mentioned previously (see Section 4.3) – by Fichte: the criminalwould be expelled into the “desert” and everyone would be allowed toshoot him like a dangerous wild animal. The death of the criminal,and everything resulting from that exile, is no longer the concern ofthe commonwealth.

At the same time, there are also elements of another logic to befound in the same section:

1 elements of compensation: “The community, the cheated creditor,will make you pay up as best it can”16

2 elements of mnemotechnic: the criminal “from now on, as is fair . . . isnot only deprived of all these valued benefits, – he is now alsoreminded how important these benefits are”17

3 elements of infliction of suffering, which in Essay ii 6 counted ascompensation for the unpaid debts.

Nietzsche summarizes:

The anger of the injured creditor, the community, makes him returnto the savage and outlawed state . . . “Punishment” at this level ofcivilization is simply a copy, a mimus, of normal behaviour towards ahated, disarmed enemy who has been defeated . . . in fact, the rulesof war and the victory celebration of vae victis! in all their mercilessnessand cruelty.18

The suffering is indeed compensation for the simply insolvent debtor.If suffering is compensation for the criminal as well, then insolventdebtors and criminals are treated equally, even though the criminal’sdebts fundamentally are qualitatively greater. This would point to arelative powerlessness of the commonwealth and the polity in respectto the criminal as well as to the fact that the compensation and theinfliction of suffering constitute not the punishment, but instead – if atall – the mnemotechnic.19 The contrast between the “anger of . . . thecommunity” and its reaction is obvious: first, the community “cast[s]”

16. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 50. Also, cf. GdM ii 5f (pp. 43–6).17. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 50. Also, cf. GdM ii 3 (pp. 41–2).18. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, pp. 50–1.19. Cf. especially GdM ii 3. Genealogy of morality, ed. Ansell-Pearson, pp. 41–2.


the criminal “away” from itself; only thereafter can “any kind ofhostile act . . . be perpetrated on him.”20

It is still unclear whose anger is directed with cruelty toward thecriminal. Nietzsche merely writes: “The anger of the injured creditor,the community, makes him return to the savage and outlawed state”(the logic of expulsion) and “now any kind of hostile act can beperpetrated on him” (the logic of infliction of suffering). The latterformulation is vague, just like the following and last lines of Essay ii 9.Does it concern two logical moments of public punishment or two actorsof punishment?

Either, the commonwealth initially expels the criminal (the firstmoment) and then handles him or her with cruelty like a defenselessenemy, whereupon one can imagine “all kinds of dishonour andtorture” (second moment). Or, the state (the first actor) expels the crim-inal and the individual citizens (the second actor) perpetrate all mannerof cruelties on the criminal – one might name this as a modernequivalent of lynch mob “justice.”

The more plausible answer certainly lies in the middle of these. Thecommonwealth perpetrates these cruelties because the criminal hasbeen excluded and because the individual citizens are angry anddemand cruelties.

The difference between the commonwealth as such and the sum ofits members reaches its clear expression at first in the second period ofpunishment. There Nietzsche writes: “the wrongdoer is no longer‘deprived of peace’ and cast out, nor can the general public vent theiranger on him with the same lack of constraint.”21 The decisionwhether the criminal should be expelled from the community orcan remain is decided by the anger of the community as such – thatis, by the ruler and not by the general anger (allgemeine Zorn) of themasses. In the first period, the ruler decides for the expulsion.Whether the ruler carries out the cruelties in person or simply allowsthem, whereupon they would be then carried out by the crowd, isactually not the core issue: in both cases the general anger dictates thepunishment – not the ruler and his or her anger. The general anger isreactive, though: it is a ressentiment that results from powerlessness.22

In this case, powerlessness results assumedly from the impossibility of

20. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 50.21. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.22. Cf. GdM ii 5. Genealogy of morality, ed. Ansell-Pearson, pp. 43ff.


reversing the crime’s effects. Anger is something different from areadiness to render compensation. Even if Nietzsche thoroughlyregards the pleasure provided by the infliction of suffering as a com-pensation,23 it is possible that the angered individuals remain unsatis-fied with the criminal’s suffering. For the victim, the pleasure from thesuffering is only a surrogate.

The fact that there is “war” between the crowd and the criminalpoints in this direction. Since this war does not occur between statesand since the state cannot wage war with individual persons, it there-fore can only be a war between the population and the criminal. A warcan, however, only occur between two communities; for that reasonhere it concerns itself with only a “copy” and with even the “mimus” ofa bellicose attitude:24 the mimus was only a farce for the Romans. Infact, in the second period, Nietzsche only talks about the avoidance of“further or more wide-spread participation [in anger] and unrest.”25

Because there is an implicit contrast to the first period in the contextof this second period, one may assert that the first period is typified bya diffusion of unrest.

The first period was, therefore, more a period of private justice. Inthe second period, punishment is no longer determined by the crowd,but instead by the commonwealth and polity as such. Only then are wedealing with a public punishment in the proper sense; in contrast tothat, in conjunction with the first period Nietzsche writes “punish-ment” only within quotation marks.

Nietzsche begins Essay ii 10 with the following words:

As a community grows in power, it ceases to take the offences of theindividual quite so seriously, because these do not seem to be asdangerous and destabilizing for the survival of the whole as they didearlier.26

For this reason – Nietzsche is here in agreement with Hegel – punish-ments can be milder. Nowadays one would essentially interpret thedanger and the risk of collapse as Hegel does in } 218 of his Elements ofthe philosophy of right: the more stable the state, the less the risk that acrime incites further crimes and, hence, makes a stricter generaldeterrence necessary. Unlike with Hegel, however, with Nietzsche this

23. GdM ii 5ff. Genealogy of morality, ed. Ansell-Pearson, pp. 43–9.24. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, pp. 50–1.25. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.26. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.


is not what the concern is; rather, the concern is about the risk that acriminal might agitate the entire society or might bring about civilwar. In which case is this a lesser risk? Perhaps it is less if the quantityand the gravity of the crimes are held as being so low or the citizens’safety and trust are so high that the crimes are not perceived as beinga direct threat by the majority of the population. This is even moreprobable, though, if the state possesses, in a stable manner, the realmonopoly on violence, which makes it possible for the state success-fully to resist the anger of its citizens. This may be the definition ofa state’s “grow[ing] in power.”

The state puts the criminal under its protection against the generalanger, in order that the crime be totally acquittable and then trulyacquitted. Unlike an enemy, with whom one remains unreconciledeven after his or her defeat and death, public punishment shouldprovide for either the complete expunction or the complete acquitt-ability of the debt contracted by the crime. In this regard, threeaspects matter.

First, the state must see to it that “the wrongdoer is isolated from hisdeed.”27 The consequence resulting from this is that, unlike anenemy, the criminal is not to be destroyed. In this manner, the criminalcannot be reduced to his or her crime, and can later become again amember of the commonwealth. Yet one should not forget the fact thatNietzsche would in no way endorse this development. For he writes inEssay i 13 that “there is no ‘being’ behind the deed . . . ‘the doer’ isinvented as an afterthought, – the doing is everything.” Were one tocite, as Michel Foucault did, a phrase from the French philosopher andpoliticianGabriel deMably as being the criterion formodern penal law,in contrast to the pre-modern penal law – that says “punishment . . .should strike the soul rather than the body” – then Nietzsche moredistinctly stands on the side of pre-modern penal law.28

Secondly, an equivalent has to be found for a deed isolated in thisway, or, more exactly, multiple “equivalents” have to be found29 – eachone probably determined by the gravity of the deed. The equivalent,therefore, can only be an illusion; (public) punishment can only be

27. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.28. Michel Foucault, Discipline and punish: the birth of the prison, trans. Alan Sheridan, second

edition (New York: RandomHouse, 1995), p. 16. The original quotation is taken fromGabriel deMably, “De la legislation,” in œuvres completes (12 vols., London, 1789), vol. 9,p. 326, and reads: “Que le chatiment . . . frappe l’ame plutot que le corps.”

29. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, pp. 51–2.


based on an illusion. Punishment is not an equivalent, but only thewaiving of an equivalent.

Thirdly, this should result in “a compromise with the anger ofthose immediately affected by the wrong-doing.” Nietzsche does notfurther specify the contents of this compromise. One may guess,however, that the compromise consists in inflicting suffering on thecriminal and declaring him or her as being inferior to the victim.Indeed, in most of the penal theories, punishment is defined as theinfliction of an evil (Ubel),30 not merely as some kind of answer tothe crime (a definition of punishment without a constitutive inflictionof suffering is nevertheless possible; rehabilitation, for example, reallygoes beyond compensation, but is not thought of as being an evil).Already in Essay ii 5, Nietzsche writes that, in the case in which “theactual power of punishment, of exacting punishment, is . . . trans-ferred to the ‘authorities,’” the victim attains “the elevated feeling ofdespising and maltreating someone as an ‘inferior.’”31 One shouldnot overlook the point that the state takes no pleasure in inflictingsuffering on the criminal or in his or her humiliation. The compro-mise can, however, partly consist in the state guaranteeing relative orincreased security.

The principle of acquittability (Prinzip der Abzahlbarkeit) still doesnot say anything about the degree of punishment. The criminal canrender compensation by a mild punishment, as well as by death aftermanifold tortures for longer durations. The principle of acquittabilityonly requires that every crime can be expiated or that any debt can beacquitted regardless of how serious it may be. Ultimately, therefore,reconciliation must occur along with the death of the criminal.

Nietzsche addresses the question of degree in punishment for thefirst time in the premises of the third period of punishment:

As the power and self-confidence of the community grows, its penallaw becomes more lenient; if the former is weakened or endangered,harsher forms of the latter will re-emerge. The “creditor” alwaysbecomes more humane as his wealth increases; finally, the amount ofhis wealth determines how much injury he can sustain without sufferingfrom it.32

30. Cf., for instance, in RL } 49e, Ak vi:331ff. Practical philosophy, ed. Gregor, pp. 472–5.Also, see above, Chapters 1– 3.

31. GdM ii 5. Genealogy of morality, ed. Ansell-Pearson, p. 45.32. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, pp. 51–2.


Nietzsche seems at this point to introduce, astonishingly, bothanother principle of penal law and another psychological basis of it.The principle of penal law no longer appears to be a compensation inproportion with the power relation, but deterrence instead. Likewise,the psychological motivation of the ruler no longer appears to bethe will to power; instead, it appears merely to be security, that is,the merely reactive will to self-preservation.

This third period comes to a climax in a fourth period:

It is not impossible to imagine society so conscious of its power that it couldallow itself the noblest luxury available to it, – that of letting itsmalefactors go unpunished. “What do I care about my parasites”, itcould say, “let them live and flourish: I am strong enough for all that!”33

At this stage even deterrence is dispensed with: the “parasites” areallowed to weaken the community. The corresponding psychologicalpremise is no longer the will to power, nor is it even the search for self-preservation and security, but instead it is only an indifference to theworld or the dissipation of luxury.

The alternation in the psychological premises conforms toNietzsche’s conception of the biological process of deployment andexhaustion of energies. In Nietzsche’s thinking, the active energies donot remain constant, but instead attempt primarily to assert them-selves: they strive for power. However, at one point they will exhaustthemselves and then they will only tend toward merely reactive self-preservation or even toward disappearance – toward the nihil of nihil-ism. The concluding judgment of Essay ii 11 can be applied to this laststage of biological development:

A system of law conceived as sovereign and general, not as a means foruse in the fight between units of power but as means against fighting ingeneral . . . this would be a principle hostile to life, an attempt toassassinate the future of man, a sign of fatigue and a secret path tonothingness.34

Not only in the contemporary debate over penal law does impunityhardly count as a serious option.35 Also, those authors who arenot supporters of maximizing general deterrence being an aim of

33. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.34. GdM ii 11. Genealogy of morality, ed. Ansell-Pearson, p. 54.35. For an exception, see Hermann Bianchi, “Abolition: assensus and sanctuary,” in

Alexander R. Duff and David Garland (eds.), A reader on punishment (Oxford:Oxford University Press, 1994), pp. 336–51.


punishment share the view that total impunity would be a weightyincentive for crime; so that at the very least, owing to this reason ofminimal general deterrence, no commonwealth can dispense with theexistence of punishment without challenging the existence of thecommunity itself. The true power of a state can never be so great thatit can dispense with every kind of punishment. In situations in which astate enjoys a growing power but does not wish completely to abdicateits security, the degree of punishment may decrease; at most, impunityremains the ever unreachable horizon of this mitigation: impunitypoints to a utopia. Nietzsche writes, in fact, not that “it is not impos-sible to imagine society so increased in power,” but instead that “it is notimpossible to imagine society so conscious of its power” (emphasismine). This power consciousness relates to no realistic perception;rather, it should be understood as being a symptom of a biological orpsychological condition – a condition of exhaustion.

No acquittal of debt takes place in the period of impunity; in thisrespect, impunity is outside of the sphere of justice defined by theprinciple of universal acquittability and comparability. The lack ofcomparison and equivalence fulfills the definition of “luxury.”A luxury is something to the price of which one pays no heed. Inthe period in which the criminal was fought against like an enemy,however, there likewise existed no justice because there was at thattime a state of war and the criminal was completely delivered to theanger of the people: there was no system of acquittability in general.

According to Nietzsche, the deeper and the lower “the position ofthe creditor in the social scale,” the more the creditor feels a lust toinflict suffering as payback, because “through punishment of thedebtor, the creditor takes part in the rights of the masters.”36 The victimsof the criminal as well as the other simple citizens sharing the generalanger enjoy this “right of the masters.” The rulers who hold themselvesto be especially powerful and decide in favor of absolute impunity – thatis, universalized pardon – also enjoy “this prerogative of the mostpowerful man”; however, he or she does this only on account of anillusion of power. Both the simple, angry citizen and the indifferent, all-pardoning ruler, on the one hand, are powerless, but, on the otherhand, they succumb to the illusion of power.37 In contrast, true powerconsists in “forc[ing]” “those who are less powerful [in this case, the

36. GdM ii 5. Genealogy of morality, ed. Ansell-Pearson, pp. 44–5.37. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, pp. 51–2.


‘less powerful’ are both the criminal and his or her victim] . . . toreach a settlement amongst themselves.”38 In fact, only in thesecond and third periods can a compromise be reached betweenthe interest of the criminal (impunity) and the interest of the victim(a bellicose, informal cruelty).

6.2. Nietzsche’s appraisal of justice

The conclusion – nearly a lesson! – that Nietzsche delivers at the endof his explanation of punishment seems at first glance to raise a directobjection to my version. He writes:

Justice, which began by saying “Everything can be paid off, everythingmust be paid off”, ends by turning a blind eye and letting off thoseunable to pay, – it ends, like every good thing on earth, by [cancelingitself]. The [self-cancellation] of justice: we know what a nice name itgives itself – mercy.39

With these lines of text, Nietzsche seems to see impunity as being thelogical conclusion of justice or acquittability, and therefore not asbeing a deviation from strict justice. For Nietzsche speaks expresslynot of sublimation, but of the “[self-cancellation] of justice” (emphasismine). One would rather expect that mercy would be designated asbeing an extrinsic cancellation of justice, being beyond justice, andthus as it is according to Nietzsche himself, “being beyond the law( Jenseits des Rechts).”40

Nietzsche’s charge against justice is that it seeks acquittal of debtwhile nevertheless allowing those unable to pay to escape punishment.There are at least two ways to understand this. Either, one sees theturning of a blind eye to those who are insolvent pertaining merely tothe fourth period, that is, to absolute impunity. But then the charge ofself-cancellation is incomprehensible. Or, one understands it in such away that the turning of a blind eye indeed completely unfolds in thefourth period, though it was already – therefore, from the outset –present in the second and third periods. In this respect, even public

38. GdM ii 8. Genealogy of morality, ed. Ansell-Pearson, p. 50.39. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, pp. 51–2. Translator’s note: in the

translation, Aufhebung is rendered as “sublimation,” but in the Nietzschean sense, as inthe Hegelian context, rendering it as “suppression,” “cancellation” or “sublation”might prevent any confusion.

40. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 52.


punishment as acquittal of debt means the turning of a blind eye; inshort, it means not truly acquitting the debt. In this context, oneought to recall two points:

1 The criminal is a promise breaker, not only a malefactor, which iswhy he or she loses “all rights and safeguards” and, furthermore, “allmercy as well.”41

2 Nietzsche stresses that “you live in a community, you enjoy thebenefits of a community (oh, what benefits! sometimes we under-estimate them today).”42

One can gather from both of these points, first, that the criminal’sdebt is qualitatively especially large and, secondly, that the creditor’sassets are likewise qualitatively especially large. Now, Nietzsche neverhimself adopts the principle of justice that “everything can be compen-sated for [acquitted].”43 On the contrary, nobler things cannot beacquitted; and the institution of the state and its establishmentwholly rank among these nobler things, as is demonstrated, forexample, by Nietzsche’s interjection “oh, what benefits!” We are,therefore, allowed to proceed under the assumption that forNietzsche the acquittability is not (at least not always) at hand. Inthe second period, the acquittal, which is then public punishment,thus means that the debt remains unpaid – at least partly: the debtorgets off relatively easily.

The following explanation of this paradox seems obvious: thejustice that leads to this renouncement of a real equivalent stemsfrom the ressentiment of the victim. Nietzsche constructs a contrastingalternative theory to this theory held by Duhring. The latter regardsthe connection of crime with ressentiment and with revenge in thefollowing manner:

In our moral deliberations, we have already regarded each action whichoriginally, inimically hurts as the object of a necessary reaction.This reaction expresses itself first internally, in a counter-sensation(Ruckempfindung) that we can also name “ressentiment” and “need forretribution” or, precisely, “revenge.”44

41. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 51.42. GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 50.43. GdM ii 8. Genealogy of morality, ed. Ansell-Pearson, p. 50.44. Eugen Duhring, Cursus der Philosophie als streng wissenschaftlicher Weltanschauung und

Lebensgestaltung (Leipzig: O. R. Reisland, 1894–5), p. 224.


Now, Duhring understands justice to mean retribution (Vergeltung).Hence justice rests on the need for revenge. Why, however, oughtthis need to remain a ressentiment or a counter-sensation instead ofleading to the action of revenge? Duhring explains the rejection of the“vendetta” by stating that it “yields a private war that is instigated andcontinued over and over again.”45 In order to avoid a civil war, oneestablishes a “system of conflict resolution in virtue of which one seeksappeasement of revenge by means of atonement and compensa-tion.”46 In return for forgoing revenge, the victim receives, in additionto public peace, the guarantee that evils will be imposed upon thecriminal and that he or she will feel guilt and remorse:

But the high rates with which one allows one’s own bodily injury and thekilling of dependants to be bought off after the fact should not lead topassing over the fact that the readiness to make an earnest materialsacrifice could also vouch for the dignity of the changed will and, hence,for true remorse and for a peaceful disposition. The need for revengedoes not disappear, however, only by subduing and harming the injurer,but it is also balanced out if the malefactor himself assumes hiscastigation honestly by humbling himself by the true confession ofguilt and by the accomplishment of an apology and a punishment.47

Nietzsche refutes this explanation in two steps. First, he shows thatjustice and public punishment have different origins. Secondly,according to Nietzsche, the punishment does not give rise to eitherany feeling of guilt or any bad conscience in the criminal, but ratherthe opposite.

Immediately, at this point, a new paradox arises. Should the pun-ishment in accord with justice originate not from “emotional reactions”but instead from “active emotions,”48 then how can the renunciationof a real equivalent be explained on the basis of these “active emo-tions”? The answer lies in the aforementioned distinction betweentwo actors, that is, between the population and those who wield power.In general, Nietzsche always responds to inquiries into the natureor into the aim of an institution with a counter-inquiry into thepower relation out of which the institution arises. The “attempt” ofthe person in power “to localize the matter and head off further or

45. Duhring, Cursus der Philosophie, p. 225.46. Duhring, Cursus der Philosophie, p. 225.47. Duhring, Cursus der Philosophie, pp. 225–6.48. GdM ii 11. Genealogy of morality, ed. Ansell-Pearson, p. 52.


more wide-spread participation and unrest” is implemented by thepunishment.49 This striving for civil peace, which Duhring also saw asthe ground for public punishment as compromise and renunciationof retribution, is not a final aim in Nietzsche’s conception – unlike inDuhring’s conception. In Nietzsche’s conception, the rule of lawand civil peace were not established for the sake of the safety of theindividual citizens, but instead as a means for ensuring power forthose in power. Nietzsche emphasizes this distinction:

A system of law conceived as sovereign and general, not as a means foruse in the fight between units of power but as a means against fighting ingeneral . . . [requiring] that every will should regard every other will asits equal, this would be a principle hostile to life . . . a sign of fatigue and asecret path to nothingness.50

Therefore, the state’s efforts at establishing a civil peace – amongother things, by means of public punishment – should be regarded asa weapon. The prevalence of power can be internal as well as external.The internal prevalence of power consists in the situation in whichindividual human beings – for instance, the victims – no longer inflictthe “punishment,” but instead the state alone; thereby, the stateenforces what has generally been later termed the “monopoly onviolence.” The external prevalence of power consists in the fact thatcivil peace represents a considerable advantage in struggles againstother states. Nietzsche’s praise of war as a principle that is favorable tolife finds repeated expression in his works.51

The result of the evolution from private revenge to public punish-ment first appears negative, though, for the population in general andfor the victim especially. This is because “[gruesome] punishment . . .has such strong festive aspects!”; and “to see somebody suffer is nice,to make somebody suffer even nicer – that is a hard proposition, butan ancient, powerful, human-all-too-human proposition.”52 At thispoint, the people lose the ability “to make someone suffer” (Leiden-machen) and receive as consolation only the ability “to see somebodysuffer” (Leiden-sehn) in which the suffering constantly becomes milderandmore discrete themore powerful the state becomes. The population

49. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.50. GdM ii 11. Genealogy of morality, ed. Ansell-Pearson, p. 54.51. For instance, in Human, all too human (i } 224); Thus spoke Zarathustra (iv 312); and

Twilight of the idols (} 38).52. GdM ii 6. Genealogy of morality, ed. Ansell-Pearson, p. 46.


experiences a clear frustration because “the wrong-doer is [instead]carefully shielded by the community from this anger, especiallyfrom that of the immediate injured party, and given protection.”53

Previously, private punishment meant “tak[ing] part in the rights ofthe masters,”54 which disappears at this point. The fiction of equiva-lency or of justice is introduced as compensation for this frustration,as a “compromise” or as a “restriction . . . of the true will of life”;however, first and foremost, it is the fiction of the bad conscience or ofthe remorse of the criminal that provides consolation to ressentiment.This remorse, which Duhring holds both to be something real andto be the aim of punishment, proves to be a mere means to theprevalence of the (state’s) power for Nietzsche.

Nietzsche stresses that change of perspective that plays out betweenthe philosophy of right and the philosophy of penal law, on the onehand, and his own conception, on the other hand. According toNietzsche, punishment has no “purpose” (Zweck); instead, the conceptof punishment represents “a whole synthesis of ‘meanings.’”55 Punish-ment refers neither to the purpose of the population (that is losing itsability “to make someone suffer”) nor to the purpose of power (that islimiting itself), nor does it effect any remorse in the criminal, as weshall later see. Not long before Nietzsche’s On the genealogy of morality(1887) appeared, two of the most significant penal law theoristsplaced the purpose of punishment at the center of penal law: Rudolfvon Jhering and Franz von Liszt.56 All of the great outlines of theory inthe tradition, from Plato’s Gorgias to Hegel’s Elements of the philosophy ofright, through Beccaria’s On crimes and punishments and Bentham’sTheory of legislation, attempted to justify the institution of punishmentwith an end. In this respect, Nietzsche represents a radical break in thetheory of penal law; only a narrow bridge remains between Nietzscheand the tradition: both approve of the institution of punishment.Nietzsche’s conception, however, rejects every dimension of justiceand any justification of punishment as being mere illusions.

Admittedly one may tend toward the illusion that Nietzsche praisesjustice when he writes: “If it actually happens that the just man

53. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.54. GdM ii 5. Genealogy of morality, ed. Ansell-Pearson, p. 45.55. GdM ii 13. Genealogy of morality, ed. Ansell-Pearson, p. 57.56. Rudolf von Jhering, Law as a means to an end, ed. Joseph H. Drake, trans. Isaac Husik

(Union, NJ: Lawbrook Exchange, 1999). Franz von Liszt, Der Zweckgedanke im Strafrecht,(Berlin, 1883).


remains just even towards someone who has wronged him . . . well,that is a piece of perfection, the highest form of mastery to be had onearth.”57 In so doing, however, three basic aspects should not beoverlooked. First, Nietzsche is referring in no case to justice, butinstead to “the just man.” Secondly, the sentence is formulated hypo-thetically (“If it actually happens . . .”), and its condition is not fulfilledsince it is “something which we would be wise not to expect andshould certainly find difficult to believe.”58 Furthermore, Nietzschenever takes an idealistic position such as that: Nietzsche categoricallydisapproves of any idea whatsoever. Thirdly, the just character of ahuman being who stands the closest to justice does not stem from hisor her appreciation for justice. The human who stands close to justiceis only “a hundred paces nearer to justice than the man who reacts.”59

Only in so far as he or she is “the active, aggressive, overreaching man[or woman]” is this person close to justice. Therefore, he or she isprimarily powerful and aggressive, and only secondarily just.

We have already given the explanation for this. For Nietzsche,justice is a means to power. A justice that is an autonomous end orthat is a value simply does not exist in Nietzsche’s thought.

6.3. The feeling of guilt arises not in the convict –but instead in the normal citizen

We have seen that Nietzsche holds remorse to be a mere fiction.According to Nietzsche, either the convicts become later recidivists andonly more skilled in crime (the punishment “strengthens the power toresist”; “we must certainly seek the actual effect of punishment primarilyin the sharpening of intelligence”),60 or their wills are broken and theybecome dulled (“if it does happen that a man’s vigour is broken,resulting in his wretched prostration and self-abasement”).61 A thirdpossibility is as good as excluded: “the real pang of conscience, preciselyamongst criminals and convicts, is something extremely rare.”62

In order to understand this, it must briefly be mentioned whereand in whom the feeling of guilt eventually arises. Nietzsche pointedly

57. GdM ii 11. Genealogy of morality, ed. Ansell-Pearson, pp. 52–3.58. GdM ii 11. Genealogy of morality, ed. Ansell-Pearson, p. 53.59. GdM ii 11. Genealogy of morality, ed. Ansell-Pearson, p. 53.60. GdM ii 14 and 15. Genealogy of morality, ed. Ansell-Pearson, pp. 59, 60.61. GdM ii 14. Genealogy of morality, ed. Ansell-Pearson, p. 59.62. GdM ii 14. Genealogy of morality, ed. Ansell-Pearson, p. 58.


summarizes the explanation when he speaks, in Essay ii 22, of the manor woman who has been “incarcerated in the ‘state’ to be tamed, andhas [invented] bad conscience so that he can hurt himself, after themore natural outlet of this wish to hurt has been blocked.”63 We haveseen how the state has monopolized the infliction of suffering bypublic punishment. Now the state is exercising a monopoly of poweragainst the criminal as well. The criminal was indeed not a creditor,but instead a debtor and “someone who has broken his . . . word.”He or she therefore cannot appeal to justice against the punishment.Nevertheless, he or she is still affected like every other human beingby the will to power and feels the infliction of suffering as a joy. Andthe convict is just as well locked up as the normal citizens, if not evenmore so! Then why can the convict, unlike the normal citizens,develop a feeling of guilt?

Nietzsche provides the following explanation:

Nor must we underestimate the degree to which the mere sight of thejudicial executive procedures inhibits the criminal himself fromexperiencing his act, his mode of conduct, as reprehensible as such:because he sees the same kind of action practised in the service ofjustice and given approval, practised with a good conscience.64

At first glance, this explanation seems puzzling because it appears toconfuse the question of guilt and punishment with the question ofdiscrimination. The convict is deliberately discriminated from the restof the citizens because he or she has committed a crime. Normal citizensare handled differently. This normative explanation is irrelevant inNietzsche’s conception, because normative elements do not provideany explanation; instead, they themselves must be descriptivelyexplained.

Yet Nietzsche’s explanation contains two important elements. First,unlike the normal citizens, the convict sees his or her “power to resist”being strengthened most of the time, because of being more heavilyoppressed than the normal citizens; unlike them, the criminal muststruggle against the power of the state so that he or she has nonecessity to require the internalization of what Nietzsche terms theLeiden-lassen, literally “mak[ing] someone suffer.” Secondly, unlikenormal citizens – and especially unlike the victims, who are required

63. The translation renders “erfunden” as “discovered,” whereas it should be rendered as“invented.” GdM ii 22. Genealogy of morality, ed. Ansell-Pearson, p. 68.



to go without a real equivalent (the consoling equivalent is only a“compromise”) – the criminal experiences something that can be heldto be an equivalent by him or her. The criminal is handled in the sameway he or she acted; with the punishment, “debts” are cancelled. Thesuffering of the victim is not seen by the convict, since the state hasseparated the convict from the victim as well as from the revenge ofthe victim. Thus, there is no room left over for “debt.”

The observations of such penal law specialists as George P. Fletcher,according to whom “high recidivism rates suggest that in contrast tothe ideal of rehabilitation, prisons serve more to harden the antisocialinclinations of criminals than to convert them to a law-abiding way oflife,”65 admittedly seem to corroborate Nietzsche’s assessment ofimprisonment. Yet even if the convict is only extremely seldomreformed, then, thanks to the punishment, at least the will of anumber of the criminals will be broken. At least some convicts willtherefore not be recidivists, that is, they will be rendered harmless orrehabilitated thanks to the punishment after their term has beenserved – which is the goal of specific deterrence.

6.4. Does Nietzsche provide a theory of punishment?

What does Nietzsche himself then suggest? The acquittance (Abzah-lung) rests on an illusion. I also do not believe that Nietzsche’s answerwould be expulsion from society. Neither the right of the mastersnor pardon can be an option. Although Nietzsche does not declarehimself to be against cruelty and against the joy that comes frominflicting suffering, anger as a counter-sensation, as ressentiment, is alsonot recommendable in Nietzsche’s perspective. One alternativeremains: “mercilessness and cruelty” against the “hated, disarmedenemy who has been defeated,”66 but indeed without anger: ratherwith joy, without reactive feeling, without ressentiment against theperson who has broken his or her word. The criminal simplyattempted to prevail by his or her breach of promise. Nietzsche hasno moral objections against that, because only the power relation canbe considered as being a norm. The criminal, whose crime is joyfullysanctioned by a gruesome death, simply has had a stroke of luck. Only

65. George P. Fletcher, Basic concepts of criminal law (Oxford: Oxford University Press,1998), p. 31.



the internalization of ressentiment, that is, the internalization of theanger of the powerless human being toward the criminal, is harmfulto the whole process of the prevailing of power, meaning symptomaticfor the triumph of powerlessness.

In fact, Nietzsche speaks of someone who has broken his or herword not as a base, abject human being, but instead like a soldier ina war, a metaphor that most of the time carries positive connotationsin Nietzsche’s works and refers to something noble. The criminalrejects the system of equivalence and as a result considers him- orherself as being incomparable. If the criminal fails, one may indeedfeel joy over the defeat, though one may not judge him or her morally.Nietzsche had already mentioned this option more extensively inthe Second Essay:

Not so long ago it was unthinkable to hold a royal wedding or full-scalefestival for the people without executions, tortures, or perhaps anauto-da-fe . . . To see somebody suffer is nice, to make somebody suffereven nicer – that is a hard proposition, but an ancient, powerful,human-all-to-human proposition . . . No cruelty, no feast: that is whatthe oldest and longest period in human history teaches us – andpunishment, too, has such very strong festive aspects!67

It only begs the question whether one may even speak of a publicpunishment under these conditions. Provided that the criminal ispunished not by private citizens – for example, in the form of alynching, which, incidentally, displays more anger than joy – butinstead by the state, one could, however, finally speak of a publicpunishment without at the same time approving of Nietzsche’s view.

What does Nietzsche’s amoral view show us about public punish-ment as a feast of cruelty? Nietzsche’s argument seems bipartite andtwo-sided. On the one hand, he has recourse to the classical justifica-tion of punishability: the criminal did not comply with the reciprocitythat is fundamental to the commonwealth, and he or she attackedthe commonwealth; thus, the criminal is no longer able to be anormal member of the commonwealth. On the other hand, inNietzsche’s conception, the further classical premise is lacking thatthe criminal being a human being can acquit his or her crime, that is,again become a member of society after the expiation of the crime(not only the theories of rehabilitation but also the theories of



deterrence and the various forms of retributivism allow – except inthe highly problematic and controversial case of the death penalty –the possibility of rehabilitation, as previously stated). In its place,power and cruelty reign. My conclusion is as follows: Nietzsche’s ideaof the most desirable punishment would be a mixed form of justiceand cruelty. Such a mixed theory would admittedly be implausibleas such.

In reality, the matter looks slightly different. Nietzsche’s primarypremise is not the kind of reciprocity that is a form of non-positive,natural law justice, but instead a reciprocity that is one of the forms ofthe “requirements of social life” that it constitutes.68 It is not theseverity of the punishment, according to Nietzsche, that initially dem-onstrates the necessary and powerful “aid to mnemonics”;69 instead,from the outset, it is the exclusion from society or the inacquittabilityof the crime. Nietzsche cites as one of the means of mnemonicsonly the death penalties with torture: stoning, breaking on the wheel,impaling, drawing and quartering or trampling by horses, boilingin oil or wine, flaying, cutting flesh from the body, the smearing ofhoney on a criminal under the burning sun.70 These punishments,being death penalties, all mean expulsion from the commonwealth.The death penalty is exclusion and torture is cruelty; briefly stated,pain is “the most powerful aid to mnemonics”71 that teaches strictobedience.

Nietzsche’s conception of punishment as being a mnemonic isconcerned merely with the prevailing of power of the commonwealthand leaves justice and the rule of law to the respective power relationsin whatever forms they may finally find themselves. A genealogy ofpunishment is what Nietzsche’s explanation is; by the same token, it isalso a resolute recommendation for punishment. However, there issurely no inspiration for a justice oriented debate about punishmentto be found in this recommendation. Apart from that, in our age itwould potentially not be an efficient mnemonic aid to implementNietzsche’s suggestion; instead, it would rather provoke – hopefully– a great deal of resistance as well as major social unrest. If our legalsystems, with all their structural and functional problems, are morestable than Nietzsche’s conception then his single core argument has

68. GdM ii 3. Genealogy of morality, ed. Ansell-Pearson, p. 42.69. GdM ii 3. Genealogy of morality, ed. Ansell-Pearson, p. 41.70. Cf. GdM ii 3. Genealogy of morality, ed. Ansell-Pearson, p. 42.71. GdM ii 3. Genealogy of morality, ed. Ansell-Pearson, p. 41.


miscarried. Thereby our age would differentiate itself fromNietzsche’s conception just as those two situations that Voltaire some-what sarcastically described in respect to retaliative punishment weredifferentiated from one another:

All these laws of blood-soaked politics have their moment in history,and one sees clearly that they are not true laws since they are short-lived.They are akin to that need to eat other people which you sometimeshave in a situation of extreme famine: you eat them no more onceyou have bread.72

72. Chapter 14 of Voltaire, “Commentary on the book On crimes and punishments , by aprovincial lawyer,” in Voltaire, Political writings, ed. and trans. David Williams(Cambridge: Cambridge University Press, 1994), pp. 244–79 (p. 264).


7

WHAT IS THE PURPOSEOF PUNISHING CRIMESAGAINST HUMANITY?

7.1. The characteristics of crimesagainst humanity

The concept of crimes against humanity fulfills two roles. On the onehand, it is intended to remedy the loopholes in the internationallegal system. On the other hand, it constitutes a new kind of crime,that is, a kind of crime that entails characteristics absent from anyother sort of crime.

The first loophole that the concept of crimes against humanity wasintended to close in the positive international legal system was the onethat arose as a result of the impossibility of prosecuting not only acrime committed against the combatants and against the civilianpopulation of the enemies, but also those committed against theirown civilian populations. In this way, the concept of crimes againsthumanity extends the concept of war crimes to include new categoriesof victims. The second loophole that was supposed to be closed wasthe impossibility of applying this extension of the humanitarian inter-national law to crimes already committed. The Nuremberg and Tokyotribunals, which were charged with prosecuting war crimes and crimesagainst humanity committed by Germans and Japanese, respectively,during the Second World War, prosecuted crimes committed beforetheir own institution, that is, before the concept of crimes againsthumanity arose prior to their own founding, thus injuring the basiclegal precept of nulla poena sine lege in favor of the enforcement of aminimal natural law legal framework in international matters. Interest-ingly enough, the International Criminal Court (hereafter ICC) doesnot contain this retroactivity in its Statute, though the national penal

171

legislations contain it.1 A third loophole is only now being graduallyclosed, whereby the ICC was given a jurisdiction encumbered neitherby a statute of limitations nor by the place where the crime has beencommitted, at least in those states where the statute calling it intoexistence was ratified. This jurisdiction is subsidiary, however, sincethe ICC prosecutes exclusively either those cases that the individualnational judiciary authorities did not try, or those in which a clearlyinappropriate decision was made.

The definition of crimes against humanity seems totally able toclose these loopholes, at the very least for the following three reasons.First, an international criminal tribunal such as the ICC, which is byno means a supreme court but a court that has a narrower jurisdiction,and which decides cases without any possibility of appeal, that is, asthe last resort, cannot be responsible for hearing those cases related tothe highest crimes that have already been heard and not grosslyunfairly handled by any domestic court. Not only is the definitionand the treatment of the gravest crimes (in most countries the gravestform of homicide) very different from one country to another (take,for example, American first-degree murder, GermanMord and Frenchmeurtre avec premeditation, which are quite different concepts and arehandled in very different ways); even if they were consistent acrossnational borders, the numbers of cases to be handled, even if only inappeal, would widely exceed the capacity of a single court. Secondly,because crimes against humanity are considered according to theirdefinition as “systematic attacks,” which implies that they are serialcrimes, they are the most significant crimes from a quantitative pointof view. Thirdly, more often than not such “systematic attacks” occureither in a failed state or in a criminal state, that is, for the former, in astate that cannot hinder such attacks, or for the latter, in a state thatcommits such crimes itself. As a result of that, there is a special needfor the subsidiary intervention of an international institution. Inshort, the concept of crimes against humanity is useful as a criterionto allow only the most severe and the most unpunished crimes to betried before such international tribunals. Were it only for these veryreasons, a new category of crime would be required and it has beenestablished in the criminal justice systems of each nation as the mere

1. Retroactivity is still the main reproach addressed to the Nuremberg Tribunal; see LarryMay, Crimes against humanity: a normative account (Cambridge: Cambridge UniversityPress, 2006), p. 211.


consequence of its existence in positive international law in the formof the verdicts pronounced in Nuremberg and Tokyo.

The concept of crimes against humanity, being a pragmatic criterion,appears to be a necessary first step on the path toward harmonizationof criminal prosecution made possible by the establishment of aglobal standard and of an appropriate priority rule for criminal pros-ecutions. Although these pragmatic reasons seem sufficient to justifythe existence of crimes against humanity – a specific sort of crime ininternational law – I suspect, however, that some of the characteristicsof this new kind of crime, as well as the justification given for itspunishment, are most often misinterpreted or are inappropriate.

These aforementioned characteristics are as follows:1. According to the Rome Statute of the International Criminal

Court, crimes against humanity are “widespread and systematicattacks”2 committed “pursuant to or in furtherance of a state ororganizational policy to commit such attack.”3 An individual criminaldestroying an entire village for personal reasons does not commit withthat act any crime against humanity. A crime against humanity is notdirectly a crime against the institutions, as is high treason, abuse ofpower, bribery, etc. Yet crimes against humanity are not merely viola-tions of the rule of law, but instead an attack against the politicalinstitutions themselves; among other reasons, because they clearlyintend to exclude entire groups from obtaining political representa-tion and from the process of political deliberation, as well as possiblyto deprive them of access to due process before impartial courts. I donot agree with Antoine Garapon’s qualification of this as the violationof the “droit d’avoir des droits” (“the right to have some rights”),4

because any victim of an ordinary murder definitively loses this “rightto have rights,” although he or she is not necessarily a victim of a crimeagainst humanity. Crimes against humanity are crimes against basicpolitical institutions.

2. Crimes against humanity abandon the usual sanction structurefound in penal law between the individual crimes, as well as betweenthe individual punishments. Unlike the domestic criminal law of each

2. Art. 7.1 of the Rome Statute of the International Criminal Court (hereafter, RomeStatute): www.icc-cpi.int/NR/rdonlyres/EA9 AEFF7 -5752-4 F84-BE 94- 0A655 EB30E16/0 /Rome_Statute_English.pdf.

3. Rome Statute, Art. 7.4. Antoine Garapon, Des crimes qu’on ne peut ni punir ni pardonner (Paris: Editions Odile

Jacob, 2002), p. 134.

punishing crimes against humanity? 173

www.icc-cpi.int/NR/rdonlyres/EA9AEFF7-5752-4F84-BE94-0A655EB30E16/0/Rome_Statute_English.pdf

www.icc-cpi.int/NR/rdonlyres/EA9AEFF7-5752-4F84-BE94-0A655EB30E16/0/Rome_Statute_English.pdf

Administrator

Sticky Note

Marked set by Administrator

country, such varied crimes as homicide, kidnapping, torture andrape are subsumed under Article 7 of the Rome Statute. Accordingto Article 26.3, committing, soliciting or facilitating the crime, or evenproviding the means for its commission, are all considered to becommitting the crime against humanity. The “applicable penalties”as defined by Article 77 are the same for all of these deeds, whereasthe amount of punishment greatly varies in the domestic legal con-texts, according to which of these deeds was committed. Neither ofthe two options for assigning the amount of punishment (eitherimprisonment for a certain number of years or life imprisonment) isin any way specifically related to one of the categories of crimes thatare treated separately in the basic penal codes of the individualnations. We can see one of the consequences in the case of a personA committing a crime incited by a person B. The nature of theincitement by B, however, does not amount to creating a state ofduress for A. Ordinary penal law shall not always punish incitementto a crime; in such cases it punishes only incitement to the gravestcrimes, and then it punishes them mostly less than the commission ofthe crime itself. In the Rome Statute there is no mention of a mitiga-tion of punishment for the incitement to a crime compared withcommission of a crime. On the contrary, the Rome Statute does notcontain anything that would prohibit a reverse sanction structure(such as punishing someone less for a crime depending on the levelof participation of that person, which goes against the usual sanctionstructure in most countries). In fact, a political leader who has com-manded large-scale massacres without having ever killed anybodyhimself may be subject to a severer punishment than a soldierwho has carried out his orders and is also declared by the court notto have acted under duress.

3. Crimes against humanity are not subject to the statute of limita-tions, whereas in most countries even the highest crimes are subject tothe statute of limitations. Even in countries such as Germany, in whichmurder is not subjected to any statute of limitation, most of the crimes(rape, kidnapping, etc.) included in Article 7 of the Rome Statute ofthe ICC are subject to the statute of limitations when they are notcommitted in the context of a crime against humanity. First andforemost, at the time Germany decided to exclude murder from thestatute of limitations, it did so precisely in order to ensure thatthe prosecution of crimes against humanity that were not yet partof the penal code could still be made possible; were this not the


intention at the time, murder would most likely have been subject tothe statute of limitations. In Austria, there is no statute of limitationsfor murder, but there is a dramatic reduction of the amount ofpunishment should the case be brought twenty years after the execu-tion of the crime.5

4. The exclusion of crimes against humanity from the statute oflimitations is often considered as being morally justifiable by themonstrousness that all crimes against humanity pose. This monstrous-ness is usually considered as implying two consequences. First, itimplies that those who commit crimes against humanity are supposedto be monsters who deserve extraordinary punishment. Secondly, itimplies that the monstrousness that crimes against humanity repre-sent should remain unforgettable, because guilt is allegedly incom-mensurable. Thus, it is held that punishment should have anexpressive function, that is, that it should represent an expression ofthe inextinguishable, most deep-seated abhorrence of all humankind.In this view, no punishment can fully fit the crime, when the crime is acrime against humanity, yet an extraordinary punishment can at leastbe regarded as an expression of the inextinguishable guilt of thecriminals. The monstrousness of some crimes against humanity, aswell as the impossibility of finding any commensurable punishment tofit the gravity of the crime, is often employed as the justification forthe existence of the legal category of crimes against humanity, andit implies not only the exclusion of the statute of limitations (asmentioned in the first point), but also the elements mentioned inthe first and second points, though interpreting them in a peculiarway. Indeed, the absence of the usual precisely codified sanctionstructure (the second point) is explained in this context as the mani-festation of the fact that the crime is so monstrous that it lies beyondthe reach of the sanctions in ordinary penal law and of any quantita-tive scale of guilt and amount of punishment. But what about pointone, that is, about the political significance of crimes against humanityas it is treated by the argument that the monstrousness of crimeswarrants sanctions above those against conventional crimes? Thosewho explain the specificity of crimes against humanity as resultingfrom their monstrousness do not emphasize the endangerment of thepublic institutions brought about by the commission of the crime;

5. Austrian Penal Code (Strafgesetzbuch), Sec. 6, Art. 57(1). See May, Crimes againsthumanity, p. 216.


rather, they emphasize that committing the crime is a moral evil, thatis, the alleged inability of the criminal ever to belong again to thepolitical community under the rule of law.

7.2. An alternative conception of crimes against humanity

In most conceptions of crimes against humanity, the preceding fourcharacteristics create a complex that I would like to challenge.I believe the first characteristic to be true, as well as the second, butI would justify the second in a completely different way, and I willentirely reject the third and fourth ones. Therefore, I will begin myinquiry into all four characteristics with the first one.

Crimes against humanity are crimes against the basic political insti-tutions of states governed by the rule of law. I would like to introducea distinction that is essential for determining the appropriate amountof punishment for this crime: I shall distinguish between (A) the timeat which it is being committed and (B) the time after it has beencommitted.

(A) We find in the tradition of natural law two rights that areintended to provide a remedy to a situation in which a crime againstthe very institutions of a legitimate political community is beingcommitted, that is, tyrannicide and the right of resistance. By defin-ition, this situation is one in which the rule of law is not enforced, andthus no action before a court is possible. Therefore, tyrannicide andright of resistance cannot be codified by positive law. The GermanBasic Law (Grundgesetz) declares: “All Germans shall have the right toresist any person seeking to abolish this constitutional order, if no otherremedy is available.”6 Yet, for the reason justmentioned, the law cannotorganize the exercise of this right to resistance; and it belongs not tothe penal law, but to the constitutional law. Tyrannicide and rightof resistance are no punishment of the tyrant, but rather provide forthe tyrannicide a legal excuse for performing an action that wouldbe, under normal circumstances, a punishable offense. I would like todiscuss, however, the punishment of crimes against humanity.

What would then comprise the goal of punishing a person who is atany given time in power and is committing a crime against humanitywhile holding office? It seems to me that the purpose of punishmentcan be either – most importantly – incapacitation, that is, in this case

6. Art. 20(4) of the German Basic Law.


the removal of the criminal who commits a crime against humanityfrom power, or general prevention, that is, in this case, deterrence ofthe members of foreign governments from committing a crimeagainst humanity. Yet, the prosecution under the charge of crimeagainst humanity, as it is defined not only in international law butalso in internal law with its already mentioned characteristics, is eithera superfluous or a useless tool. It is superfluous when it is possible toindict and to depose a member of the government or an entiregovernment under charges derived from ordinary penal law, and theparliament is not so corrupt as either to tolerate or to support thecrimes committed by the government. It is useless when either no suchprocedures of destitution or impeachment are foreseen by the consti-tution, or the criminal members of the government exercise such apower or threaten such violence that nobody will dare to prosecute.Prosecution may be impossible not only for reasons of internal dys-function, but also for reasons of internal or international immunity.Members of governments, members of parliaments, diplomats, etc.,enjoy an immunity grounded in international law that protects themeven against prosecution of crimes against humanity. Since anyhumanitarian intervention employing military force is prohibited,except in the presence of aggression and with the consent of theUnited Nations Security Council, international law seems clearly notto head in the direction of a prosecution of crimes against humanitywhile the criminals are still in office. The concept of crimes againsthumanity, as it is understood by today’s human rights law, does notprovide the tools necessary for the clearing of the legal obstacleson the path toward intervention against a government that violatesmassively the fundamental rights of its citizens.

In fact, if the arguments made in favor of an unconventionalpunishment for crimes against humanity were to emphasize a deter-rent effect at all, then it is the general deterrent effect that is supposedto result from sentencing the criminal political leader after he orshe has left office. Effective prosecution and conviction of politicalleaders allegedly deter other political leaders from committing crimesagainst humanity. Yet, first, no evidence has ever been provided thatpolitical leaders who see in crimes against humanity the only meansavailable either to promote political goals – the righteousness of whichthey may be truly convinced of – or simply to remain in office can bemotivated to change their mind under the influence of the punish-ment of their foreign colleagues, although the motive for promoting


political goals they are deeply committed to or of remaining in officeis certainly very strong.7 Secondly, were the aim of punishment deter-rence, punishment should then be imposed directly following thecommission of the crime, that is, while the perpetrator still occupiesoffice, and not after he or she has eventually been deposed.

(B) Let us now investigate the situation after the criminal hasbeen deposed. Once the criminal has been divested of power, thepreventive aim of punishment may be either: general deterrence –about which I have expressed my aforementioned doubts – or specificdeterrence, which implies hindering the criminal in regaining power.Traditionally, three options are available in the case of a leader whohas committed crimes against the institutions of his own state and hasbeen deposed (for example, in cases of conspiracy, high treason,abuse of power, etc.), with each option corresponding to a character-ization of the former leader’s traits as a leader, as well as of thesituation in which he or she is presently. Let us assume that the formerleader poses an immediate and acute threat for the institutions,because of having at his or her disposal as many supporters, as manymeans, as many weapons, etc., as he or she sees fit in order to have achance to regain power. The first possible characterization impliesthat the former leader would lose this ability to regain power again ifhe were no longer to be present within the state’s territory. Theclassical solution for the aforementioned: banishment. This solutiondoes not have anything to do with the suggestion made by AaronFichtelberg to banish the criminal from all of human society.8 In fact,banishment from all of human society ultimately amounts tocondemning a person to an all-too-likely death, like that of a danger-ous, wild animal, as Fichte demonstrated. The second characterizationis that, even in exile, the criminal would remain as much of a dangerfor the state as if he or she were actually present in the state. Theclassical solution for the aforementioned: death. Finally, the thirdcharacterization of the former leader is that, after being deposed,the former leader no longer poses a threat to the state’s institutions.The former leader is thus condemned in order to sever the tiesbetween the former leader and the people, as well as to demonstratethat the succeeding government does not share the same intentions as

7. See Aaron Fichtelberg, “Crimes beyond justice? Retributivism and war crimes,” CriminalJustice Ethics, 24, no. 1 (2005), 31– 46 (p. 34); on p. 35 he rightly says that the threat ofpunishment might even be counterproductive.

8. Fichtelberg, “Crimes beyond justice?”


the preceding leader so that the crimes the earlier government commit-ted can be strictly outlawed. But, in general, the condemnation is shortand a pardon is rapidly granted.9 If necessary, a punishment may beadded to someperiod of ineligibility for office. The old adage states thatmost of the time in politics punishment is either death (the secondcharacterization) or a short prison term (the third characterization).

It is interesting to note that in positive international law none ofthese three options is available to a tribunal such as those of theformer Yugoslavia, of Rwanda or of the ICC. Capital punishment hasbeen abolished by a significant proportion of the members of the UN,as well as by the ICC in The Hague, and no exception is made forcrimes against humanity – unlike the exception made for the statuteof limitations; the gravest crimes against humanity carry no greatersanctions than the worst instances of first-degree murder. After theNuremberg and the Tokyo trials, capital sentences were no longerpassed by an international criminal court. Were capital sentencesallowed to be passed, most Western democracies would not participatein those trials, because most of them are expressly prohibited fromextraditing to a court that can pass capital sentences – even in casesof crimes against humanity. Banishment is not an option becausethe Universal Declaration of Human Rights (Article 9), the EuropeanConvention on Human Rights (Article 7(1)) and the Final Act ofthe Conference on Security and Cooperation in Europe (HelsinkiAccords of 1975, Principle 10) guarantee an individual’s access tohis or her own country at all times. Admittedly, were the criminalleader to be judged by an international court, rather than by a domes-tic jurisdiction of the country of which he or she is a national, thesanction would then probably be carried out outside the country,which may be considered as de facto banishment. Finally, the sanctionfor the leaders of crimes against humanity cannot be a short-termsentence, but must be either a life sentence or a long period of time inprison. Two astounding points come to light when all of the afore-mentioned is considered as a whole:

1 The punishment for crimes against humanity radically divergesfrom the traditional punishment of crimes against legal and against

9. An amnesty may even be a necessary means with which to ensure a just peace; seeMichael P. Scharf, “Justice vs. peace,” in Sarah B. Sewall and Carl Kaysen (eds.), TheUnited States and the International Criminal Court: national security and international law(Lanham, Md.: Rowman and Littlefield, 2000), pp. 213–36 (p. 189).


political institutions, even though both crimes are closely related bytheir very nature.

2 The severity of sanctions handed down for crimes against humanitydoes not exceed the most severe sentences already existing forother crimes, whereas those other crimes, in most countries, aresubject to the statute of limitations.

I will come back to a discussion of the aforementioned observations ata later point. At this point, I will consider punishment of executantsmerely acting upon orders given to them by the politicians. I will againdistinguish between the situation during the commission of the crimeand the situation after the crime has been committed.

It seems obvious that during the commission of the crime nosignificant number of executants, if any, can be prosecuted, if thereshould happen to be a systematic “state or organizational policy” inthe framework in which they find themselves carrying out their crim-inal actions (Rome Statute Article 7.1). After the crime has beencommitted, the most important specific deterrent measure that mustbe taken is the dissolution of all criminal political organizations thathad planned the crimes against humanity, guided them and, finally,carried them out. Yet, what might be the purpose of punishing thenumerous individual executants? As I have already mentioned, one ofthe characteristics of crimes against humanity – as is also the case withwar crimes – is that, on the one hand, command cannot be employedas a justification of the crime, and on the other hand, the commissionof crime is not considered as being more serious than mere solicita-tion of an offense, in short, implying that leaders and the executantsare all responsible for the crime against humanity.

Yet, this is in the realm of pure theory, since only a very smallproportion of the executants are prosecuted and convicted, and theyare usually sentenced to less severe punishment than the leaders,moreover to sentences not any severer and often more lenient thanfor ordinary crimes (one thinks of a case heard before the Inter-national Criminal Tribunal for the Former Yugoslavia in 1996, inwhich Drazen Erdemovic, a multiple murderer, was sentenced to onlyfive years in prison owing to mitigating circumstances).10 Even volun-tary membership in an organization officially classified as “criminal”by the Nuremberg Tribunal, such as the Schutzstaffel (SS) or the

10. See the Sentencing Judgment in the case of Prosecutor v. Drazen Erdemovic, available atwww.un.org/icty/erdemovic/trialc/judgement/erd-tsj980305e.htm.


http://www.un.org/icty/erdemovic/trialc/judgement/erd-tsj980305e.htm

Gestapo, has not been systematically punished, only the gravest casesbeing prosecuted. The sanctions enforced on a few executants can bereally efficient neither from the point of view of specific deterrence(a few isolated individuals alone are not able to rebuild the criminalorganizations to which they belonged), nor from the point of viewof general deterrence (were the probability to be sentenced extremelylow, the deterrence effect would not be sufficient to justify suchmeasures). A much more efficient solution, from the point of view ofgeneral deterrence, would be a massive punishment of all executants,were it possible at all.

If such a massive punishment would meet with resistance fromthe criminal who was indicted and sentenced, that would threatenthe existence and stability of the state under the rule of law, and if thekind of very limited moderate punishment that I have describedwould allow it to restore the state institutions to the rule of law, alltheories of punishments, whether deterrent or retributive, wouldprefer the latter option. However, there is a quintessential differencebetween deterrent and retributive theories with respect to this solu-tion. A retributive theory of punishment considers it as an exceptionmade to the law, which can be made only in order to protect theexistence of the law in a situation in which the enforcement of the lawis not possible. In the Doctrine of right, Kant evokes the situation inwhich “the number of accomplices” involved in a homicide “is so greatthat the state, in order to have no such criminals in it, could soon finditself without subjects; and . . . the state still does not want to dissolve,that is, to pass over into the state of nature, which is far worse, becausethere is no external justice at all in it.”11 According to Kant, who,at this point, undoubtedly belongs to the retributivist school, themalefactor in such a situation should be sentenced to a mitigatedsentence, yet not through the letter of the law, but instead througha decree of pardon issued by the monarch. For a theory of deterrence,these more lenient punishments provide no exception to the law ofcriminal justice, but are instead part of it, because penal law must, asits sole purpose, ensure the enforcement of the rule of law as quicklyas possible. In this regard, once the danger has passed that themalefactor who committed crimes against humanity might regainpower under normal rule of law, the probability of recidivism willdisappear. The probability of recidivism is much lower than with other



crimes, and because of that the punishment should likewise be muchless severe than what is usually prescribed by the normal penal code.

I will come back later to the retributivist position. The conclusionthat I will first draw from these points may be perceived prima facieas repugnant. Because the characteristics that typify crimes againsthumanity are formed by their political dimension and by the lackof the usual sanction structure, they ought to lead to more lenientpunishments than do ordinary individual crimes. General andspecific deterrence simply do not require the same means, in casesinvolving crimes against humanity, as cases involving ordinary crimesperpetrated on a more individual scale.

This conclusion might well become even more repugnant were weto carry out the following thought experiment. At this point, beforegoing any further, I would like to stress the point that I have absolutelyno sympathy for the person, or the ideas, mentioned below, whosepolicies – especially his policy advocating wholesale extermination ofseveral ethnic, political and religious groups – rightly deserve thedeepest moral condemnation and the most vehement rational rejec-tion. But I would like to make reference to the reasoning used by theFrench defense attorney Robert Badinter, who finally obtained theabolition of capital punishment after he became the French Ministerof Justice, after having struggled against it for his clients beforenumerous criminal courts. Badinter explained that the more antip-athy a criminal case provokes, the more clearly pleas may be made onthe grounds of principle.12 Let us begin our thought experiment byimagining that the New York Times discovered, over forty years after theclose of the Second World War, that Adolf Hitler did not die in hisbunker in Berlin in April 1945, but instead escaped and ever since hadbeen living inconspicuously somewhere in the Irish countryside.Except for retributivist arguments, I see no criminal justice theoreticalargument that could be presented to demonstrate that he should stillbe punished after forty years; at that point, he no longer poses a threatto any political institutions. In order to be absolutely sure that hewould not pose any further threat, one could consider depriving him,at the very most, of his right to free speech. Drawing this repugnantconclusion, that is, pleading for the imposition of the statute of limita-tions, even in cases involving the most despicable criminals, contradictsone of the aforementioned characteristics of crimes against humanity,

12. See Robert Badinter, L’Abolition (Paris: Fayard, 2000).


but it does not necessarily mean making a plea for the lifting of thefinding of guilt, as demonstrated by the Commission for Truth andReconciliation, a body that was intended to establish the facts but notto make any condemnations.13 However, I suspect that a certainamount of time in prison would have been necessary to rehabilitatesome very serious criminals of the apartheid era because they hold themost firmly rooted convictions. I will confine myself only to forward-ing the thesis that in cases of crimes against humanity the conven-tional prescriptive period should never be exceeded, but instead evenreduced.

7.3. The boundaries of our moral intuitions

As I mentioned above, a rejection of this repugnant conclusion canonly rely on a retributivist theory of punishment. Such a theory typic-ally refers to the fourth characteristic of crimes against humanity thatI mentioned in Section 7.1. Retributivism rejects my repugnant con-clusion, because the punishment can never be equal in gravity to thecrime; this is, for instance, Hannah Arendt’s thesis. Clearly, the mon-strousness of all crimes against humanity is apparent, but the mon-strousness of the crime cannot define a wholly new category of crime.

A cannibal who has eaten his consenting victims is, in my view, moremonstrous than someone who kills for stealing a million dollars. Yet,the latter will be sentenced to a more serious punishment. The lifesentence will be (for instance, in Germany) the same as for someonewho tortured and killed a baby before the eyes of its mother, whichmost of us will find more monstrous. And there are many actions thatare monstrous and unpunishable. Someone may cynically declare,while eating in a luxury restaurant, that what makes him particularlyenjoy the meal is the thought that so many “useless” people are dyingof starvation somewhere in Africa. Or someone may declare to herdying life partner at the hospital that she never loved him, hadbetrayed him, and then leave without even saying good-bye. Both aremonstrous actions, though fully legal and protected rights. Whether ahomicide, a rape, a kidnapping, a manslaughter, etc. is monstrous isneither a criterion for ascribing it to the category of crimes againsthumanity, nor an element of the crime itself. The purpose of penallaw, as is the purpose of the legal system in toto, is not to retribute moral

13. See May, Crimes against humanity, p. 239.


evils and perversions, but to protect society, that is, the rights of thecitizens. This implies that penal law amounts to a means to restore therule of law between all persons – including the criminals – as quicklyand as efficiently as possible.

If we were now to formulate the retributivist objection to my repug-nant conclusion in the guise of a classical criticism, we should saythat the retributivist would postulate that crimes against humanityare deserving of severer punishment than similar crimes not barring“widespread and systematic attacks”14 committed “pursuant to or infurtherance of a State or organizational policy to commit such attack.”15

However, I see several serious difficulties with such a view being utilizedas a justification for disregarding the statute of limitations.

1 Were we to compare a homicide committed within the frameworkof crimes against humanity with an “ordinary” homicide perpet-rated on the individual level, we would be hard pressed to postulatethe former as being more deserving of severer punishment. Yet,should a campaign of discriminatory, forced sterilization be pun-ished more severely than, say, an ordinary first-degree murder? Theanswer is not at all obvious.

2 Should a crime against humanity consisting of a massive extermin-ation of human beings utilizing industrial methods be punishedmore severely than a conventional first-degree murder? Were thelatter already punished by a life sentence, I see only a death sen-tence, or more likely a long and public torture, as a severer punish-ment. Yet, would that not constitute in itself a crime againsthumanity? The letter and the spirit of international law punishingcrimes against humanity are clearly not shaping its course towardsuch extreme punishments. (For this reason I disagree with LarryMay, who sees proportionality as requiring restraint in punishmentin relation to “emotional factors”: I see it, on the contrary, asheading in the same direction.)16

3 Retributivism is not compatible with the lack of an outline of thetraditional sanction structure inArticles 7 and 25 of theRomeStatute.

One may object to my repugnant conclusion stating that penal lawshould not offend popular sentiment. Allowing Adolf Hitler refuge in

14. Rome Statute, Art. 7.1.15. Rome Statute, Art. 7.2.16. May, Crimes against humanity, p. 215.


the harbor of the statute of limitations would severely offend popularsentiment at least as much as not punishing crimes against humanitymore so than conventional crimes. One can understand this argumentin one of the following two ways. Either this argument posits that nomoral judgment can be morally true, were it to contradict popularsentiment. Or the argument posits that such a repugnant conclusionwould lead the people to overthrow the judiciary power and possiblyeven the rule of law.

Popular sentiment cannot be allowed, however, to be a criterionfor the moral judgment of punishment, among others thingsbecause it is inconsistent in many regards. Popular sentiment mixesretributive, general and specific deterrent elements. It can protestagainst punishments that seem too lenient, but it can allow the victimof a car accident to receive much greater compensation than thevictim of rape or of serious physical injuries perpetrated by an insolv-ent criminal. Furthermore, common sentiment is more often prone toimmoral judgments. For centuries it authorized criminal trials andsentences against animals or against witches; it authorized tortureeither as a punishment or as a method for interrogating a suspect;it authorized the punishment of an entire family for a crime commit-ted by one of its members. All this and more was fully acceptable tothe common sentiment. Nowadays, death by starvation of millionsof people is still not perceived in common feeling as a violation ofbasic human rights, although such texts of international law asthe Universal Declaration of Human Rights (1948) rightly recognizesubsistence as a fundamental human right. Furthermore, in manycountries, a majority of the citizens is still in favor of capitalpunishment, as well as of increasing restrictions on the right ofasylum.

Were one to consider the risk of the people possibly overthrowingthe institutions of the judiciary, as well as the political institutions,if crimes against humanity were to be subject to the protections ofthe statute of limitations, there would be no evidence in favor ofthis view. South African institutions of state have suffered no harmto their prestige by not punishing the criminals of apartheid,17

and demonstrations against the impunity of dictators who

17. The opposite is true: without the amnesty granted to a section of the perpetrators ofthe apartheid regime by the new government there would have been a real threat ofmilitary upheaval in South Africa: see chapter 6 of Amy Gutmann and DennisThompson, Why deliberative democracy? (Princeton: Princeton University Press, 2004).


committed crimes against humanity have never endangered thedemocratic rule of law.

For these reasons I see no valid moral objection against applyingthe statute of limitations and against mitigating punishment. Thoughthe existence of the concept of crimes against humanity is a prag-matic criterion for judging the gravest crimes, and though it is thenecessary first step on the path to the harmonization of criminalprosecutions by setting a global standard and an appropriate priorityrule, these pragmatic reasons seem to me not to justify the existence ofcrimes against humanity as a fundamentally new type of crime thatrequires new types of punishment and goals of punishments, as well asjustifying especially severe prosecutorial precepts.

Treating crimes against humanity pragmatically does not imply,however, that they do less harm than conventional crimes. In fact,exactly the opposite is true: crimes against humanity are especiallymonstrous. This paradox reminds us that it is not the role of penallaw to take revenge on criminals for our past powerlessness. Rather,the role of penal law is to protect against future crimes and tosecure the rights of the victim – as well as of those of the perpetrator– within the rule of law. The victims of crimes against humanity mightwell be more fortunate if governments who self-sufficiently congratu-late themselves for having created the ICC – which I certainly find amost useful institution of global justice – actually developed the rightof asylum rather than reduced it. The central issue should be notwhether or not Milosevic should be sentenced to life in prison or toonly four years behind bars; instead, the central issue should be howbest to wipe torture off the face of the earth.18

18. In this regard, Andrew Altman and Christopher H. Wellman, in “A defense ofinternational criminal law,” Ethics, 115, no. 1 (2004), 35–67, confuse the role of thepunishment of criminals against humanity with the role that ought to be played bymilitary humanitarian intervention.


CONCLUSION

In Part i, we examined the Kantian, retributivist thesis, according towhich the criminal is punished, because he or she merits it. Two interpret-ations of the Kantian concept of right are possible (see Chapter 1).On the one hand, according to the liberal interpretation, which doesnot refer to the internal disposition of human beings, right is definedmerely as being the coexistence of the freedom of action of allhuman beings according to the principle of equality of right. Onthe other hand, according to the moral interpretation, which substan-tially refers to the internal dispositions of human beings, rightshould implement the content of the categorical imperative as far aspossible through the application of coercion. In accordance with bothof these interpretations, two dimensions can be distinguished in theKantian retributivist thesis: one genuine dimension of legal ethics,which is a dimension independent of the internal dispositions ofindividuals (see Chapter 2), and a dimension of personal ethics (seeChapter 3). The dimension of legal ethics of the Kantian retributivistthesis contradicts the requirement for the coexistence of all freedoms,of which the freedom of the criminal is a part too, and of which itought again to become a part – at least in the most direct way. Thedimension of personal ethics contradicts the postulate of the highestgood, which rather requires forgiveness toward the criminal as long asthis forgiveness is compatible with the concern for the safety of theother citizens. Hence, Kant’s concept of right as well as his concept ofvirtue require rehabilitation instead of retribution, which ought tooccur after the shortest possible period of specific deterrence, that is,after a period of incapacitation.

In Part ii, I showed how this justification of punishment as arehabilitation occurring after a period of specific deterrence was

187

supported on the basis of the liberal interpretation of Kant’s conceptof right by Fichte (see Chapter 4) and Hegel (see Chapter 5): thecriminal is punished, in order that he or she might once again become partof the commonwealth. Fichte begins with the question of the treatmentof the criminal and by a thought experiment initially draws thelogical conclusion of crime. The logically consistent consequenceof it would not be retaliation, but instead the exclusion of thecriminal from the commonwealth and the revocation of his or herlegal status. The public punishment derived from right as suchemerges as an alternative to this fate that is better for the criminalso long as the alternative punishment amounts neither to the deathpenalty nor to torture but rather to a sentence of incarceration. Suchan expiation, therefore, is not an evil, but rather a chance, whichFichte constructs in such a way that the criminal can be reformedand then finally reintegrated into the commonwealth. Hegel, in turn,chooses as his guideline not the treatment of the criminal, but thereestablishment of right as the negation of the negation – in whichcrime is the negation and the punishment is the negation of thenegation. In so doing, the “negation” of the negation is in no case, asit is often mistakenly assumed to be, of the same sort as the simple“negation” is, and thus it should not be taken to be a response to theevil of the crime by a second evil. Contrary to many interpretations,Hegel is not a retributivist, even though he – like Fichte as well –assigns to retributivism a (merely) instrumental, psychological role inconveying to the convicted criminal the significance of the punishment.

In Part iii, it finally turned out that retributivism could often demandpunishment without any respect for humanity in the person of the criminal.Nietzsche offers a plausible account of the genesis of retaliativepunishment (see Chapter 6), in which the motive of the institutionof such punishments stems not from the respect for human dignity,but rather from a universal human cruelty toward the criminal,which hinders rather than awakens the emergence of bad conscienceand remorse in the criminal. When the state establishes a monopolyon violence and guarantees public security, the people are deprivedof the exercise of this cruelty and punishments become comparativelymild, such as those we have known since the end of the eighteenthcentury. In order to keep the promise I made in the introductionto draw conclusions resulting from the rehabilitative alternative toretributivism, I finally argue (see Chapter 7) for a treatment of even

188 conclusion

the most serious and most inhumane criminals – in other wordsthose who have perpetrated crimes against humanity – that is nomore severe than how other criminals are treated. Even thesecases ought to be subject to the statute of limitations: the degreeof punishment ought not to be higher – which is to say, proportionalto the crime – but ought, instead, to tend to become less; theidea ought to be abandoned that the monstrousness of such crimesjustifies the punishment – and indeed, the most severe possiblepunishment.

Further conclusions can be drawn from my critique of retributivismas well as from my alternative proposal. At this point, I would likebriefly to outline how the penal system should be reformed accordingto my alternative proposal.

1. There ought to be no fixed duration within the judgment forincarceration. Instead of designating a certain duration, the degree ofpunishment ought to be determined by a goal. The criminal ought toexpiate the punishment for as long as he or she still poses more of adanger to the commonwealth than the other citizens do. In order forthat to occur, the criterion employed to suspend punishment maynot be the absolute safety of society, because every “normal” citizenwho consistently remains true to the law could, at any time, commit acrime. Even now, punishments are reduced by sentences of parole,and in the case of good behavior prisoners are released before theirsentences are up. Others are allowed to work outside the prisonduring the day. Such leeway, already customary nowadays, oughtto be expanded to the general abolition of every designation ofduration.

2. The principle of proportionality between crime and punishmenthas to be rejected. Admittedly, it is to be expected that, on average,thieves can be rehabilitated in a shorter period of time than multiplemurderers who have tortured their victims. One would, therefore,probably observe a clear, statistical correlation between the gravity ofthe deed and the length of the process of rehabilitation. But this linkshould not be a matter of principle, and there would certainly also beexceptions.

3. A reform of penal law ought not to involve punishments becom-ing generally severer or generally milder. In effect, some punishmentswould become milder, others severer, all depending on the behaviorof the individual convict.

conclusion 189

4. Psychological assessments of the convicts should occur regularlyand they should be thorough. A process such as the following isconceivable. At the beginning of serving a sentence the criterion fora discharge would be especially strong. The criterion would weaken astime passes and would finally reverse so that gradually the burden ofproof might stand on the side of the commonwealth to prove that theconvict must, in fact, remain in prison.

5. Retributivism likes to contend that only retaliation is just for thevictim and every other punishment would be unjust, for the criminalas well as for the victim. In the current penal system, this argumentrather appears to be an empty consolation for the fate of the victimthat turns out to be scandalous. Anyone who suffers grave bodilyinjury is better off having it occur in the course of an automobileaccident for which he or she is not liable, rather than by a serialkiller, because the damages awarded to victims of criminal acts areespecially low. One should make it clear to the public what a criminaltrial really means. Whereas potential compensation for damages,including pain and suffering, should satisfy the claims of the victims,punishment applies only to the interests of the commonwealth asa whole regarding public security. Therefore, a twofold process isconducted, even if both concerns are to be dealt with in the sametrial. Instead of exclusively and systematically calling for harsherpunishments as a way to win popular support, politicians shouldprimarily think of a better and more respectful compensation forthe victims.

6. An attempt ought always to be punished in the same degree asthe crime itself. This is not the case nowadays. Instead, attempts atonly some of the graver crimes are punishable, while carrying a lesserdegree of punishment. Feinberg convincingly traced back this situ-ation to the conjunction that retributivism wants to create between thegravity of a deed and the wickedness of the criminal:

I find no intuitive plausibility at all in first basing criminal liabilityon moral blameworthiness, but then basing moral blameworthiness inturn upon the actual harm or absence of harm caused. The ancientview, that liability should be based not upon blameworthiness at all butinstead directly upon the amount of harm caused, seems to me morehonest, though no more plausible. It seems almost as if the retentionistis so fixated on actual harm that he keeps searching for the questionto which it is the right answer. Not the question: “What ought to bethe basis of criminal liability?” Moral blameworthiness is a plausible

190 conclusion

answer to that question. Not the question: “What is the basis of moralblameworthiness?” The traditional multiplicity of culpability conditions,motives, mitigations, aggravations, and so on, answers that question. Howabout the question: “What is a necessary condition for tort liability?” Now,that is more like it.1

In this, modern retributivism encounters the question of how theinner, moral guilt of a person should be punished in a legal system,if the system intends to be in charge only of freedom of action, andnot of morality.

Beyond these concrete suggestions for the reform of penal law, asthe result of my inquiry I must address four points of more generalinterest.

First, penal law cannot be about the (purely moral) wickedness ofhuman beings, but instead only about the observance of the laws orabout the infringement of right. Even if we were all “devils,”2 as Kantnotes in Toward perpetual peace, it would still not mean that we all oughtto be punished with deterrence in mind, before having committed anymisdeed. Only those who cannot be hindered from committingcrimes by moral or external reasons should be punished.

Secondly, it is not retributivism, but the justification of punishmentas a rehabilitation that is compatible with that path via which Kantwants to reach global “perpetual peace,” the condition by which thestate of law would reign over the entire world. The modern traditionof natural right that Kant critically draws upon recognizes three justintentions for waging war: legitimate self-defense, recovery of one’sown goods stolen by the enemy and punishment of an unjust enemy.3

This “punishment” of the vanquished enemy is intended to meandeterrence by the theorists of just war – for instance, Hugo Grotius,Samuel von Pufendorf, Emer de Vattel, etc. The aim is a peace that isjust toward both parties. In this respect, punishment is allowed only inso far as to bring about security for the victor. This theory pursues thegoal of ensuring that no future injustice is carried out that wouldpresent a just ground for a further war. If this deterrent intent turnsout to be efficacious, there will then be a continuously decreasing

1. Joel Feinberg, Problems at the roots of law: essays in legal and political theory (Oxford: OxfordUniversity Press, 2003), p. 100.

2. Cf. ZeF Ak viii:366. Practical philosophy, ed. Gregor, p. 355.3. Cf. Book 2, Chapter 1, ii.2 of Hugo Grotius, The rights of war and peace, ed. Richard Tuck

(3 vols., Indianapolis: Liberty Fund, Inc., 2005), pp. 395–6.

conclusion 191

number of wars. Since the victor has already obtained supremacy,the punishment is often mild. Herein lies the strategy of the theoryof just war for the preservation of peace.

Let us now imagine what consequences a retributivist conceptionof punishment might have for this theory. If the vanquished enemyshould fail to understand that the victor’s case is just, the retributivepunishment by the victor will be perceived by the vanquished asbeing unjust. Unlike in national law, in international law there areno higher-ranking judges, so that every sovereign is his or her ownjudge, and although there cannot be two parties whose suits areultimately equally just, there can be two bona fide parties waging warwhose reasons for war appear completely legitimate and prima faciejust. Unlike a punishment that is meant to deter and to protect againstthe recurrence of the grounds that led to war, retributivism requiresthe vanquished not only to assume liability for all damages caused inthe war (compensation, reparations, etc.), but also to serve a severepunishment. If the vanquished enemies waged the war in good faith,this retributive punishment will then be perceived by them to beunjust and thus may consitute a possible causa iusta for a future war.For this reason, such a retributive punishment stumbles into a contra-diction with Kant’s primary objective, that is, with perpetual peace.Kant’s first preliminary article deprives every retaliation of its pre-requisite by requiring that possible discoveries about the past shouldnot be taken into consideration:

Causes for a future war, existent even if as yet unrecognized by thecontracting parties themselves, are all annihilated by a peace treaty, nomatter how acute and skilled the sleuthing by which they may be pickedout of documents in archives.4

Even though Kant had the material claims of the warring partiesin mind in this passage, this remark certainly also applies in the realmof penal law. Even in the Doctrine of right, Kant recommends the samestrategy for (civil) peace on the national level, as the following repeat-edly cited passage shows:

Accordingly, every murderer – anyone who commits murder, ordersit, or is an accomplice in it – must suffer death; this is what justice, asthe idea of judicial authority, wills in accordance with universal laws thatare grounded a priori. – If, however, the number of accomplices (correi)

4. ZeF Ak viii:344f. Practical philosophy, ed. Gregor, p. 317.

192 conclusion

to such a deed is so great that the state, in order to have no suchcriminals in it, could soon find itself without subjects; and if the statestill does not want to dissolve, that is, to pass over into the state ofnature, which is far worse because there is no external justice at all init . . . then the sovereign must also have it in his power, in this case ofnecessity (casus necessitatis), to assume the role of judge (to representhim) and pronounce a judgment that decrees for the criminals asentence other than capital punishment, such as deportation, whichstill preserves the population. This cannot be done in accordance withpublic law but it can be done by an executive decree that is, by an act ofthe right of majesty which, as clemency, can always be exercised only inindividual cases.5

Here abiding by the law is set aside in favor of preserving civil peace forthe very reason of preserving it, because civil peace is the sole prerequis-ite for a state of law in which laws can be issued and enforced.

This problem, to which retributivism is exposed within the Kantianphilosophy of right, is actually related, thirdly, to a fundamentalproblem: retributivism is not compatible with an important principleof the Kantian philosophy of right, that is, with the permissive law(see Sections 2.4 and 4.4). The permissive law allows an exceptionto be made to law’s implementation, so long as this exception onlyconcerns what has occurred up to then and not that which either isoccurring in the present or will occur in the future. Such an exceptionshould provide for the peaceful establishment of a state of law wherethe application of the law to the existing situation would lead toregression to the state of nature or into civil war. However, the imple-mentation of the law in the present and in the future is not allowed tobe postponed ad calendas graecas,6 but instead should occur in themost direct way possible.7 Now, Kant defines right as being the “sumof the conditions under which the choice of one can be united withthe choice of another in accordance with a universal law of freedom,”8

by which the power of choice (Willkur) of all human beings – there-fore, also of the criminal – is meant. Thus, if there is in fact a wayenabling one to reintegrate the criminal into society in the most direct

5. RL Ak vi:334. Practical philosophy, ed. Gregor, p. 475.6. ZeF Ak viii:344f. Practical philosophy, ed. Gregor, p. 321.7. Cf. RL Ak vi:247. Practical philosophy, ed. Gregor, pp. 401–2. Also, cf. Reinhard Brandt,

“Das Erlaubnisgesetz, oder: Vernunft und Geschichte in Kants Rechtslehre,” in Brandt(ed.), Rechtsphilosophie der Aufklarung (Berlin: De Gruyter, 1982), pp. 233–85.


conclusion 193

way, then this way is commanded. If this way actually exists at all, thenit is, per definitionem, rehabilitative punishment.

The categorical imperative of right consists not only in an ideallegal system, but also in the steps that are necessary for either theestablishment or the reestablishment of this legal system. Writingssuch as Toward perpetual peace and Idea for a universal history with acosmopolitan purpose belong to the latter dimension of the categoricalimperative of right. Even though } 49e devoted to penal law is aneditorial section of the Doctrine of right, which has to do primarily withthe ideal legal system, and therefore with the first dimension, penallaw can be categorized into the second dimension.

In this respect, Gustav Radbruch’s dichotomy between right as arealm of justice in compensation (commutative justice) and an extra-legal realm of distributive justice and social utility is to be resolutelyrejected.9 Penal law, which incontrovertibly constitutes an essentialcomponent of right, pertains neither to commutative justice nor tocorrective justice. The former presupposes the agreement of allparties over the conditions and the items of the trade. Be it only forthis reason, punishment cannot rest on commutative justice. Not evenretributivism can be described as a trade. “Eye for eye, tooth fortooth”10 or “the retribution of what one merits” does not meantrading eyes or teeth, or trading the deserts of a human being forthe infliction of evils by another human being. Now, the return to thestatus quo ante pertains to corrective justice. However, how can the statusquo ante of a murder victim be reestablished even to a small extent?In this respect, Cesare Beccaria’s rhetorical question is irrefutable:“Can the wailings of a wretch, perhaps, undo what has been doneand turn back the clock?”11 The punishment, rather, has to do withdistributive justice, because it assures every human being of integrationor of reintegration into the commonwealth – regardless of his or herachievements or actions up to then – for the simple reason that everyhuman being as a being capable of reason has an innate, unforfeitableright to such.

9. Cf. Gustav Radbruch, Rechtsphilosophie, ed. Erik Wolf, seventh edition (Stuttgart:K. F. Koehler, 1970), p. 265.

10. Lev. 24:20 (KJV).11. Chapter 12, entitled “The purpose of punishment,” in Cesare Beccaria, On crimes and

punishments, in Beccaria, On crimes and punishments and other writings, ed. RichardBellamy, trans. Richard Davies (Cambridge: Cambridge University Press, 1995), p. 31.

194 conclusion

Fourthly, the categorical legal imperative of integration or reinte-gration into the commonwealth likewise derives a deeper groundingin the Kantian foundation of morality, namely in the empirical cir-cumstances of the exercise of autonomy in a limited being capable ofreason who can also act against his or her duty, irrationally. On theone hand, a human being should bear the responsibility for his or heractions because he or she has freedom of action. On the other hand,the consequences of actions carried out in the past should not lead tothe consequence that a person can no longer act freely in the future.These two dimensions do not limit one another. The consequences ofthe actions hitherto done are limited by the commandment thatrequires treating every human being as a being capable of reason withan inalienable freedom of action. Now, this inalienable freedom ofaction is not limited by the consequences of past actions. Rather, it islimited by the same inalienable freedom of action of the other humanbeings. It is only for this reason that the Kantian concept of right is anobject of the categorical imperative, which we may also term the“categorical imperative of right,” to borrow Otfried Hoffe’sexpression.12

We therefore find a twofold hierarchization of moral goods thatcan collide with one another. First, the commonwealth would receiveprecedence over individual freedom in the case of a collision, whichis why it is allowed provisionally either to limit or to suspend thefreedom of the criminal. Secondly, in the case of a collision, individualfreedom receives precedence over the consequences, which theactions of individuals should otherwise imply; this is why crime shouldbe forgiven by society – with consideration for the first priority, that is,with consideration for the security of the commonwealth – in the mostdirect manner possible.13

At the center of the Kantian idea of autonomy, under the guidanceof which we always ought to act and to which, therefore, the legalsystem is subject, is the forward-looking project of the full

12. Cf. chapter 5 in Otfried Hoffe, Categorical principles of law: a counterpoint to modernity(University Park: Pennsylvania State University Press, 2002).

13. This very arrangement of priority holds equally, by the way, for other realms in theKantian doctrine of right, for example property: cf. Jean-Christophe Merle (ed.),“Eigentumsrecht,” in Merle (ed.), Johann Gottlieb Fichte: Grundlage des Naturrechts(Berlin: De Gruyter, 2001), pp. 159–72; and Section 2.3 in Jean-Christophe Merle,Justice et progres: contribution a une doctrine du droit economique et social (Paris: PressesUniversitaires de France, 1997).

conclusion 195

development of the human being as a being capable of reason, andnot the painfully exact accounting of the misdeeds of humanity in theepoch of the – as Kant terms it – “baby walker.”14 As Albert Camusobserved in the debate over the death penalty, one could equallysay of retributivists that: “we all know today’s executioners are human-ists.”15 It is just that we do not hold the same concept of humanityas they do.

14. The epoch of the Gangelwagen, a device used by children learning to walk. Wasist Aufklarung? Ak viii:35. Practical philosophy, ed. Gregor, p. 17: “walking wagon.”Also found in KrV B 174. Critique of pure reason, ed. Guyer and Wood, p. 269:“leading-strings.”

15. Albert Camus, Reflections on the guillotine: an essay on capital punishment by the 1957Nobel Prize winner, trans. Richard Howard (Michigan City, Ind.: Fridtjof-KarlaPublications, 1959), p. 50.

196 conclusion

BIBLIOGRAPHY

Primary texts

Beccaria, Cesare. On crimes and punishments, in Beccaria On crimes and punishmentsand other writings, ed. Richard Bellamy, trans. Richard Davies (Cambridge:Cambridge University Press, 1995), pp. 1–113.

Bentham, Jeremy. An introduction to the principles of morals and legislation, reprintof 1823 edition (Oxford: Oxford University Press, 1907).

Camus, Albert. Reflections on the guillotine: an essay on capital punishment by the1957 Nobel Prize winner, trans. Richard Howard (Michigan City, Ind.:Fridtjof-Karla Publications, 1959).

Cicero, Marcus Tullius. On duties, ed. M. T. Griffin and E. M. Atkins (Cambridge:Cambridge University Press, 1991).

Duhring, Eugen. Cursus der Philosophie als streng wissenschaftlicher Weltanschauungund Lebensgestaltung (Leipzig: O. R. Reisland, 1894–5).

Fichte, Johann Gottlieb. Foundations of natural right, ed. Frederick Neuhouser,trans. Michael Baur (Cambridge: Cambridge University Press, 2000).

Grotius, Hugo. The rights of war and peace, ed. Richard Tuck (3 vols., Indianapolis:Liberty Fund, Inc., 2005).

Hegel, G. W. F. Elements of the philosophy of right, trans. H. B. Nisbet (Cambridge:Cambridge University Press, 1991).

Naturrecht und Staatswissenschaft (1818/19), in Hegel, and Vorlesungen uberRechtsphilosophie, ed. Reinhard Lauth (4 vols., Stuttgart and Bad-Cannstatt:Frommann-Holzboog, 1973–4), vol. 1, pp. 217–351.

Philosophie des Rechts (1824/25), in Hegel, Vorlesungen uber Rechtsphilosophie,ed. Reinhard Lauth (4 vols., Stuttgart and Bad-Cannstatt: Frommann-Holzboog, 1973–4), vol. 4, pp. 67–752.

Principes de la philosophie du droit, trans. Jean-Francois Kervegan (Paris: PressesUniversitaires de France, 1998).

Hobbes, Thomas. Leviathan, ed. A. R. Walter (Cambridge: Cambridge UniversityPress, 1904).

Jhering, Rudolf von. Law as a means to an end, ed. Joseph H. Drake, trans. IsaacHusik (Union, NJ: Lawbrook Exchange, 1999).

197

Kant, Immanuel. Kants gesammelte Schriften (Berlin: Koniglich PreußischeAkademie der Wissenschaften, 1902).

Lectures on ethics, ed. Peter Heath and J. B. Schneewind, trans. Peter Heath(Cambridge: Cambridge University Press, 1997).

Klein, Ernst Ferdinand. Grundlinien des gemeinen deutschen und preußischenpeinlichen Rechts (Halle: Hemmerde und Schwetschke, 1796).

Krause, Karl Christian Friedrich. Abhandlung uber die Idee und die Eintheilung derPhilosophie und der Mathematik und den innigen Zusammenhang beider, in PaulHohlfield and August Wunsche (eds.), Philosophische Abhandlungen: aus demhandschriftlichen Nachlasse (Leipzig: Schulze, 1889), pp. 5–40.

Grundlage des Naturrechts oder philosophischer Grundriß des Ideals des Rechts,(Jena: Gabler, 1803).

Mably, Gabriel de. “De la legislation,” inœuvres completes (12 vols., London, 1789).Mill, John Stuart. “April 1868 speech on capital punishment,” in Mill,

Utilitarianism, ed. George Sher (Indianapolis: Hackett, 2001), pp. 65–70.Nietzsche, Friedrich. On the genealogy of morality, ed. Keith Ansell-Pearson, trans.

Carol Diethe (Cambridge: Cambridge University Press, 2007).Plato. Laws, trans. Trevor J. Saunders, in John M. Cooper and D. S. Hutchinson

(eds.), Plato: Complete works (Indianapolis: Hackett, 1997), pp. 1318–1616.Pufendorf, Samuel von. De jure naturae et gentium, ed. James Brown Scott, trans.

C. H. and W. A. Oldfather, Classics of International Law (2 vols., Oxford:Clarendon Press / London: Humphrey Milford, 1934).

Schopenhauer, Arthur. The world as will and idea, trans. R. B. Haldane andJ. Kemp, seventh edition (4 vols., London: Kegan Paul, Trench, Trubnerand Co., 1909?).

Schroeder, Friedrich-Christian (ed.). Die Carolina: die Peinliche GerichtsordnungKaiser Karls V. von 1532 (Darmstadt: Wissenschaftliche Buchgesellschaft,1986).

Seneca.De ira, inMoral essays, trans. JohnW.Basore (3 vols., London:Heinemann/Cambridge, Mass.: Harvard University Press, 1928), vol. 1, pp. 106–355.

Sidgwick, Henry. The elements of politics, reprint of 1891 edition (New York:Cosimo Classics, 2005).

Voltaire. “Commentary on the book On crimes and punishments, by a provinciallawyer,” in Voltaire, Political writings, ed. and trans. David Williams(Cambridge: Cambridge University Press, 1994), pp. 244–79.

Secondary texts

Altman, Andrew and Wellman, Christopher Heath. “A defense of internationalcriminal law,” Ethics, 115, no. 1 (2004), 35–67.

Ataner, Attila. “Kant on capital punishment and suicide,” Kant-Studien, 97, no. 4(2006), 452–82.

Badinter, Robert. L’Abolition (Paris: Fayard, 2000).Bedau, Hugo. “Retribution and the theory of punishment,” Journal of Philosophy,

75, no. 11 (1978), 601–20.Benn, Stanley. “An approach to the problems of punishment,” Philosophy, 33

(1958), 321–41.

198 bibliography

Bianchi, Hermann. “Abolition: assensus and sanctuary,” in Alexander R. Duffand David Garland (eds.), A reader on punishment (Oxford: OxfordUniversity Press, 1994), pp. 336–51.

Bosanquet, Bernard. The philosophical theory of the state (London and New York:Macmillan, 1899).

Braithwaite, John. Crime, shame and reintegration (Cambridge: CambridgeUniversity Press, 1989).

Braithwaite, John and Pettit, Philip. Not just deserts: a republican theory of criminaljustice (Oxford: Clarendon Press, 1990).

Brandt, Reinhard. “Das Erlaubnisgesetz, oder: Vernunft und Geschichte inKants Rechtslehre,” in R. Brandt (ed.), Rechtsphilosophie der Aufklarung:Symposium Wolfenbuttel 1981 (Berlin: De Gruyter, 1982), pp. 233–85.

Breazeale, Daniel. Philosophy and truth: selections from Nietzsche’s Notebooks of theearly 1870s (Atlantic Highlands, NJ: Humanities Press, 1979).

Brugger, Winfried. “Darf der Staat ausnahmsweise foltern?” Der Staat, 35 (1996),67–97.

“Vom unbedingten Verbot der Folter zum bedingten Recht auf Folter?”Juristenzeitung, 35, no. 4 (18 February 2000), 165–73.

Byrd, Sharon. “Kant’s theory of punishment: deterrence in its threat, retributionin its execution,” Law and Philosophy, 8, no. 2 (1989), 151–200.

Clark, Michael. “A non-retributive approach to punishment,” Ratio, 17, no. 1(2004), 12–27.

Cohen, Hermann. Schriften zur Philosophie und Zeitgeschichte, ed. Albert Gorlandand Ernst Cassirer (Berlin: Akademie Verlag, 1928).

Deith, John. “On the right to be punished: some doubts,” Ethics, 94 (1984),191–211.

Feinberg, Joel. Problems at the roots of law: essays in legal and political theory (Oxford:Oxford University Press, 2003).

Fichtelberg, Aaron. “Crimes beyond justice? Retributivism and war crimes,”Criminal Justice Ethics, 24, no. 1 (2005), 31–46.

Flechtheim, Ossip Kurt. “Die Funktion der Strafe in der Rechtstheorie Hegels,”in Flechtheim, Von Hegel zu Kelsen: rechtstheoretische Aufsatze (Berlin: Dunckerund Humblot, 1963), pp. 9–20.

“Zur Kritik der Hegelschen Straftrechtstheorie,” Archiv fur Rechts- undSozialphilosophie, 54 (1968), 539–48.

Hegels Straftheorie (Berlin: Duncker und Humblot, 1975).Fletcher, George P. Basic concepts of criminal law (Oxford: Oxford University

Press, 1998).Forster, Wolfgang. Karl Christian Friedrich Krauses fruhe Rechtsphilosophie und ihr

geistesgeschichtlicher Hintergrund (Ebelsbach: Aktiv Druck und Verlag, 2000).Foucault, Michel. Discipline and punish: the birth of the prison, trans. Alan Sheridan,

second edition (New York: Random House, 1995).Garapon, Antoine. Des crimes qu’on ne peut ni punir ni pardonner (Paris: Editions

Odile Jacob, 2002).Guillarme, Bertrand. Penser la peine (Paris: Presses Universitaires de Paris, 2003).Gutmann, Amy and Thompson, Dennis. Why deliberative democracy? (Princeton:

Princeton University Press, 2004).

bibliography 199

Hampton, Jean and Murphy, Jeffrie. Forgiveness and mercy (Cambridge:Cambridge University Press, 1988).

Hart, H. L. A. Punishment and responsibility (Oxford: Clarendon Press, 1968).Hassemer, Winfried. Einfuhrung in die Grundlagen des Strafrechts, second edition

(Munich: C. H. Beck, 1990).Hill, Thomas E. “Kant on punishment: a coherent mix of deterrence and

retribution,” Annual Review of Law and Ethics, 5 (1997), 291–314.“Kant on wrongdoing, desert, and punishment,” Law and Philosophy, 18 (1999),407–41.

Human welfare and moral worth: Kantian perspectives (Oxford: Clarendon Press,2002).

Hoffe, Otfried. Gibt es ein interkulturelles Strafrecht? Ein philosophischer Versuch(Frankfurt a.M.: Suhrkamp, 1999).

(ed.). Immanuel Kant: Metaphysische Anfangsgrunde der Rechtslehre (Berlin:De Gruyter, 1999).

“Vom Straf- und Begnadigungsrecht,” in O. Hoffe (ed.), Immanuel Kant:metaphysische Anfangsgrunde der Rechtslehre (Berlin: De Gruyter, 1999),pp. 213–33.

“Konigliche Volker”: zu Kants kosmopolitischer Rechts- und Friedenstheorie (Frankfurta.M.: Suhrkamp, 2001).

Categorical principles of law: a counterpoint to modernity (University Park:Pennsylvania State University Press, 2002).

Gerechtigkeit: eine philosophische Einfuhrung (Munich: C. H. Beck, 2004).Holtman, Sarah. “Toward social reform: Kant’s penal theory reinterpreted,”

Utilitas, 9 (1997), 3–21.Honderich, Ted. Punishment: the supposed justifications (New York: Harcourt,

Brace and World, 1969).Honneth, Axel. The struggle for recognition: the moral grammar of social conflicts,

trans. Joel Anderson (Cambridge, Mass.: MIT Press, 1996).Hosle, Vittorio. “Was darf und was soll der Staat bestrafen?”, in Hosle,

Rechtsphilosophie des deutschen Idealismus (Hamburg: F. Meiner, 1989).Kaufmann, Matthias. Rechtsphilosophie (Freiburg i.Br.: K. Alber, 1996).

“Zwangsrecht (}} 13–16),” in Jean-Christophe Merle (ed.), Johann Gottlieb Fichte :Grundlage des Naturrechts (Berlin: Akademie Verlag, 2001), pp. 125–37.

Kersting, Wolfgang. “Sittengesetz und Rechtsgesetz: die Begrundung des Rechtsbei Kant und den fruhen Kantianern,” in Reinhard Brandt (ed.),Rechtsphilosophie der Aufklarung: Symposium Wolfenbuttel 1981 (Berlin: DeGruyter, 1982), pp. 147–77.

Wohlgeordnete Freiheit: Immanuel Kants Rechts- und Staatsphilosophie, secondedition (Berlin: De Gruyter, 1993).

Kleinig, John. “Punishment and moral seriousness,” Israel Law Review, 25, no. 3(1991), 401–21.

Klug, Ulrich. “Abschied von Kant undHegel,” in Jurgen Baumann (ed.), Programmfur ein neues Strafgesetzbuch: der Alternativ-Entwurf der Strafrechtslehrer (Frankfurta.M.: Fischer, 1968), pp. 36–41.

Koller, Peter. “Probleme der utilitaristischen Strafrechtfertigung,” Zeitschrift furdie Gesamte Strafrechtswissenschaft, 91 (1979), 45–95.

200 bibliography

Kostlin, Christian R. Neue Revision der Grundbegriffe des Kriminalrechts, reprint(Aalen: Scientia, 1970).

Kuhl, Kristian. Die Bedeutung der Rechtsphilosophie fur das Strafrecht (Baden-Baden:Nomos, 2001).

Landau, Peter. “Karl Christian Friedrich Krauses Rechtsphilosophie,” in Klaus-Michael Kodalle (ed.), Karl Christian Friedrich Krause (1781–1832):Studien zu seiner Philosophie und zum Krausismo (Hamburg: Meiner, 1985),pp. 80–92.

Lazzari, Alessandro. “Eine Fessel, die nicht schmerzt und nicht sehr hindert:Strafrecht,” in Jean-Christophe Merle (ed.), Johann Gottlieb Fichte: Grundlagedes Naturrechts (Berlin: Akademie Verlag, 2001), pp. 173–86.

Lesch, Heiko Hartmut. Der Verbrechensbegriff: Grundlinien einer funktionalenRevision (Cologne: Heymanns, 1999).

Liszt, Franz von. Der Zweckgedanke im Strafrecht (Berlin, 1883).Ludwig, Bernd. Kants Rechtslehre (Hamburg: F. Meiner, 1988).May, Larry. Crimes against humanity: a normative account (Cambridge: Cambridge

University Press, 2006).McTaggart, John E. Studies in Hegelian cosmology (Cambridge: Cambridge

University Press, 1918).Merle, Jean-Christophe. “Il punto di vista educativo e religioso dei Contributi

destinati a rettificare il giudizio del pubblico sulla Rivoluzione francese:la dimensione politica del Saggio di una critica di ogni rivelazione,” in AldoMasullo and Marco Ivaldo (eds.), Filosofia trascendentale e destinazione etica(Milan: Guerini, 1995), pp. 303–25.

Justice et progres: contribution a une doctrine du droit economique et social (Paris:Presses Universitaires de France, 1997).

(ed.). “Eigentumsrecht,” in Merle (ed.), Johann Gottlieb Fichte: Grundlage desNaturrechts (Berlin: Akademie Verlag, 2001), pp. 159–72.

Mohr, Georg. “Unrecht und Strafe,” in Ludwig Siep (ed.), G. W. F. Hegel:Grundlinien der Philosophie des Rechts (Berlin: De Gruyter, 1997), pp. 95–124.

Murphy, Jeffrie G. “Does Kant have a theory of punishment?” Columbia LawReview, 87 (1987), 509–32.

Neumann, Ulfried and Schroth, Ulrich. Neuere Theorien von Kriminalitat undStrafe (Darmstadt: Wissenschaftliche Buchgesellschaft, 1980).

Ottmann, Henning. Philosophie und Politik bei Nietzsche (Berlin: De Gruyter,1987).

Pelczynski, Zbigniew A. (ed.). Hegel’s political philosophy (Cambridge: CambridgeUniversity Press, 1971).

Piontkowski, Andrei A. Hegels Lehre uber Staat und Recht und seine Strafrechtstheorie,trans. Anna Neuland (Berlin: De Gruyter, 1960).

Pogge, Thomas W. “Is Kant’s Rechtslehre comprehensive?” Southern Journal ofPhilosophy, 36, supplement (1997), 161–87.

“Is Kant’s Rechtslehre a ‘comprehensive liberalism’?” in Mark Timmons (ed.),Kant’s Metaphysics of morals (Oxford: Oxford University Press, 2002),pp. 133–58.

Primorac, Igor. “Punishment as the criminal right,” Hegel-Studien, 15 (1980),187–98.

bibliography 201

Primoratz, Igor. “Banquos Geist: Hegels Theorie der Strafe,” Hegel-Studien, 29,supplement (1986).

Pugsley, Robert A. “A retributivist argument against capital punishment,”HofstraLaw Review, 9 (1981), 1501–23.

Quinton, Anthony M. “On punishment,” in Harry Burrows Acton (ed.),The philosophy of punishment (London: Macmillan, 1969), pp. 55–64.

Radbruch, Gustav. Rechtsphilosophie, ed. Erik Wolf, seventh edition (Stuttgart:K. F. Koehler, 1970).

Ricœur, Paul. Le Juste (Paris: Editions Esprit, 1995).Ripstein, Arthur. Equality, responsibility and the law (Cambridge: Cambridge

University Press, 1999).Rosen, Fred. “Utilitarianism and the punishment of the innocent: the origins

of a false doctrine,” Utilitas, 9, no. 1 (March 1997), 23–37.Roxin, Claus. Strafrechtliche Grundlagenprobleme (Berlin: De Gruyter, 1973).

“‘Schuld’ und ‘Verantwortlichkeit’ als strafrechtliche Systemkategorien,” inRoxin(ed.), Grundfragen der gesamten Strafrechtswissenschaft (Berlin: DeGruyter, 1974), pp. 171–97.

Scala,Klaus. “HegelsBegriffderStrafeunddiemoderneStrafvollzugsproblematik,”Hegel-Jahrbuch (1987), 164–70.

Scharf, Michael P. “Justice vs. peace,” in Sarah B. Sewall and Carl Kaysen (eds.),The United States and the International Criminal Court: national security andinternational law (Lanham, Md.: Rowman and Littlefield, 2000),pp. 213–36.

Scheid, Don E. “Kant’s retributivism,” Ethics, 93 (1983), 262–82.Schild, Wolfgang. “Ende und Zukunft des Staatsrechts,” Archiv fur Rechts- und

Sozialphilosophie, 70 (1984), 71–112.Schwarzschild, Steven S. “Kantianism and the death penalty,” Archiv fur

Rechts- und Sozialphilosophie, 70 (1985), 343–77.Seelmann, Kurt. “Versuch einer Legitimation von Strafe durch das Argument

selbstwiderspruchlichen Verhaltens des Straftaters,” Jahrbuch fur Recht undEthik, 1 (1993), 315–26.

“Wechselseitige Anerkennung und Unrecht: Strafe als Postulat derGerechtigkeit?” Archiv fur Rechts- und Sozialphilosophie, 79, no. 2 (1993),228–36.

Shue, Henry. “Torture,” Philosophy and Public Affairs, 7, no. 2 (1977–88),124–43.

Sorell, Tom. Moral theory and capital punishment (Oxford: Blackwell, 1987).Stillman, Peter G. “Hegel’s idea of punishment,” Journal of the History of

Philosophy, 14 (1976), 169–82.Tunick, Mark. Hegel’s political philosophy: interpreting the practice of legal punishment

(Princeton: Princeton University Press, 1992).van den Haag, Ernest. “Why capital punishment?” Albany Law Review, 54, nos.

3–4 (1990), 501–14.Willascheck, Marcus. “Why the Doctrine of Right does not belong in the

Metaphysics of Morals,” Annual Review of Law and Ethics, 5 (1997), 205–27.Wolf, Jean-Claude. Verhutung oder Vergeltung? Einfuhrung in ethische Straftheorien

(Freiburg i.Br.: Alber, 1992).

202 bibliography

“Strafe als Widerherstellung eines Gleichgewichts,” Jahrbuch fur Recht undEthik, 11 (2003), 199–216.

Wood, Allen W. Hegel’s ethical thought (Cambridge: Cambridge University Press,1990).

“Kant’s Doctrine of right,” Introduction to Otfried Hoffe (ed.), ImmanuelKant: Metaphysische Anfangsgrunde der Rechtslehre (Berlin: Akademie Verlag,1999), pp. 19–39.

Zaczyk, Rainer. Das Strafrecht in der Rechtslehre J. G. Fichtes (Berlin: Duncker undHumblot, 1981).

bibliography 203

INDEX

abolitionism, 20; see also impunityacquittal of a debt, see debtadiaphoron, moral, 27agreement, 21Altman, Andrew, 186Arendt, Hannah, 183Aristotle, 21, 100Ataner, Attila, 59attempt, 92, 190autonomy of the will, see freedom

of action

Badinter, Robert, 182banishment, 178, 179; see also exclusion

from societyBeccaria, Cesare, 2, 18, 20, 70, 90–1,

164, 194Benn, Stanley, 108Bentham, Jeremy, 18–19, 20, 164bestiality, 42Bianchi, Herman, 1, 158Bosanquet, Bernard, 107Braithwaite, John, 5, 7, 13Brandt, Reinhard, 59, 103, 193Brugger, Winfried, 2burden of proof, 69Byrd, Sharon B., 44, 45–9

Camus, Albert, 196cancellation (Aufhebung), 115, 124, 126,

129, 134–5, 136–8, 160categorical imperative, 25–43, 58–9, 72,

90, 119, 130, 194, 195child, 64–5, 66, 67Christianity, 7Cicero, Marcus Tullius, 100Clark, Michael, 102coercion, 30–1, 33, 38, 39–41, 42, 46,

67, 74, 91–4, 125, 128–45Cohen, Hermann, 59–60

Commission for Truth andReconciliation, 183

common sense, 113, 122–4, 143, 149,184–5

compensation, 21, 54–5, 73, 92, 127,136–7, 151, 152–3, 155, 156–8,161, 192

conscience, bad, see remorseconsciousness, common, see common

senseconsent, 113, 114consequences, internal to the act, 138–9,

143–4contract, 29–31, 54, 152contribution, 57conversion, 79–80Cooper, David, 129corporal punishment, see torturecrime against humanity, 149cruelty, 153–4, 160, 167, 168

death penalty, 1, 5, 20, 31, 47–8, 49–50,53, 54–5, 59–60, 60–1, 68, 70, 94, 95,106, 117, 133–4, 157, 169, 178–9,182, 184, 185, 193, 196

debt, 21, 54, 74, 150, 152–3, 156, 157,159–61, 167, 168

Deith, John, 80desert, see meritdeterrence: general, 2–3, 8, 18, 19,

57–71, 94, 97–8, 101–2, 104, 107, 110,119, 135, 138, 140–1, 177–8, 180–1,181–2, 191; specific, 4

Duhring, Karl Eugen, 150, 161–4duty: of right, 32–3; of virtue,

Ebbinghaus, Julius, 26enslavement, see slaveryErdemovic, Drazen, 180evil, see wickedness

204

example, deterrence through, 93, 95,104, 138

exclusion from society, 94, 96, 99, 100,102–3, 105, 153–4, 167;see also banishment

expiation contract, 94, 99, 102expressivism, 72, 175

Feinberg, Joel, 191Feuerbach, Anselm, 2, 142Fichte, Johann Gottlieb, 14, 42, 83,

85–106, 116–18, 134, 144, 153,178, 188

Fichtelberg, Aaron, 178Flechtheim, Ossip K., 108, 111–13Fletcher, George P., 167force, 132–4forgiveness, 78, 81Forster, Wolfgang, 105, 106Foucault, Michel, 1, 156freedom of action, 195freedom of the will, 35, 62–4, 65

Garapon, Antoine, 173goal of punishment, 72–3, 162, 164,

165, 189, 191Grotius, Hugo, 19, 21, 191Grundgesetz (German Basic Law), 176Guillarme, Bertrand, 111guilt, 150, 175, 183, 191Gutmann, Amy, 185

Hampton, Jean, 72, 127happiness, 55, 75–6, 77–8, 80harm principle, 37Hart, H. L. A., 70, 73, 82Hassemer, Winfried, 4, 5, 10, 107Hegel, Georg Wilhelm Friedrich, 10,

13, 14, 42, 83, 85, 105, 107–45,149, 155, 164, 188

highest good, 55, 75–6, 90Hill, Thomas E., 44, 55, 72, 80, 149Hitler, Adolf, 182, 184Hobbes, Thomas, 2, 18, 19, 151–2Hoffe, Otfried, 4, 6, 25, 34, 44, 51, 56,

107, 108, 195Hohfeld, Wesley Newcomb, 92Holtman, Sarah, 44Honderich, Ted, 108Honneth, Axel, 126Hosle, Vittorio, 42Hufeland, Gottlieb, 26humanity, crime against, see crime

against humanityhumanity as an end in itself, 32–3, 40,

46, 61, 64, 93, 119humiliation, 157

impunity, 158–60; see also abolitionismimputability, 62–4incapacitation, 3, 19, 71, 80–1, 94, 104, 176incorrigible prisoners, 94–5innocents, punishment of the, 19, 22, 48, 74insight of the criminal, 114, 145internalization, 147, 151, 168International Criminal Court (ICC),

171–5, 179International Criminal Tribunal for

Former Yugoslavia (ICTFY), 180intervention, military humanitarian, 177ius strictum, 137

Jhering, Rudolf von, 164Joseph II, Holy Roman Emperor, 133justice: commutative, 21, 151, 159–62,

164–5, 194; corrective, 135;distributive, 21, 174; Hobbesian, 152;in general, 169; natural law, 169

Kant, Immanuel, 10, 13–14, 21–3, 24–43,44–71, 72–83, 90, 92, 101, 118–20,136, 149, 181, 187–8, 191, 192–6

Kaufmann, Matthias, 88, 108Kersting, Wolfgang, 25, 26, 28, 39–41Kervegan, Jean-Francois, 108Klein, Ernst Ferdinand, 132, 134, 135, 142Kleinig, John, 127Klug, Ulrich, 107Koller, Peter, 3Kostlin, Christian Reinhard, 107Krause, Karl Christian Friedrich, 105–6Kuhl, Kristian, 3, 10, 30

Landau, Peter, 3, 106Lazzari, Alessandro, 88Lesch, Heiko H., 110liberalism, 25, 34–43, 106Liszt, Franz von, 164Locke, John, 34Ludwig, Bernd, 28, 29lynching, 168

Mably, Gabriel Bonnot de, 156McTaggart, John Ellis, 107marriage, 42maxim, 28, 33, 56, 75May, Larry, 172, 175, 183, 184mercy, 160–1, 167merit, 74, 91Merle, Jean-Christophe, 43, 87, 90, 195Mill, John Stuart, 20, 34, 37Milosevic, Slobodan, 186Mitnick, Kevin, 69mixed theories, 5, 12, 44–53, 88–9, 104,

107–20, 169

index 205

mnemotechnic, 153, 169Mohr, Georg, 108–9, 118–19monstrous crimes, 175, 183morality, 23–4, 25–34, 38, 42, 44, 45, 142–3Moses, 19Murphy, Jeffrie G., 72

natural law, 18, 169, 176, 191necessity, see right of necessitynegation: of the crime, 107–45; of the

negation, 112, 122Neumann, Ulfried, 6, 9Nietzsche, Friedrich, 14, 147–70, 188nulla poena sine lege, 171nullity of crime, 115, 116, 117–18,

129, 137Nuremberg, tribunal of, 171, 179, 180

Ottmann, Henning, 151

pardon, 31, 68, 159, 167, 179, 181parole, release on, 69Paul, Saint, 80peace, civil, 163, 193; see also war, civilpermissive law, 103, 193personality, 61–4Pettit, Philip, 13Pfaelzer, Marianna, 69Piontkowski, Andrei A., 107Plato, 17, 38, 73, 116, 164Pogge, Thomas W., 34, 62, 65, 66positivism, legal, 38Primorac (Primoratz), Igor, 108, 111,

115, 122promise, 29–31, 152, 161property right, 41, 60, 65proportionality principle, 5, 19–20, 21, 43,

45, 49, 70, 74, 82, 113, 120–2, 124,127, 159, 168, 174, 183, 184, 189

Pufendorf, Samuel von, 2, 19, 21, 191Pugsley, Robert A., 59

Quinton, Anthony M., 108

Radbruch, Gustav, 194recidivism, 9, 10, 13, 57, 68, 69, 95, 147,

165, 181recognition, 99–100, 116–18, 131reconciliation, 113reform, see rehabilitationrehabilitation, 3, 5, 7, 9, 13, 57–8, 61–71,

81, 85–106, 108, 117, 135–6, 145,147, 167, 189, 194

remorse, 14, 147, 149, 162, 164, 166republic, 25resentment, see ressentimentresistance, 176

responsibility, 6–7, 13ressentiment, 154, 161restoration of right, 112, 113, 127–8, 194retaliation, 6, 72, 90, 112, 115, 117, 120–1,

125–7, 141, 143–4retribution, 6, 8, 49, 50–7, 74, 92, 127,

141, 161retributivism, 2, 3–4, 6, 7, 21–4, 43, 51–7,

72, 81–2, 110, 119, 141, 143, 147,181, 183, 192

revenge, 56, 78, 81–2, 114, 125–6, 135,137 , 161– 2, 163; see also vengeance

Ric�ur, Paul, 5right, concept of, 85–7, 89, 90, 97, 99,

101, 116, 144–5right of necessity, 46, 52, 94, 193Ripstein, Arthur, 80, 108, 139Rosen, Fred, 3, 74Roxin, Claus, 7, 9, 11, 12, 13, 107, 147–8

savage, 67Scala, Klaus, 108Schadenfreude, 149Scharf, Michael P., 179Scheid, Don E., 44, 51Schild, Wolfgang, 108Schmalz, Theodor, 26Schopenhauer, Arthur, 2, 107Schroth, Ulrich, 6, 9Schwarzschild, Steven S., 59security, public, 95, 140, 157–9, 190,

191, 195Seelmann, Kurt, 108, 111, 116–17, 141self-defense, right of, 91, 137, 191Seneca, 17, 73Shue, Henry, 2Sidgwick, Henry, 17, 20, 23slavery, 56, 60–1, 66, 95social contract, 85, 94, 102Sorell, Tom, 60sovereign, 91state of nature, 52, 53, 57, 58–9, 101,

154, 181, 193statute of limitation, 174–5, 184–6, 189Stillman, Peter G., 108suffering, see torturesuicide, 37, 59, 61

talion law, 17, 19–20, 21–3, 51–2,56–7, 87, 95–6, 101, 113, 121–2,127, 142

teleology, 39–41, 65Thompson, Dennis, 185threat of punishment, deterrence

through, 2, 46–8, 70, 93–4, 98,104–5, 119

Tokyo, tribunal of, 171, 179

206 index

torture, 1, 5, 20, 48, 96, 106, 128,152–3, 154, 157, 163, 166, 168–9,184, 185, 186

Tunick, Mark, 108tyrannicide, 176

universalization, 56usefulness, 99, 100–1utilitarianism, 18, 44, 51, 68, 100

van den Haag, Ernest, 2Vattel, Emer de, 191vengeance, 76–7victim, 9, 81, 92, 127, 130, 137, 155,

157, 159–60, 162, 163, 171, 173,186, 190, 194

virtue, 74–6, 77Voltaire, 170

war: civil, 156, 159, 193; international,191 –2; see also peace, civil

Wellman, Christopher H., 186wickedness, inner, 43, 55, 73–4, 75,

76, 77–8, 79, 80, 101, 102, 122,176, 184, 190–1

Willascheck, Marcus, 34Wolf, Jean-Claude, 55, 107Wolff, Christian, 2, 39Wood, Allen W., 36, 108–9, 114worthiness of punishment, 73, 81–2

Zaczyk, Rainer, 93

index 207

Documents

German Idealism and the Concept of Punishment Modern European Philosophy