German Idealism and The Concept of Punishment Modern European Philosophy

This page intentionally left blank
GERMAN IDEALISM AND THE

CONCEPT OF PUNISHMENT
Against the background of early modernism a period that justified
punishment by general deterrence Kant is usually thought to
represent a radical turn toward retributivism. For Kant, and later for
Fichte and Hegel, a just punishment respects the humanity inherent
in the criminal, and serves no external ends: it is instituted only because
the criminal deserves it. In this original study, Jean-Christophe Merle
uses close analysis of texts to show that these philosophers did not in
fact hold a retributivist position, or even a mixed position; instead
he traces in their work the gradual emergence of views in favor of
deterrence and resocialization. He also examines Nietzsches view that
morality rests on the rejection of retribution. His final chapter offers a
challenge to the retributivist position, and a defense of resocialization,
in the context of current legal theory and practice concerning the
punishment of crimes against humanity.
j e a n - ch r i s t o p h e m e r l e is a senior researcher in philosophy at the
University 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 Dame
ROLF-PETER HORSTMANN, Humboldt University, Berlin
Some recent titles

Daniel W. Conway: Nietzsches Dangerous Game
John P. McCormick: Carl Schmitts Critique of Liberalism
Frederick A. Olafson: Heidegger and the Ground of Ethics
Gunter Zoller: Fichtes Transcendental Philosophy
Warren Breckman: Marx, the Young Hegelians, and the
Origins of Radical Social Theory
William Blattner: Heideggers Temporal Idealism
Charles Griswold: Adam Smith and the Virtues of Enlightenment
Gary Gutting: Pragmatic Liberalism and the Critique of Modernity
Allen Wood: Kants Ethical Thought
Karl 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: Heideggers Concept of Truth
Michelle Grier: Kants Doctrine of Transcendental Illusion
Henry Allison: Kants Theory of Taste
Allen Speight: Hegel, Literature, and the Problem of Agency
J. M. Bernstein: Adorno
Will Dudley: Hegel, Nietzsche, and Philosophy
Taylor Carman: Heideggers Analytic
Douglas Moggach: The Philosophy and Politics of Bruno Bauer
Rudiger Bubner: The Innovations of Idealism
Jon Stewart: Kierkegaards Relations to Hegel Reconsidered
Michael Quante: Hegels Concept of Action
Wolfgang Detel: Foucault and Classical Antiquity
Robert M. Wallace: Hegels Philosophy of Reality, Freedom, and God

Johanna Oksala: Foucault on Freedom
Beatrice Longuenesse: Kant on the Human Standpoint
Wayne Martin: Theories of Judgment
Heinrich Meier: Leo Strauss and the Theologico-Political Problem
Otfried Hoffe: Kants Cosmopolitan Theory of the Law and Peace
Beatrice Longuenesse: Hegels 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,

So Paulo, Delhi, Dubai, Tokyo
Cambridge University Press
The Edinburgh Building, Cambridge CB2 8RU, UK
Published in the United States of America by Cambridge University Press, New York
www.cambridge.org
Information on this title: www.cambridge.org/9780521886840
Jean-Christophe Merle 2009
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.
First published in print format 2009
ISBN-13
978-0-521-88684-0
Hardback
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.
. . . we all know todays executioners are humanists.

Albert Camus
CONTENTS
page xi
Preface
xiii
List of abbreviations
Introduction
part i desert as the sole justification

for punishment
1 The two Kantian concepts of right
17
2 Kants legal justification of punishment
44
3 Kants moral justification of punishment
72
part ii punishment as a means of

rehabilitation
4 Fichtes expiation contract
87
5 Hegels negation of crime
107
part iii retributivist inhumanity

6 Nietzsche and punishment without remorse
149
7 What is the purpose of punishing crimes

against humanity?
171
Conclusion
187
Bibliography
197
Index
204
ix
PREFACE
Nemo prudens punit quia peccatum est sed ne peccetur, says Seneca
in De ira, and many philosophers who have come after him recommend
such a justification of punishment by deterrence. Since Immanuel
Kant, 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 should
not read: For what purpose punish? Rather, according to Kants
absolutist or categorical imperative regarding punishment, punishment can only be carried out because the malefactor is deserving of
the punishment. Everything else is allegedly unjust, and is detrimental
to the malefactors human dignity as a moral subject. Such a theory of
retributive justice, which draws not only from Kant but also from
G. W. F. Hegel, inspires a great deal of fascination in many philosophers, but that notwithstanding it still stands on shaky ground.
A precise analysis of Kants and Hegels philosophy of law and morality leads rather to a special form of deterrence theory.
I will attempt to conduct this analysis within the confines of this
book. The analysis begins with Kant, continues with J. G. Fichte and
Hegel, leads to Friedrich Nietzsche, and then concludes with a discussion of the justification of punishment for crimes against humanity.
This closing discussion should be seen as the touchstone. Should my
position be able to explain this difficult case, then it should be even
more able to explain cases of lesser difficulty.
I would like to thank Manfred Frank and Anton Schindling for
their comments on the manuscript as well as two anonymous referees
from Cambridge University Press. Special thanks go to Sharon Byrd,
Philippe Coppens, Roman Eisele, George Fletcher, Thomas Grundmann,
Jan C. Joerden, Matthias Kaufmann, John Kleinig, Eugenio Pacelli de
xi
xii
preface
Oliveira, Herve Pourtois, Alexandre Travessoni Gomes, Luiz Moreira

and Jean-Claude Wolf. I have also benefited from questions and
remarks from: European Network Applied Global Justice, KantGesellschaft, Fichte-Gesellschaft, CAPPE (ANU Canberra), Chaire
Hoover and Centre de Philosophie du Droit (Louvain), University of
Grenoble, PUCRS Porto Alegre, UFMG Belo Horizonte, UFSC Florianopolis, Ankara Bar Association, University of Fribourg, Graduiertenkolleg Globale Herausforderung (Tubingen), Charles Sturt
University at Wagga Wagga, and a conference of the European Network Global Justice in Otzenhausen, as well as from my seminars on
these topics in Aachen, Saarbrucken, Tours and Tubingen. Last, but
not least, I am very grateful to Hilary Gaskin of Cambridge University
Press, as well as to Frances Brown.
ABBREVIATIONS
Kant
GMS
Idee
KpV
KrV
Pad
Rel
Groundwork of the metaphysics of morals

(Grundlegung zur Metaphysik der Sitten)
(1785, Ak iv:385464)
Immanuel Kant, Practical philosophy, ed. Mary Gregor
(Cambridge: Cambridge University Press, 1996), pp. 37108
Idea for a universal history with a cosmopolitan purpose
(Idee zu einer allgemeinen Geschichte in weltburgerlicher Absicht)
(1784, Ak viii:1532)
Immanuel Kant, Political writings, ed. Hans Reiss, trans.
H. B. Nisbet, second edition, (Cambridge: Cambridge
University Press), pp. 4153
Critique of practical reason
(Kritik der praktischen Vernunft)
(1788, Ak v:1164)
Critique of pure reason
(Kritik der reinen Vernunft)
(1st edn 1781, 2nd edn 1787, Ak iii:1552)
Page numbers are from the second edition
Immanuel Kant, Critique of pure reason, ed. and trans.
Paul Guyer and Alan W. Wood (Cambridge: Cambridge
University Press, 1997)
Lecture On pedagogy
(Padagogik)
(1803, Ak ix:43799) (no translation)
Religion within the boundaries of mere reason
xiii
xiv
RL
TL
VE
ZeF
Fichte
GNR
Hegel
GPhR
NRSW
list of abbreviations
(Religion innerhalb der Grenzen der bloben Vernunft)

(1793, Ak vi:1202)
Immanuel Kant, Religion within the boundaries of mere reason,
in Kant, Religion within the boundaries of mere reason and
other writings, ed. Allen Wood and George di Giovanni
The doctrine of right (Part 1 of The metaphysics of morals)
(1st edn 1797, 2nd edn 1798, Ak vi:203372)
The doctrine of virtue (Part 2 of The metaphysics of morals)
(1st edn 1797, 2nd edn 1798, Ak vi:373493)
Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge: Cambridge University Press, 1996), pp. 507615
Immanuel Kant, Lectures on ethics, ed. Peter Heath and J. B.
Schneewind, trans. Peter Heath (Cambridge: Cambridge
University Press, 1997)
(Eine Vorlesung Kants uber Ethik)(c. 187580, Ak xxvii:286)
Toward perpetual peace
(Zum ewigen Frieden)
(1795, Ak vii:34186)
Foundations of natural right, according to the principles of the
Wissenschaftslehre
(Grundlage des Naturrechts nach Principien der Wissenschaftslehre)
Johann Gottlieb Fichte, Foundations of natural right: Grundlage
des Naturrechts nach Principien der Wissenschaftslehre, ed.
Frederick Neuhouser, trans. Michael Baur (Cambridge:
Cambridge University Press, 2000)
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)
Lecture on Natural law and the science of state
(Vorlesung uber Naturrecht und Staatswissenschaft)
(181819)
(No translation)
list of abbreviations
PhR
xv
Lecture on The philosophy of right

(Vorlesung uber Philosophie des Rechts (18245))
(No translation)
Nietzsche
GdM
On the genealogy of morality
(Zur Genealogie der Moral)
Friedrich Nietzsche, On the genealogy of morality, ed. Keith
Ansell-Pearson, trans. Carol Diethe (Cambridge: Cambridge
University Press, 2007), pp. 1128
WuL
On truth and lies in a nonmoral sense
ber Wahrheit und Luge)
(U
Friedrich Nietzsche, Writings from the early notebooks, ed. Ladislaus
Lob, 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 of
all 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 work
for which no English-language citation is given, then the text has been
newly translated. In the interests of simplifying the footnote citations,
this will not always be noted unless there is a specific need for
clarification.
INTRODUCTION
A nearly total unanimity prevails with regard to the fundamental

necessity of a public penal system.1 Even among those few who advocate the abolition of all punishments, a large majority advocates
instituting alternatives to the usual prison sentence, rather than
calling for the abolition of punishment without anything to replace
it. When seen in this light, the existence of public penal law can be
regarded as being completely justified. The manner in which punishment might actually be justified, however, remains just as controversial
a subject as determining the appropriate amount of punishment. This
is because these issues are closely related to one another.
Every theory of punishment currently advocated shares the rejection of the system of punishment which was prevalent in the early
modern age. This rejected system, illustrated by such penal provisions
as the Constitutio criminalis Carolina, enacted in 1532, was placed in
opposition to the modern system of punishment by Michel Foucault
in Discipline and punish. The early modern system differs from the
modern system in the sense that the latter prefers either prison
sentences or (if any) the most painless and most decent death sentences possible.2 It is worth noting that well into the eighteenth
century more than one hundred crimes were capital offenses. Torture,
1. For an example of the few exceptions, see Herman Bianchi, Abolition: assensus and
sanctuary, in Alexander R. Duff and David Garland (eds.), A reader on punishment
(Oxford: Oxford University Press, 1994), pp. 33651.
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 the
Constitutio criminalis Carolina in Friedrich-Christian Schroeder (ed.), Die Carolina:
die Peinliche Gerichtsordnung Kaiser Karls V. von 1532 (Darmstadt: Wissenschaftliche
Buchgesellschaft, 1986).
german idealism and the concept of punishment
which was already systematically employed as an interrogation method,

was often a component of the punishment, as well as constituting an
intensification of the death sentence. Even though torture as a means
for investigation and security is currently being propagated again3 and
even though the death penalty is still supported,4 there are no theorists
to be found who would come out in support of a return to early
modern practice. All contemporary theorists show themselves to be
guided by the humanitarian spirit of the Eighth Amendment to the
United States Constitution: Excessive bail shall not be required, nor
excessive fines imposed, nor cruel and unusual punishments inflicted.
What exactly this humanism consists of, and where the limits of this
humanism might lie, are still controversial issues. However, the fundamental discrepancy is generally seen to be somewhere between the
theories of retributive justice and general deterrence.
Retributivist theories justify the punishment of a criminal on the
grounds that retribution is demanded by justice to compensate for the
inequities created by the crime. On this basis, one may make a subsidiary distinction and ask whether what needs to be compensated for is
the gravity of the offense itself or the malevolence of the criminal that
was illustrated by the said offense. Theories of general deterrence, on
the other hand, justify the punishment on the grounds that all of the
citizens will be deterred from carrying out an offense before it occurs,
either through the threat of a certain punishment or through the
enforcement of the said punishment; the latter option relies on the
example it displays.
On the one hand, the modern advocates of general deterrence
(examples include Thomas Hobbes, Samuel von Pufendorf, Christian
Wolff, Cesare Beccaria, Anselm Feuerbach and Arthur Schopenhauer) consider that the uselessness inherent in punishments that
do not serve to deter others from committing crimes is inhumane.
The current system of positive criminal law also requires that each
punishment contains elements of general prevention as stated in the
initial paragraphs of the German Penal Code:
3. See, for example, Winfried Brugger, Darf der Staat ausnahmsweise foltern?, Der Staat,
35 (1996), 6797; and Vom unbedingten Verbot der Folter zum bedingten Recht auf
Folter? Juristenzeitung, 35, no. 4 (18 February 2000), 16573. For a detailed
repudiation of torture, see, for example, Henry Shue, Torture, Philosophy and Public
Affairs, 7, no. 2 (197788), 12443.
4. See Ernest van den Haag, Why capital punishment?, Albany Law Review, 54, nos. 34
(1990), 50114.
introduction
By serving a prison term, a prisoner should eventually become able to

lead a life in which he or she commits no crimes (which is the goal of the
execution of punishment). The completion of a term of imprisonment
also serves to protect the general public from further crimes.5
On the other hand, theories of general deterrence often draw the

criticism that they treat criminals inhumanely, because the intended
aim of punishment is conceived of solely to serve the interests of
the other citizens without taking into consideration the dignity of
the criminal. This criticism is leveled in its most intense form when
it is postulated that general deterrence allows the punishment of an
innocent 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 reason
that the aim of punishment ignores the interests of the convicted, or on
the other hand, in accordance with retributivism, for the reason that
any kind of interest whether of the criminal or of the fellow citizens
should be disregarded by the penal sentence, because a punishment
justified in a retributivist way is merely about inflicting on the criminal
what he or she intrinsically deserves because of the deed. A punishment
situated in the retributive model should be concerned with inflicting a
punishment that is in line with what the prisoner merited because a
certain crime was perpetrated. The latter objection is raised by retributivism. The former objection is raised by positions that hold that the
rehabilitation of the perpetrator should be the punishments goal.
Admittedly, the latter position recognizes that for the purpose of
reaching rehabilitation a certain period of time of specific deterrence
may become necessary, in which society is protected from further
crimes 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 passage
is the precept to the later statutes in the codex. Compare as well with the judgment
of the German Federal Court of Justice (Bundesgerichtshof) on December 8, 1970, 1 StR
353/70. (Translation mine.)
6. Cf. Peter Koller, Probleme der utilitaristischen Strafrechtfertigung, Zeitschrift fur die
Gesamte Strafrechtswissenschaft, 91 (1979), 4595; Kristian Kuhl, Die Bedeutung der
Rechtsphilosophie 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 (17811832): Studien zu seiner Philosophie und zum
Krausismo (Hamburg: F. Meiner, 1985), pp. 8092 (p. 29). For a refutation of this
objection, compare Fred Rosen, Utilitarianism and the punishment of the innocent:
the origins of a false doctrine, Utilitas, 9, no. 1 (March 1997), 2337.
Retributivism criticizes theories of general deterrence as well as

theories of rehabilitation considered as the aim of punishment for
much the same reason, that is, because these theories treat punishment as a mere means to an end. Retributivists themselves hold that
punishment ought to be justified without any reference to further
goals, invoking the rationale that the criminal deserves it because he
or she knowingly violated the law. It is for this reason that legal
theorists term retributivism an absolute theory, because, according
to retributivism, punishment represents a good that does not depend
on any goal. On the contrary, in the relative theories the justification
of punishment always depends on its relation to a goal. According
to the proponents of retributivism,7 it derives its superiority from
the fact that it alone as the only theory of criminal justice that views
the punishment solely as a goal in itself treats the malefactor not as
a 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 that
it is not retributivism but rather rehabilitation that meets this
requirement.
Proponents of rehabilitation obviously consider the aim of punishment to be to grant the criminal the best possible status, by which
it is understood that this is the way of treating the criminal that is
both the most benevolent and still compatible with the protection of
society against further crimes. Thus, with rehabilitation as an aim in
punishment, clear limits are set for specific deterrence. Without specific deterrence, rehabilitation would be unthinkable, for if there were
no public enforcement of the law, there could in turn be no reintegration back into society, for there would be no rule of law into which
a criminal could be reintegrated after the rehabilitation has been
completed.
Unlike rehabilitative and specific deterrent punishments, retributivism does not concern itself with the future of the malefactor beyond
the duration of the punishment. In this respect, the theory of rehabilitation is the only one that can categorically exclude those sorts of
punishment 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 the
Federal Constitutional Court of Germany Winfried Hassemer rightly
7. See, for example, Otfried Hoffe, Gerechtigkeit: eine philosophische Einfuhrung (Munich:
C. H. Beck, 2004), p. 83.
introduction
observes: The goal of rehabilitation is necessarily contained in opting

for prison sentences. Corporal and capital punishments do not need
any further justification than retribution.8
Even when one assumes that the punishment is merited and therefore justified as being compensation either for the crime or for the
criminals own wickedness, and asserts the claim that one should treat
the criminal humanely, this allegedly humane view of the criminal
should also concern itself with the situation in which the criminal will
be after the complete term of imprisonment has been served. The
absence of appropriate rehabilitative measures leads not only directly
to the creation of a durable criminal environment with an ensuing
reduction in public safety,9 but also to a state in which the criminal is
punished twice for a crime, by being continually stigmatized, instead
of enabling him or her to express remorse and to reach a reconciliation with society.10 Therefore, concern for the convicts future after
the sentence necessitates appropriate treatment while the sentence is
being served. Retributivism may attempt to fulfill this requirement of
humaneness, along the lines of what Paul Ricur attempted.11 In
doing so, retributivism stumbles upon what Hassemer terms the antinomy of punitive goals, which refers to the fact that in many cases the
various existing theories of punishment do not allow for the same
amount of punishment.12 Hassemer observes that
A period of punishment limited by the proportionality principle and
as required by the goal of retribution normally does not suffice for a
treatment, 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 be

set. Either retributivism should be deemed to be the primary aim and
rehabilitation the secondary aim meaning that the rehabilitation will
be 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 University
Press, 1989), p. 102.
10. Cf. Braithwaite, Crime, shame and reintegration, p. 101; Hassemer, Grundlagen des
Strafrechts, p. 289.
11. Paul Ricur, Le Juste (Paris: Editions Esprit, 1995), p. 203.
12. Hassemer, Grundlagen des Strafrechts, p. 291.
13. Hassemer, Grundlagen des Strafrechts, p. 291.
The form of retributivism that is given priority regards its moral

superiority as stemming from its being the only genuinely humane
justification for punishment. It sees itself as better than rehabilitation
for 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 his
or her guilt, responsibility or malevolence. However, in doing so,
retributivism ignores the following points.
First, the following differentiation is necessary. Retribution can be
understood, on one hand, in its minimalist sense, as meaning that the
guilt of the punished, without any exception, should be the prerequisite for any given punishment.14 According to this understanding, all
theories of punishment, that is, theories both of general deterrence
and of rehabilitation, are retributivist theories.15 On the other hand,
retribution can also be understood to mean (1) a response to the
offense that strives to provide equal compensation for the criminals
merit or guilt and (2) that this equal compensation is the sole just
punishment. Thus, it excludes any goal in punishment (for instance,
general deterrence, rehabilitation and specific deterrence). This is
what one usually understands under the term retributivism. Retributivism consists in the acceptance of the latter (disputable) thesis.
When I speak of retribution in this book, I will be making reference
to the latter understanding of retribution.
Secondly, in the justification of the punishment as retaliation, it is
not the criminals future, but rather his or her past that is taken exclusively
into account. As we have seen, this occurs with appeal either to the
criminals merit or to his or her guilt. The concept that lies at the root of
this guilt can be from a modern, humane perspective only the concept
of responsibility. In fact, in every contemporary retributive theory as
opposed to the cases of the deterrent and rehabilitative theories there
is the imperative to respect the responsibility of the criminal as constituting his or her own dignity and to respond to this dignity with equal
compensation.
Throughout this book, I will proceed as would a retributivist,
assuming that a human beings responsibility for his or her actions,
as opposed to other living beings, is actually what constitutes the
special 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.
introduction
differentiation must be taken into account. There are always two

dimensions to the assertion that a human being bears responsibility
for his or her actions. On the one hand, this responsibility means that
different actions will lead to different consequences, and especially
that all actions, which from the perspective of morality are valued
differently, will also lead in some cases to different consequences.
Among other things, it means that actions violating the law necessarily
could or should lead to a worse situation. On the other hand, the
humans status as a being capable both of reason and of assuming
responsibility is an inalienable status that cannot be taken away from
that human or any other human. The assertion that the human beings
past illegal actions should carry consequences should not lead us to stop
treating this human being as a being capable of reason, except when
this prevents his or her fellow human beings from exercising that same
status. Otherwise, the perpetrator would be treated as a person capable
of assuming responsibility only up to the crime; after the conviction, however, the criminal would lose this status, that is, his or her worth. In this
way, retributivism would lack exactly the sort of behavior toward the
criminal which it regards as its moral superiority over the other theories
of punishment. In short, retributivism does not sufficiently differentiate
between the actor and the action. In this respect, the viewpoint that
would trace a retributivist influence back to Christianity proves not
to be truly convincing.16 Rehabilitation is equally rooted in the
Christian tradition, as the rehabilitation theorist John Braithwaite
suggests in his plea for reintegrative shaming:
It is shaming which labels the act as evil while striving to preserve the
identity of the offender as essentially good. It is directed at signifying
evil deeds rather than evil persons in the Christian tradition of hate the
sin and love the sinner.17
Thirdly, retributivism relies on the assumption that the consequences

of the crime, for which the criminal should take responsibility, should
consist of providing equal compensation, because that is what justice
demands. While retributivism focuses in this manner on demanding
an equivalent punishment, it misses an indisputable consequence of the
crime : at least temporarily, the existence of a commonwealth between
the 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.
A short thought experiment may illuminate this point. Let us

assume the retributivist position as the starting point for this experiment, which is a position that holds that the severity of punishment
should be equivalent to the gravity of the crime and that this alone
should represent without any specific aim of punishment the only
justified solution. Then, let us assume that in this fashion we reach the
conclusion that a criminal has merited a twenty-year prison sentence.
It is usually the case that serving the prison term begins, at the very
latest, after the judgment has taken effect and after all recourse has
been exhausted (in reality, however, the criminal still considered
officially innocent has already been in custody either since the
investigation commenced or since the issue of the arrest warrant). In
this case, retributivism should raise no objections if the enforcement
of the sentence were to be postponed. If the criminal were, for
example, twenty years old, this would allow him or her to be in prison
from age thirty to age fifty, but to remain a free person from twenty to
thirty.18 It is not only for pragmatic reasons (for instance, because of
administrative concerns regarding prison capacity) that our society
would reject such a reform of the prison system; our society would find
this fundamentally and wholly unacceptable because it would severely
endanger public safety. Above all, such a reform would be rejected
because of a specific deterrent rationale. Now, the danger to society is
clearly consequent to the crime. Following the specific deterrence
model, the criminal is therefore liable for this consequence caused
by his or her criminal offense. Whereas it is controversial whether the
moral demand of retribution for the guilt that is addressed by retribution can be seen as a consequence of the crime, the danger to society
is an unquestionable consequence arising from the actions of the
criminal. Retributivism, however, does not actually take into account
the degree to which society is endangered by these crimes, which is at
odds with retributivisms own self-portrayal of itself as being the only
theory of criminal justice that requires the criminal to shoulder the
burden of the consequences of his or her actions.
Fourthly, the disregard shown for the consequences of the action
attests to a disregard for the legal dimension of crime as well as for the
punishment of a crime. Retributivism focuses on the guilt of the
perpetrator. Admittedly, it also emphasizes that the punishment of
18. Such a postponement is actually possible under German law, though admittedly only
in a limited number of cases with short sentences.
introduction
the criminal represents justice for the victim.19 Apart from that, concern for the commonwealth does not play a role in retributive theorys
justification of punishment. Consequently, if in the retributivist justification of punishment only the criminal is taken into account, this
begs the question of why the punishment falls under the jurisdiction
of the judicial powers, that is, why it belongs to the commonwealth
which, in other cases, maintains distance from the private sphere of
the individual. If, in the retributivist justification of punishment, only
the criminal and the victim are taken into account, then it is inevitable
to ask why the punishment could not just be decided in a civil trial.
Under these circumstances, we should not be surprised to observe that
retributivism is hardly ever supported by legal theorists, even though it
enjoys a wide esteem among philosophers for its supposed morality, as
well as majority support.20 This discrepancy between the view of
philosophers and the view of legal theorists does not, unfortunately,
receive much attention from philosophers.
Unlike the theories of deterrence and rehabilitation, the justified
implementation of retributivism is not accountable to performance criteria, or to its output (put differently: According to retributivism . . . the
significance of punishment lies outside the realm of social reality).21 If
especially high recidivism figures were noticed in relation to one sort of
crime that was being handled with measures involving rehabilitation
of the criminal, there would then be questions as to the legitimacy of
this sort of penal mechanism. However, the implementation of retributivism is, by its very essence, not dependent on its effectiveness. The
justification for a system of retributive justice is not empirically verifiable through criminological studies, for example. Rather, the criticism
directed toward retributivism must be at the conceptual level. For this
19. This perspective is very questionable, because, in a modern constitutional state, the
criminal proceedings are differentiated from civil proceedings, among other things,
by the former being carried out by the states attorney as the representative of the
interests 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 in
no way seen to be compensation for the infringement on their rights. Since the notion
that punishing the malefactor represents justice for the victim seems not to be central
to the core of retributivism I will disregard this aspect of it, at least in this work, and
concentrate only on the main argument of retributivism. Were this main argument
omitted, the (alleged) justice on the part of the victim for a proponent of
retributivism would also not be enough justification for the ills that the criminal
would 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.
10
reason, the criticism of retributivism that will be made in this book

will present hardly any empirical and interdisciplinary aspects.
Instead, it will concentrate on the conceptual arguments, that is, on
arguments dealing with legal ethics and moral philosophy. In this
respect, this book will proceed no differently than do the depictions
of those legal theorists who still treat absolute theory in relation
to Kants and Hegels argumentation or more exactly, to its theoretical foundation. But my conceptual critique will aim to facilitate
the convergence of the philosophical with the legal debate over the
justification of punishment.
Last but not least, the importance of empirical studies for the
justification of punishment should not be overestimated. Hassemer
calls to our attention
that reliable knowledge about the successes of rehabilitation can hardly
be obtained. The favorite argument based on the recidivism figures,
which because they fluctuate between 30 and 40 percent are said to
discredit the concept of rehabilitation, is untenable upon closer
examination. First, from such statistics one knows only the manifest,
determined and judged criminality . . . Secondly, experiments and
empirical lines of argument face a fundamental problem in penal law:
one cannot isolate the intervening variables; one cannot try out how it
would be if one were to seek to achieve the reform of the offender with
another form of penal consequence.22
Even if an especially high recidivism rate were to be observed in regard

to rehabilitative punishments of certain crimes, it would still be impossible to draw a reliable conclusion from that observation. Is the
limited efficiency of the punishment still better than nothing? Or, on
the contrary, should the punishment of these crimes be completely
abolished? Should a retributivist degree of punishment be substituted
in place of a rehabilitative punishment? Or should an effective general
deterrence punishment be introduced, such as the death penalty?
The decision remains to be made, and it requires overall legal and
moral guidance regarding the justification of punishment. The consequences for the commonwealth should also be brought into consideration. I differ in this point somewhat from the following view of
Kristian Kuhl, for example: All the theories of punishment that are
geared 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.
introduction
11
punishment for reaching these goals, and this assertion really must
be proven with empirical data.23 In my view, the debate about justification of punishment is first and foremost a legal and a moral discussion. In this work I will conduct the discussion accordingly.
The four aforementioned points urge skepticism in relation to the
claim of moral superiority that is made by retributivism, because
retributivism, unlike rehabilitation, does not pay heed to the criminals interests, but instead only to his or her merits or responsibility for
the consequences of every act. First, equal compensation is neither
necessarily nor self-evidently the consequence of a criminal offense.
Secondly, retributivism considers only one dimension of responsibility, but not the inalienable status of a human being as a being capable
of reason. Thirdly, retributivism would allow important consequences
of the criminal offense such as those for the commonwealth to be
disregarded. Fourthly, an explanation is missing in retributivism for
why a punishment intended retributively, whose justification only
concerns the criminal and possibly the criminals victim as well
does not, however, concern the commonwealth, but yet is still
imposed by the commonwealth.
All of these four objections form the basis of my critical debate with
retributivism. Many exponents of it are already cognizant that these
four points are at least problematic. They respond to skepticism by
dispensing with the exclusivity of retributivism in justifying punishment and in determining the degree of punishment. The opinion is
often expressed that retributivism only represents one particular ethical framework for criminal law in which goals of punishment are also
allowed to be fully pursued. When this point is made in order to limit
the application of the other theories of punishment through a kind of
limitation of power, there is a presupposition that retributivism enjoys
legal and moral superiority, for it allegedly respects the rights of the
individual 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 which
the needs of the many as opposed to individual freedom can be reined
in, this seems to me to be a more important question for criminal law
than the question about the existence of guilt in general. The answer
must be compatible with the second option.24
23. Kuhl, Die Bedeutung der Rechtsphilosophie, p. 30.
24. Roxin, Strafrechtliche Grundlagenprobleme, p. 21.
12
The restriction, which Roxin admits partly contradicts retributivism,

proceeds, according to Roxin, by prohibiting a punishment from
exceeding the retributivist degree of punishment, while allowing it
to fall short of the retributivist degree.
The punishment is not allowed to exceed the degree of guilt . . . On the
contrary, it is allowed that the punishment befitting the guilt is
undercut. Actually, such a practice shows itself to be unacceptable for
a consistent retributivist theory, because it means a partial relinquishment
of the suffering in punishment that provides equal compensation.25
The justification for this mixed theory is found by Roxin, as by so

many other mixed theorists who have dominated the debate since
the 1980s, in a form of division of labor between each of the theories
of punishment. This division of labor proceeds by assuming the
following observation:
Each theory of punishment focuses unilaterally on certain aspects of
criminal law the specific deterrent theory focuses on the enforcement,
the retributive notion on the judgment and the general deterrent
conception on the goal of threatening punishment and neglects
the remaining guises of the penal power, even though, however, each
one of them implies specific interferences with the freedom of the
individual.26
According to the division of labor in the mixed theories, every theory

will allegedly confine itself to its main focus. However, such a construct obviously succumbs to the aforementioned antinomy of punitive goals. My investigation will therefore select the search for a single
primary justification of punishment as the guideline. In doing so, it is not
out of the question that the demands of the other theories of punishment may be partly met; however, if they are met at all, they will only
be partly met. This is because the other theories of punishment are
incompatible with one another. They are not, however, contrary theories
of punishment.
In what follows, I will attempt not to lose sight of the conceptual
arguments concrete meaning and its concrete consequences. However, arguing from the point of view of legal ethics, my critique
will, whenever necessary, allow itself to diverge from positive law, as
retributivism also does. In this respect, I will not, for example, accept
25. Roxin, Strafrechtliche Grundlagenprobleme, pp. 234.
26. Roxin, Strafrechtliche Grundlagenprobleme, p. 12.
introduction
13
out of hand the criticism against rehabilitation theory that it allows

neither a minimum nor a maximum time-limit for the degree of
punishment. Roxin regards it as an objection against rehabilitation
that rehabilitation should consequently aim for a treatment until
an ultimate reform . . . even when it is of unforeseeable duration, and
that every punishment has to be dispensed with, even with the gravest
crimes, if there is no danger of recurrence.27 It may be true that
many theorists of rehabilitation28 do not want to draw these conclusions, and are, on account of that, inconsistent. This does not mean,
however, that rehabilitation as an aim of punishment should be abandoned, but instead that one should defend it if one is to defend it at
all in a consistent manner, and for this reason accept and assert
these consequences. I want to do exactly this.
Therefore, in this text, I would like to criticize and refute retributivism in that very aspect which it sees as being the core of its legal and
moral superiority over the other theories of punishment: I will attempt
to show that it is not retributivism, but instead solely rehabilitation, that takes
the responsibility truly seriously that a human being and especially a criminal
has for his or her actions. For the aforementioned reasons, I will relate my
critique first and foremost to Kants and Hegels arguments.
My critiques foundation will be the Kantian concept of right that
builds directly upon the very human dignity to which retributivism
appeals: Right is therefore the sum of the conditions under which the
choice (Willkur) of one can be united with the choice of another in
accordance with a universal law of freedom.29
Kants concept of right can be interpreted in two ways, which
I present and critically discuss in the first chapter, that is, either as a
liberal concept of right, or as a moral concept of right, which is moral
in the narrowest sense of the term. According to the liberal interpretation, the Kantian concept of right (Rechtsbegriff ) means the coexistence of the freedoms of action according to the principle of equal
rights, without any consideration of the content of the choice (Willkur), that is, not dependent on how humans want to employ this
freedom. However, according to the moral interpretation in the
narrow 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.
14
complete categorical imperative to such an extent in the legal system

as the will of a single human being will allow itself to be moved to the
morally right through coerced implementation. In the rest of Part i, the
attempt will be made to show that Kants own retributivist theory is
inconsistent with both the first, liberal interpretation of his concept of
right (see Chapter 2) and the second, moral (in the narrower sense)
interpretation of it (see Chapter 3). Rather, the theory of rehabilitation
is compatible with both interpretations.
In Part ii, I shall proceed from the statement that the Kantian
concept of right comes to be adopted by Fichte as his concept of
right, and by Hegel as the abstract right. Out of Kants concept of
right, Fichte and Hegel develop similar consequences to those that
Kant, in my view, should have logically drawn. Whereas in Fichte the
fate of the criminal being punished is central (see Chapter 4), Hegel, on
the contrary, follows restoration of rights as a guideline and through
this offers a more systematic presentation of a position that, in its results,
is very similar to the Fichtean theory of punishment (see Chapter 5).
The retributivist way of interpreting Hegel, which is frequently deemed
to be axiomatic, will prove to be unfounded and wrong.
In Part iii, I shall attempt to show that, in opposition to Kants
claims to it, retributivism respects the human dignity in the criminals
person less than the alternatives of deterrence. I will deal in Chapter 6
with Nietzsches radical critique of retributivism, according to which
the original motive for the institution of a retributivist punishment is
to be found by no means in a respect for human dignity, but instead in
a cruelty to the criminal free of morality, which rather hinders than
encourages the onset of bad conscience and remorse. In contrast to
that, I advocate, in Chapter 5, out of the perspective of a combination
of specific deterrence and rehabilitation, a treatment of criminals
that respects the dignity in human beings. In order to comply with
the promise that has been made in this introduction to accept all of
the consequences of the alternative to retributivism that I support
and even, as the case may be, the most radical, the most unusual and
the most unpopular consequences I shall even use, in the seventh
chapter, the example of the gravest and most inhuman crimes, that is,
of crimes against humanity.
PART I
DESERT AS THE SOLE
JUSTIFICATION FOR PUNISHMENT
1
THE TWO KANTIAN
CONCEPTS OF RIGHT
1.1. The retributivist turn

The lex talionis is a rule of distribution for the degree of punishment
that is at least as ancient as the biblical verse Breach for breach, eye
for eye, tooth for tooth: as he hath caused a blemish in a man, so shall
it be done to him again.1 It never really was a self-evident, absolute
principle for all punishments that needed no further justification.
Plato, along with several major schools of ancient philosophy, adapted
that premise formulated by Seneca into the famous sentence Nemo
prudens punit quia peccatum est sed ne peccetur.2 It contains two
decisive ideas. First, it implies that punishment is an evil.3 If punishment is to be allowed at all, this evil must be outweighed by the good it
brings with it. Secondly, punishment is, as any institution, a means
toward the realization of the bonum commune, which is the good that
makes all other goods realizable in a sustainable way.
Early modern thought in penal law drew important consequences
from these premises. I would like quickly to highlight the commonalities of the early modern theorists of penal law, however significant
their divergences from other aspects of law and penal law may be.
First, punishment is a means to obtain compliance with the law
from individuals. This is independent of the question whether the
purpose of the law is primarily self-conservation, or perhaps the
1. Lev. 24:1820; see also Exod. 21:2325; Deut. 19:21 (KJV).
2. A sensible person does not punish a man 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. 106355
(pp. 1589).
3. See, for instance, Henry Sidgwick, The elements of politics, reprint of 1891 edition
(New York: Cosimo Classics, 2005), p. 109.
17
18
desert as the sole justification for punishment
pursuit of happiness (Hobbes), or happiness through the least possible

limitation of individual freedoms with an emphasis on the less privileged (Beccaria), or happiness as the greatest happiness of the greatest
number (Bentham), or any other form of bonum commune. Indeed, selfconservation and, in general, protection of citizens is a condition for
pursuing happiness, however one may conceive of the latter. Classical
utilitarian thinkers certainly do belong to this tradition of thought in
penal law: for example Jeremy Bentham, as well as Cesare Beccaria, who
some see as a forerunner to utilitarianism (according to him, the cool
observer of human nature recommends the following maxim of public
ethics: the greatest happiness shared among the greatest number).4 However,
this tradition is not limited to utilitarian authors, but also encompasses
most of the authors of the natural law tradition. Thus, I shall sketch the
most prevalent elements of early modern thought without addressing
the many differences between natural law and utilitarianism that exist
in the realm of public ethics, public law and the foundations thereof.
Punishment is considered to be a sanction, meaning that it attaches
a consequence to certain behaviors and certain deeds by means of a
legal institution in order negatively to influence the motivations that
lead to these deeds. As a sanction, its purpose is to deter a citizen from
breaking the law; thus, punishment must be felt by its addressees to be
an evil, and to that end the punishment must be impressive. In this case,
the addressees are both the convicted and all other citizens whose
passions may incline them to violate the law5 and to disregard reasons
that demonstrate that obedience to the law benefits the entire commonwealth. While Grotius,6 Pufendorf7 and Bentham explicitly mention both goals, the latter states that example is the most important
end of all, in proportion as the number of the persons under temptation 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 and
other writings, ed. Richard Bellamy, trans. Richard Davies (Cambridge: Cambridge
University Press, 1995), pp. 1113 (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. James
Brown Scott, trans. C. H. and W. A. Oldfather, Classics of International Law (2 vols.,
Oxford: Clarendon Press / London: Humphrey Milford, 1934), vol. 2, pp. 11756.
8. Footnote in Chapter 13, } 1 of Jeremy Bentham, An introduction to the principles of morals
and legislation, reprint of 1823 edition (Oxford: Clarendon Press, 1907), p. 171.
the two kantian concepts of right
19
mentions making an example of a criminal; as a means to specific

deterrence he mentions reformation of the criminals will as well as
disablement of his or her physical power.9
Secondly, innocents are not to be punished, for the following two
reasons. To begin with, there can arrive no good for the CommonWealth, by Punishing the Innocent.10 In fact, the punishment, no
longer being linked to the infringement of the law, would not be a
sanction meaning a disincentive any longer. Thus, the punishment
of innocents would contradict the very goal of punishment: deterrence. Furthermore, for seeing all Soveraign Power, is originally
given by the consent of every one of the Subjects, to the end they
should as long as they are obedient, be protected thereby; the Punishment of the Innocent, is a rendring of Evill for Good.11 Thus, the
punishment of innocents would damage the very reason for establishing the commonwealth, the preservation of which is the proper aim
of punishment.
Thirdly, as for the degree of punishment, on the one hand, it is
not (according to Pufendorf) necessary that a man suffer exactly
what he has done to another, that is, that crimes be always punished
by talion.12 Hugo Grotius stresses the point that most often penal law
did not in practice enforce the law of talion, and that this is even the
case with Moses, who most definitely never instituted it in reality. In
fact, according to Grotius, the practice of penal law inflicted, rather, a
degree of punishment proportionate to the crime, rather than equal
to it: most of the time, the true degree of punishment exceeds the
severity of the crime by many multiples.13 On the other hand, the
early modern thinkers make a plea for introducing more differentiation into the degree of punishment that is in accordance with the
severity of the crime committed. There is a sort of ordinal proportionality
to punishment of crimes, wherein the severity of each crime is pegged
to an ordinal scale of punishment. The purpose of this scale, upon
which a less severe crime is always punished more mildly than a more
severe 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:
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. 121318.
13. Cf. Book ii, Chapter xx, Section xxxii(f) in Grotius, The rights of war and peace, vol. 2,
pp. 101013.
20
more severe crimes rather than lesser crimes, as well as to deter

criminals who are committing a crime from committing a further
crime, for instance to deter robbers from murdering the victims who
could identify them.14 Yet this rule for determining the degree of
punishment is not to be confused either with talion law which would
imply an absolute equality between crime and punishment or even
with an arithmetic equality. (An arithmetic equality would imply proportionality between the difference in severity among the crimes, on
the one hand, and the difference of severity between the degrees of
punishment to which criminals are sentenced, on the other hand.)
Since early modern penal theorists argue against punishments
harsher than those required for deterrence, and since they assume
that the practice of talion law exceeds the alleged equality of talion
law, they see themselves, in comparison with talion law, as promoting
clemency toward the convicts. The characteristic of avoiding superfluous and needless pain from being inflicted on the convict is
termed frugality by Jeremy Bentham.15
Fourthly, the inner logic of the requirement for punishment to
be no harsher than deterrence requires leads Cesare Beccaria to
reject punitive torture and to prefer the punishments that simultaneously make the most efficient and durable impression on the people
and are the least painful for the convict. The goal is not the real
infliction of an evil or of pain, but the impression made first of
all on other citizens, secondarily on the criminal himself or herself.
For example, as the utilitarian Henry Sidgwick puts it, from a utilitarian point of view . . . punishment should be, so far as possible, what
Bentham calls exemplarity, that is greater in appearance than
in reality, since it is chiefly appearance that deters.16 Mill defends
the same view when he rejects the abolition of the death penalty and
its replacement by a life sentence because he reproaches the abolitionists with being driven to inflictions less severe in appearance, and
therefore less efficacious, but far more cruel in reality.17
14. Cf. Chapter 27 of Beccaria, On crimes and punishments, p. 63: The harsher the
punishment and the worse the evil he faces, the more anxious the criminal is
to avoid it, and it makes him commit other crimes to escape the punishment of
the 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. 6570 (p. 65).
21
Fifthly, punishment is not considered as following any of the traditional principles of justice such as commutative justice, distributive
justice or corrective justice. Grotius and Pufendorf present the
following arguments for rejecting the classification of punishment in
any of the usual categories of justice. Since the aforementioned proportionality, which would correspond to a geometric equality in an
Aristotelian sense has been rejected, punishment does not follow
any principle of distributive justice. Does punishment follow a
principle of commutative justice? Punishment is not owed to the
criminal as being either the payment of a debt or the fulfillment of
any other agreement.18 Nor is punishment to be thought of as compensation for damage. Also, there exists a duty that requires one to
give back what each person merits only in the case of a positive merit
and not in the case of demerit.19 Instead of only following a principle
of justice, punishment is subject, first and foremost, to prudence
(prudentia) and to public utility.20 Admittedly, punishment can be said
to be just, but rather in a negative sense, by which no injustice arises
by punishing someone who deliberately caused an evil to the other
members of the commonwealth. Yet it is not justice but rather prudence that evaluates the degree of punishment necessary for
obtaining a deterrent effect. Furthermore, although criminals ought
to understand that being punished is not unjust, justice does not
require them either to inflict punishment upon themselves or to
cooperate in the infliction of punishment upon their own person. It
is solely the other citizens who must cooperate in the latter case.
Kant is the first theorist of penal law who derives from talion law
not only a rule for determining the degree of punishment, but also a
justification for punishment. In so doing, he introduces an alternative
theory that is external to the idea of deterrence, to which all competing theories hitherto belonged. Therefore, Kant is not only a central
figure in retributivism, but also its founding father.
The Kantian turn concerns all five elements mentioned. Concerning the first point, Kant excludes neither all the deterrent effects nor
all the deterrent goals of punishment. On the contrary, in the Doctrine
of 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; and
Book viii, Chapter iii(4f) in Pufendorf, De jure naturae et gentium, vol. 2, pp. 11526.
19. Cf. Book viii, Chapter iii(15) in Pufendorf, De jure naturae et gentium, vol. 2, pp. 11867.
20. Cf. Pufendorf, Book viii, Chapter iii(24), in Pufendorf, De jure naturae et gentium,
vol. 2, p. 1210.
22
means (rigorosa) for preventing the recurrence of wrongs by others;

for then a human being would be throwing away his rights and letting
others trample on them.21 What Kant actually rejects is the consideration of punishment as merely being a means toward the fulfillment of
an end, for instance toward the fulfillment of an end of preventing
crime before it occurs. According to Kant, first, punishment is justified
only by the fact that the criminal violated the law, and, secondly, so
that the only such punishment that complies with this justification
is the one that corresponds to justice as defined by ius talionis. In
all cases in which punishment cannot be justified in this way, it is
unjust and therefore prohibited by Kant, however strong the deterrent
effect may be.
The consequence of this is that, concerning the fifth point, Kant
regards the justification of punishment as concerning only justice
and not prudence. According to Kant, not only is it not unjust that
criminals are punished, but justice even requires their punishment
according to talion law. However, it is unclear whether for Kant this
requirement of justice belongs to the realm of legal justice or of moral
justice, which will become apparent below (see Section 1.2).
As for the fourth point, the impression made by punishment on the
citizens and the criminal does not matter to Kant. What matters to him
is, instead, solely the evil inflicted upon the criminal who is deserving
of the punishment. Concerning the third point, Kant unrestrictedly
adopts talion law, although with some adaptations. He considers that
talion law inflicts more severe degrees of punishment than deterrence
theory, in particular more severe than Beccaria does for Beccaria is
allegedly moved by overly compassionate feelings of an affected
humanity.22 One must note that todays defenders of Kants theory
consider that talion law imposes an upper limit on the degree of
punishment, which, they allege, is exceeded by deterrence theories.
Interestingly enough, in regard to the second point, it was not
Kant himself but todays Kantian retributivists who reproach the
deterrence theories for tolerating the punishment of innocents under
certain circumstances.
Owing to the contrast in the first point, that is, between the deterrence theories and Kant, who comes to link the punishment and its
21. TL Ak vi:461. Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge:
22. RL Ak vi:335. Practical philosophy, ed. Gregor, p. 478.
23
degree only to the moral demerit of the criminal (something that only
a human being is capable of bearing), Kantian retributivists also see
the core of their theory as presenting the only theory that respects the
humanity in the person of the criminal. In this respect, Kantian
retributivism radically diverges from the ancient function of talion
law, as described by Henry Sidgwick: at an earlier stage of social and
intellectual development, th[e] distinction [between retribution and
reparation] 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 a
kind of reparation to the person originally maimed.23 Thereby,
respect for humanity is not to be understood in the utilitarian or
natural law sense, that is, as keeping the evil or the pain inflicted
upon the criminal as low as possible. Rather, in Kants view, respect for
the humanity in human beings considered as beings capable of reason
consists in two dimensions. The first dimension is treating the humanity in the criminals person as an end in itself. In this book, I shall not
enter into Kants complex theory of autonomy and free will. Nor do
I need to deal with this theory. I limit myself to observing that there
cannot be any respect for the free will of rational beings without first
allowing them external freedom, provided that it conforms to the
limits imposed by the freedoms of the others. The second dimension
in which respect for humanity consists is the infliction of as much evil
on the criminal as is merited; it is respectful of the humanity in the
criminals person because this demerit is the choice of the criminal,
and therefore everybody ought to have the foreseeable consequences
of their actions expressed, that is, of their merits and demerits.
In } 49e of the Doctrine of right, which Kant devotes to penal law, he
assumes that the latter requirement meaning that everybody ought
to accept the consequences of their actions belongs to the former
one. Yet he does not provide any argument supporting this point,
and I think that none is possible in the Kantian system. Thus I shall
assess the questions of whether retributivism can be grounded in the
first dimension and of whether it can be grounded in the second
dimension as two separate issues.
The importance of proceeding through these two separate
approaches is enhanced by the link between the two dimensions of
respect for humanity, on the one hand, and the controversial issue of
the relationship between law and morality, on the other hand. There
23. Sidgwick, The elements of politics, p. 107.
24
are two theses that pertain to this relationship. Either the legal system
is largely independent from morality and its moral premise does not
go beyond preserving the first dimension that involves respect for
human dignity by mutually limiting external individual freedom of
each 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 and
belongs, incontrovertibly, to the competence of the law, the answer
to the question whether the second dimension can belong to the
competence of the law depends on the answer to the question about
the relationship between law and morality in general. Now, if it
appears that retributivism cannot be justified by respect for human
dignity as taken in its first dimension, one should still inquire into the
possibility of justifying it by respect for human dignity in the second
dimension, for the case of an extensive subordination of the legal
system under morality.
In what follows, this twofold question of whether retributivism can
be supported by either of the two dimensions of respect for human
dignity mentioned above will eventually be answered in the negative in
both dimensions. Thus, unlike the case in which only one of the two
questions is answered in a negative way, the other being answered
positively, I will not be obliged to choose either between the two
dimensions or between the two theses mentioned above. I shall thus
content myself with exploring the consequences of both dimensions
of respect for human dignity for the justification of punishment. It will
ultimately appear that they both support the same justification of
punishment and the same degree of punishment.
1.2. Two Kantian concepts of right

Some like to advance the notion that Kant created a revolution in
legal philosophy just as much as he did in moral and theoretical
philosophy, and that he made a significant contribution to todays
discussion. According to this notion, we have Kants Doctrine of
right24 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 refer
to the title of the work, whereas the doctrine itself, that is, Kants conception of right,
will be lower-case roman. According to the common usage in Kants age, Kant employs
the term doctrine of right in the sense of ius (RL Ak vi:229), referring therefore to law
(Recht).
25
coexistence of empirical freedoms according to the universal law of

equal rights for all legal persons. This concept of right offers two
advantages. On the one hand, it claims to comply not with a
demanding moral justification, but only with the mere reciprocity
in the coexistence of empirical freedoms. On the other hand, it
appears not to contradict morality, because it demands, just as the
categorical imperative does, a universal law. In this sense Wolfgang
Kersting, for example, writes:
The rational principle of right requires from everyone the very
curtailment of freedom to which everyone who influences each other
through actions in their freedom would agree to under fair conditions in
a situation free of threats, namely, a strictly universalizable curtailment
of freedom that limits everyone in the same way.25
Likewise, Otfried Hoffe writes about the Doctrine of right that, on the
one hand, the political and legal theory that derives from it [from the
categorical legal imperative] consists of a political . . . liberalism.26
On the other hand, according to Hoffe, the Metaphysics of morals
orients the principles of moral philosophy in the (metaphysical)
doctrine of right toward an external, and in the (metaphysical) virtue
theory, toward an internal lawgiving.27 I believe, though, that these
aspects are not compatible with one another. In believing so, I am not
in any way disputing that Kant advanced them. He just did not
advance them in the same work.
The first aspect, that is, right as coexistence of empirical freedoms
with equal rights, is advanced by Kant in, for example, Toward perpetual
peace (1795) as a republican constitution, while requiring that the
civil constitution in every state should be republican,28 and defining
the republic in the following manner:
A constitution established, first on principles of the freedom of the members
of 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, in
O. Hoffe (ed.), Immanuel Kant: metaphysische Anfangsgrunde der Rechtslehre (Berlin:
Akademie Verlag, 1999), pp. 4162 (p. 48).
28. ZeF Ak viii:349. Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge:
26

upon a single common legislation (as subjects), and third on the law of
their equality (as citizens of a state) . . . is a republican constitution.29
Let it be noted here that equality before the law is by no means

justified by the categorical imperative.
The second aspect is advanced by Kant in his Doctrine of right, which
is why the Doctrine does not offer a liberal concept of right. With
regard to the relation of right to morality in Kant, one typically
distinguishes three main options: (1) the thesis of complete independence (initially proposed by Julius Ebbinghaus); (2) the dependence thesis of the necessary derivation of right from the principle of
morality (initially advanced by Kantian legal theorists in the years in
between the Groundwork of the metaphysics of morals and the Doctrine of
right30); and (3) the thesis of limited dependence, arguing dependence on morality for the validity (Geltung) or for the adjudication of
right, and, in contrast to that, arguing independence from morality
for the realization or for the execution of right. Independence in the
execution of right has the essential feature that, whereas morality
envisions the execution that is, the carrying out of actions out of
duty (Pflicht), right relies on coercion (Zwang) as the incentive. This
third thesis appears to me to be philologically correct. The systematic
evaluation connected to it overlooks, however, the scope of rights
dependence on morality in the adjudication.
Interpreters of Kant concentrate, though, on what differentiates
Kant from the Kantian legal theorists, who, even before the publication of the Doctrine of right (1797), had tried to develop a theory of
right on the basis of the Groundwork of the metaphysics of morals (1785).
Most of the Kantian legal theorists, such as Gottlieb Hufeland and
Theodor 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 that
which regulates the external relations between humans. External
actions and relations are already objects of moral duty. The difference
between moral duty and right is placed by these Kantian legal theorists
in the differentiation between, on the one hand, what is morally commanded or forbidden and, on the other, what is morally allowed.
29. ZeF Ak viii:349f. Practical philosophy, ed. Gregor, pp. 3223.
30. Cf. W. Kersting, Sittengesetz und Rechtsgesetz: die Begrundung des Rechts bei Kant
und den fruhen Kantianern, in Reinhard Brandt (ed.), Rechtsphilosophie der
Aufklarung: Symposium Wolfenbuttel 1981 (Berlin: De Gruyter, 1982), pp. 14777.
27
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 in
the realm of the moral adiaphoron. On the other hand, the Kantian
legal theorists appeal to the implication that whatever is morally commanded is therefore morally allowed. Thus, moral duty, next to the
moral adiaphoron, belongs to right: I have the right to do the morally
allowed as well as the morally commanded. Depending on the author,
right is therefore classified into absolute or conditional, or into alienable and inalienable, or even into complete and incomplete rights. One
could possibly distinguish a first step to the contemporary differentiation between human rights and the other rights, in so far as human
rights count as being absolute rights, whereas the other rights differ
from commonwealth to commonwealth and from situation to situation.
Such a twofold derivation of right from the principle of morality
leads, however, to three major problems. The first of these is the
derivation that assembles right out of two heterogeneous parts relying
on two very different normative justifications, which calls into doubt
whether right as such is derived, that is, if right really forms a whole
and can be differentiated from morality, or whether it is, rather, a
weakened version of morality. Secondly, the authority to coerce is
lacking. Thirdly, the authority to coerce in the case of the merely
morally allowed appears more difficult to justify, if at all, than in the
case of the morally commanded.
1.3. Kants concept of right derived from the principle

of morality in the Doctrine of right
The division of Kants Metaphysics of morals into the Doctrine of right and
the Doctrine of virtue obviously attempts to propose a doctrine of right
that is not subject to the Kantian legal theorists aforementioned
problems. The core relation between the doctrine of right and the
doctrine of virtue consists of two aspects. First, the doctrine of right
and the doctrine of virtue can be derived from the same principle,
that is, from the categorical imperative; this derivation, in the case of
the law, implies no weakening of morality: the obligation remains the
same. Secondly, the difference between the doctrine of right and the
doctrine of virtue rests on the different motives for their respective
execution. The execution of the doctrine of right only has to deal with
coercion as its motivation, whereas the doctrine of virtue is responsible
for the execution out of duty.
28
It arises from the second aspect that the substantive boundaries of

the duties of right are defined . . . through the range of the external
motivation that is related to them in juridical lawgiving: wherever
coercion fails to motivate, juridical lawgiving does not rule.31 Like
most 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 is
not rescinded for the category of duties of right just because they can be
issued in a juridical manner.32
I would, however, like to accentuate the first point: the area of right
encompasses every ethical duty that can also be fulfilled through coercion. In other words, right should provide for the fulfillment of every
ethical duty that could be fulfilled through coercion by coercion itself
(that is, through both the threat of coercion and the application of
coercion).
But which duties can be fulfilled through coercion? Here too Kants
answer appears to be clear: not maxims, but actions. Bernd Ludwig
formulates this in the following manner:
Now, what is the fundamental substantive difference between the
principle of right and the categorical imperative? . . . When the
disposition (Gesinnung) [Ak vi:393] of the actor which is not
accessible to external lawgiving both cannot and is not allowed to be
included, it is also not possible to address demands to it. However, that
his or her actions at least can coexist with a universal lawgiving is indeed
externally enforceable and without having to exert influence for it on
the maxims of the person concerned.33
It appears to me that there is an inaccuracy in Ludwigs formulation.

The third formula of the categorical imperative, to which Ludwig
obviously refers, reads: act in accordance with a maxim that can at the
same time make itself a universal law.34 If right is derivable, on the one
hand, from the categorical imperative, and it is, on the other hand,
responsible for that part of this imperative that allows it to be imposed
through coercion, then the procedure for determining the content of
right should look something like the following. As a first step, maxims
31.
32.
33.
34.
Kersting, Wohlgeordnete Freiheit, p. 176.

Bernd Ludwig, Kants Rechtslehre (Hamburg: F. Meiner, 1988), p. 95.
GMS Ak iv:436f. Immanuel Kant, Practical philosophy, ed. Mary Gregor (Cambridge:
Cambridge University Press, 1996) p. 86.
29
that survive the test should be determined. As a second step, the

actions (by which I mean both active actions and abstaining from
realizing actions) that these maxims lead to should be investigated.
As a third step, an investigation must really seek to determine which of
those actions can be realized by means of coercion; right must then
see to it that these actions are met with a threat of coercion or an
execution of coercion. Therefore, Ludwigs formulation should be
modified thus: However, that his or her actions at least can coexist
with a universal lawgiving of the maxims, that is, with the categorical
imperative, is indeed partly externally enforceable and without having
to exert influence as a counterpart on the maxims of the person
concerned; and it should also be enforced by means of coercion in
so far as it is enforceable by coercion. In contrast, Ludwigs formulation (that his or her actions at least can coexist with a universal lawgiving) leaves open the possibility that law is nothing other than the
equality of all legal persons with regard to actionable rights. Such a
conception of right is not compatible with Kants Doctrine of right, as
I will show in what follows.
The premises stating that the actions, at the very least, can coexist
with the universal lawgiving of the maxims, that is, with the categorical
imperative, and that this is partly externally enforceable by means
of coercion, and also should be enforced by this, in so far as it is
enforceable, imply the conclusion not only (1) that rights are actionable by the persons concerned, but also (2) that action must be taken,
and (3) that the enforcement should be compelled by coercion from
those who would like to abstain from it. I shall content myself with
only 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 takes
the law (pacta sunt servanda) and the duty corresponding to it from the
doctrine of right, as already given there. Accordingly the giving of the
law 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 lawgiving
connects with that duty, namely external constraint, were absent, the
idea of duty itself would be sufficient as an incentive.35
It is certainly the case that no legal system would ever want to enforce
fulfillment of every promise. Not even the relevant specifications from
Kants Doctrine of right fulfill this challenging requirement (this is often
35. RL Ak vi:21920. Practical philosophy, ed. Gregor, pp. 383ff.
30
overlooked by most of the interpreters of Kant).36 The law the

Kantian 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 fulfillment
of 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 case
always is either the fulfillment of the promise enforced, or, alternatively, the delivery of compensations compelled from a promise breaker.
3 By the concerned persons suit, only is the fulfillment of some
promises (for example, only the fulfillment of contracts that fulfill
certain conditions) enforced, or are remedies compelled for nonfulfillment of some promises (for example, for the non-fulfillment
of 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, not
enforced, nor are compensations compelled.
When we take Kants formulation cited above seriously, then only the
first, or at most also the second, option should be possible in his
doctrine of right. But Kant advances different options, which is exactly
what occurs in every legal system. Here is one example for each option
that appears in Kants writings.
1 The preliminary articles of Toward perpetual peace are in no sense at
the disposal of the contractual partners, even if all partners were to
be in agreement to abrogate an article.37
2 Kant claims that if one of the partners in a marriage has left or
given itself into someone elses possession, the other partner is
justified, always and without question, in bringing its partner back
under its control.38
3 The adherence to contracts with servants is only compelled in
limited variations: The contract of the head of a household with
servants can therefore not . . . [be concluded] for life, but at most
only for an unspecified period of time, within which one party may
give the other notice.39
36.
37.
38.
39.
Cf. Kuhl, Die Bedeutung der Rechtsphilosophie, p. 41.

ZeF Ak viii:3437. Practical philosophy, ed. Gregor, pp. 31721.
RL Ak vi:278. Practical philosophy, ed. Gregor, p. 427.
31
4 The sentence sale breaks a lease40 provides for the termination of

rental contracts without compelling compensation for loss (but
with adherence to the cancellation period). In Kants view, as with
the current legal system, only in very few cases (for example, in the
case of lying in court or in the case of breach of trust, etc.) are lies
legally proscribed; still less is the truth compelled, even though with
Kant, as is generally known, the moral prohibition against telling
lies applies without exception.
Even in penal law, in which Kant counts as being the strict supporter
of a categorical retributive imperative, he does not accept all the
consequences of his concept of right. On the one hand (to name a
well-known case) if all the inhabitants of an island leave their homeland, he nevertheless requires that all convicts, including those who
have been condemned to death, should actually serve their sentences
out, but then, on the other hand, he concedes a power of pardon to
the sovereign. In this vein, he writes, on the one hand:
Punishment by a court . . . can never be inflicted merely as a means to
promote 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 categorical
imperative.41
On the other hand, however, he claims:

With regard to crimes of subjects against one another it is absolutely not
for him [the sovereign] to exercise it [the power of pardon]; for here
failure to punish (impunitas criminis) is the greatest wrong against his
subjects. He can make use of it, therefore, only in case of a wrong done
to himself (crimen laesae maiestatis).42
Here the enforcement of the legalethical consequences of a crime

stands at the disposal of other persons. At this point, the categorical
imperative turns into in opposition to Kants concept of right a
mere hypothetical imperative.
Such isolated inconsistencies muddle not only the effective range
of Kants concept of right, but also the systematic differentiation that
ultimately proves to be untenable and leads to the situation that with
Kant some actions, which, according to his Doctrine of right, should be
42. RL Ak vi:337. Practical philosophy, ed. Gregor, pp. 4778.
32
the concern of a legal command or proscription, turn out not to be so,

because Kant separates many duties toward external actions from their
corresponding actions in an inconsistent way.
As is generally known, Kant differentiates between duties of right
and duties of virtue. He claims that duties of virtue cannot be subject
to external lawgiving,
because they have to do with an end which (or the having of which) is
also a duty. No external lawgiving can bring about someones setting an
end for himself (because this is an internal act of the mind), although it
may prescribe external actions that lead to an end without the subject
making it his end.43
But if external actions are actually commanded by the corresponding categorical imperative or by the corresponding duty, then these
actions because they are external, and provided that they are
enforceable by means of coercion as such are legally dictated, that
is, legally enforceable by means of coercion.
In this respect, we should not allow ourselves to be led astray either
by the asymmetry in the presentation of duties in the Groundwork of the
metaphysics of morals for instance, in the commentary on the categorical
imperatives formula of humanity as an end in itself or by qualifying
as duties of right the duties toward the preservation of humanity as
an 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 furtherance of humanity as an end in itself. The asymmetry arises out of Kant
referring to the proscription of false promises and the assaults on
freedom and property as duties for the preservation of humanity as an
end in itself, whereas no duties toward its furtherance are named. At
least three arguments speak against combining the asymmetry in the
presentation offered by Kant with a conceptual asymmetry. First, it
would be absurd to believe that the proscription of false promises and
the attack against freedom and property apply unconditionally. If
property or freedom were to be abused as the means of a criminal
action, both could then be vastly curtailed or suspended for longer
periods of time. Secondly, the proscription of false promises is to be
understood as a duty of right. A false promise is a promise that the
33
promiser does not intend to keep. This intention is, however, not an
external action. An external action is present if, instead of observing a
false promise, one observes the non-fulfillment of a promise; in this
way, however, the moral prohibition of false promises becomes a legal
proscription of non-fulfillment of promises. Last but not least, thirdly,
the duties toward the furtherance of humanity as an end in itself can
be formulated as duties toward external action, for instance the duties
toward establishing basic schooling or basic medical care for all. Being
duties toward external and enforceable actions, these duties
should definitely belong to the duties of right, whereas in Kant they,
being duties toward the furtherance of humanity, count as being
duties of virtue. For these reasons, the identification of the differentiation between duties toward the preservation of humanity and duties
toward the furtherance of humanity, on the one hand, with a differentiation between duties of right, in the legal sense of enforceable rights,
and duties of virtue, in the sense of non-legally enforceable duties, on
the 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 single
proscribed, dictated and allowed actions, because the moral appraisal
of maxims and actions presupposes knowledge of their empirical
attributes and of their context. In contrast, however, it is always
possible to determine with the categorical imperative whether a given
maxim or action is contrary to duty or in conformity with it, and, in
the latter case, whether it is morally dictated or just morally allowed.
Apart from that, every maxim, being as general and indirect as it can
be, is focused on external actions. If we have a general maxim for
guiding our lives, we can really determine for all cases that this or that
external action is compatible or incompatible with our maxim, and if
we necessarily have to carry out the action when following our maxim.
That is why one can say about all external actions taking into
consideration their entire context whether, in each case, they are
morally proscribed, dictated or merely allowed.
All external actions can potentially be influenced by coercion to
some extent. Therefore, all external actions are not only objects of
moral judgments, but also objects of right in the sense of Kants
concept of right. Apart from that, the moral and the legal judgments
of external actions should be decided identically every time. The only
thing in the moral judgment that does not have any influence on the
34
legal judgment is the answer to the question whether the morally and
legally dictated action was executed out of duty or merely in conformity with duty.44 In this respect, morality is certainly more demanding
than law.
The range of right presented here, nevertheless, greatly surpasses
whatever belonged to the area of law. The proximity of law and
morality 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 of
personal lifestyle pertaining neither to the rights nor to the legitimate
interests of other individuals, then Kants conception of right appears
to 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 group
of authors one tends to refer to for the definition of what the liberal
state is, these implications may astound. Yet, the liberal interpretation
of Kant is based on two elements.
Regarding the first element, the derivation of the concept of right
from the categorical imperative is usually underestimated. Marcus
Willaschek, for example, goes so far as to claim:
Kant nowhere really says that the principle of right can be derived
from, or is based on, the categorical imperative. The moral law and
the categorical imperative are not even mentioned in }}AE of the
Introduction to the doctrine of right, where Kant introduced the
principle of right.45
And Thomas Pogge begins his enquiry into the concept of right
directly with the last lines from } B: Kant defines Recht as the whole
of the conditions under which the choice of one can coexist . . . with
the 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 differentiations
between morality and legality and between the duties of law and virtue are two different
differentiations.
45. Marcus Willascheck, Why The doctrine of right does not belong in The metaphysics of
morals, Annual Review of Law and Ethics, 5 (1997), 20527 (p. 230).
46. Thomas W. Pogge, Is Kants Rechtslehre a comprehensive liberalism? in Mark
Timmons (ed.), Kants Metaphysics of morals (Oxford: Oxford University Press,
2002), pp. 13358 (p. 137).
35
In the Introduction to the metaphysics of morals, the division of

the Metaphysics of morals is undertaken according to the motive of the
fulfillment of the dictates of the categorical imperative. The decisive
definition of right is found in the following formulation: The doctrine of right and the doctrine of virtue are therefore distinguished
not so much by their different duties as by the difference in their
lawgiving, which connects one incentive or the other with the law.47
This definition consists of (1) the duties that are common to morality
and 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 and
third paragraphs give a definition of right. Right at the beginning, the
duties that are common to morality and to right are recalled: The
concept of right, insofar as it is related to an obligation corresponding
to it (i.e., the moral concept of right).49 The three following points
do indeed draw on the difference between right and morality (right
pertains only to, first, the external and indeed practical relation of
people; secondly, choice [Willkur] in contrast to mere wishes; and,
thirdly, the form of choice in contrast to its matter). One usually
overlooks, though, that at the end of this very same paragraph duties
that are common to right and to morality are again mentioned:
Kant speaks of the form of choice insofar as choice is regarded
merely as free, and whether the action of one can be united with the
freedom of the other in accordance with a universal law.50 Hence,
it is about choice as form, that is, about choice insofar as choice is
regarded merely as free.
This clearly means that the freedom we are talking about here is
not the freedom of action in choice, but instead the pure freedom
required in Kantian morality, that is, a freedom of will in autonomy.
Therefore, the universal law concerned here cannot just be equality
of rights, but instead must be a law suited to the actors freedom of
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 why
Willascheks alternative thesis is much closer to the official thesis than he himself
believes, and why it is not really relevant for the treatment of our problem. Willaschek
himself believes his thesis to be a thesis of limited independence of the law from
morality; in his own words, rather an independent, basic law of practical rationality
(p. 223).
36
will: it is the same universal law as the one in the third formula of the
categorical imperative (act in accordance with a maxim that can at the
same time make itself a universal law).51 Because the individual formulas
of the categorical imperative are versions of the same imperative, and
therefore equivalent, it also says: So act that you use humanity, whether in
your own person or in the person of any other, always at the same time as an
end, never merely as a means.52 The reciprocal relation of choice is,
therefore, more demanding in relation to Kant than the mere coexistence of the powers of choice or the freedoms of action under the
principle of equality as in the liberal understanding of a just legal
system.
The reference to a demanding moral freedom is indeed present in
the concluding formulation of } B, in the Introduction to the doctrine of right, though not emphasized: Right is therefore the sum of
the conditions under which the choice of one can be united with the
choice of another in accordance with a universal law of freedom.53
One should, at this point, take note that Kant uses the expression in
accordance with a universal law of freedom, which is neither the same
as 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, of
freedom would be superfluous. At the beginning of } C, we find a
variant 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 this
point with the following argument: as to the universal principle of law itself, it is
hard to sustain the view that it can be derived from the moral imperative. This
principle is: any action is right if it can coexist with everyones freedom in
accordance with a universal law, or if on its maxim the freedom of choice of each
can coexist with everyones freedom in accordance with a universal law. . . This
principle may bear a superficial resemblance to the Formula of Universal Law: Act
only in accordance with that maxim which you can at the same time will that it become
a universal law . . . Like that formula, the principle of right provides us with a test only
of the permissibility . . . But the principle of right says nothing about willing maxims as
universal laws. A. W. Wood, Kants Doctrine of right, Introduction to Otfried Hoffe
(ed.), Immanuel Kant: Metaphysische Anfangsgrunde der Rechtslehre (Berlin: Akademie
Verlag, 1999), pp. 1939 (p. 35). Woods argument against the derivation of law from
the categorical imperative is, however, only valid against a derivation of his execution
from the categorical imperative; it is not convincing against an adjudication of actions
from 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.
37
precisely which aspect of choice is meant (the freedom of choice54

[Freiheit der Willkur], that is, the form of the freedom, or the freedom of
will in choice), and this extraneous element according to a universal
law of freedom (emphasis mine) becomes dispensable and disappears:
Any action is right if it can coexist with everyones freedom in accordance with a universal law.55 In the remaining part of } C, as well as
throughout the entire Doctrine of right, the freedom of choice and
universal law is mentioned at every repetition of the definition of right.
Explicit references to the categorical imperative and to duty, however,
appear more seldom, so that the framework, and hence the exact meaning of freedom and universal law, can easily be overlooked.
Regarding the second element: the liberal interpretation of Kant
also rests on a minimalist understanding of the second formula of the
categorical imperative. Kant in fact applies (as shown above) only one
part of the categorical imperative to law, because he is of the opinion
that the duties toward the furtherance of humanity are allowed to
relate only to an end (Zweck) and not to an external action so that
they cannot constitute any external lawgiving. In contrast, for Kant,
the duties toward the preservation of humanity as a goal in itself can
constitute an external lawgiving, and indeed all of these duties as a
whole can constitute such an external lawgiving. Now, the minimalist
understanding leads one to believe that, even if one neither overlooks
nor forgets the derivation of the Kantian concept of right from the
categorical imperative, the requirement addressed by the categorical
imperative to the legal system is strictly limited to the protection of the
fellow citizens or to the prohibition of homicide, slavery, withdrawal of
citizenship and the like.
If law means as much as the coexistence of the freedom of action of
all humans according to the law of equality of right, then what is being
dealt with here is a liberal definition of law, which is derived from the
harm principle without any reliance on Mills utilitarian background
or on any comprehensive moral justification. In such a context,
the requirements of just such a formula as the second formula of
the categorical imperative, which has been conceived of minimally,
are not especially difficult to fulfill (one of the exceptions consists of
the case, for example, of suicide, which is forbidden by the categorical
imperative).
38
A liberal concept of right and a genuine independence of law from

morality can be won not through the adoption of coercion as the
motivation for the execution, but instead, first through an independent adjudication, that is, first when the content of law is no longer
derived from the categorical imperative. It should be obvious that
such a conception of right amounts by no means to legal positivism.
The liberal concept of right comprising the set of conditions for the
coexistence of external freedoms, according to the law of equality,
contains completely normative requirements (of protection of the
existence of all individuals, of equality before the law, etc.). Many
legally positive and valid legal systems do not, however, conform to
these criteria; rather, they infringe them.
1.5. The liberal concept of right

In the introduction I argued that both of the following aspects can
be found in Kants writings: (1) a liberal concept of right and (2) a
derivation of right from the categorical imperative that leads to
another concept of right, which is incompatible with the liberal concept of right. After I have developed the second aspect, I would like to
return to the first. Here I shall content myself with presenting two
passages that document Kants liberal concept of right:
1. On the concept of right as found in Platos Republic, Kant writes:
A constitution providing for the greatest human freedom according to laws
that permit the freedom of each to exist together with that of others (not one
providing for the greatest happiness, since that would follow of itself) is
at least a necessary idea, which one must make the ground not merely of
the primary plan of a states constitution but of all the laws too.56
2. Kant designates as the highest task which nature has set for
mankind the establish[ment of] a society in which freedom under
external laws would be combined to the greatest possible extent with
irresistible force, in other words [the establishment of] a perfectly just
civil constitution.57 Kant states more precisely that it concerns that sort
of society that has not only the greatest freedom, and therefore a
continual 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 Alan
W. 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: Cambridge University Press, 1991), pp. 456.
39
specification and preservation of the limits of this freedom in order

that it can co-exist with the freedom of others.58
The context of both of these quotes, as well as further passages in
Kants works where this concept of right is mentioned, is a teleology of
history, what Kersting terms a conception of right belonging to moral
teleology.59 Kersting rejects, however, that selfsame teleological perspective as being normatively irrelevant. He raises the following objection against the teleological perspective on right, which he diagnoses as
having too strong an affinity with Wolffs 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 toward
the freedom to fulfill duty relates neither to striving for perfection nor
to self-preservation, but instead must relate itself to a duty that demands
the formally perfect treatment of others, and if the moral duty were not
to be grounded in a lex naturalis but in an internal lawgiving that would
not require the realization of natural designs, but instead require an act
out of respect for the law, and for that reason cannot depend on
external assistance?60
Kerstings rhetorical question is astonishing for three reasons.

First, Kersting overlooks that the natural design in Kants moral
philosophy is precisely none other than that the human natural capacities are developed completely and in conformity with their
end directed towards the use of his reason.61 Since a reasonable
being who follows reason and not, for instance, his or her impulses
is really obeying the categorical imperative, the natural design in
humans, according to Kant in his Idea for a universal history from a cosmopolitan point of view (1784), is clearly not only the cultivation of his or her
relation to nature, the civilization of his or her behavior toward fellow
humans and the formation of a civil society or state of law, but also the
moralization of humans themselves, in short, exactly the acting out of a
respect for the law required by Kersting, as well as the moral duty as
an 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 of
the greatest human freedom according to laws without guaranteeing
58.
59.
60.
61.
Idee, Proposition 5, Ak viii:22. Political writings, ed. Reiss, p. 45.

Idee, Propositions 1 and 2, Ak viii:18. Political writings, ed. Reiss, pp. 423.
40
this freedom to the individual only under the condition that he or she
fulfills any duties: individuals only need to heed the laws of coexistence in order to enjoy these freedoms guaranteed by coercion. That
means that they enjoy them equally, no matter if they act morally or
immorally, just not contrary to right. However, freedom is enabling
for an individual striving for the fulfillment of moral duty to fulfill in
an appropriate way the duties, which Kant derives from the categorical
imperative, toward the preservation and the furtherance of humanity
as an end in itself. The individual receives explicitly through freedom
more and better means to preserve and to advance humankind as an
end in itself. In this respect, the freedom to fulfill duty contains neither
a duty to achieve perfection nor a duty toward self-preservation (otherwise we would not be dealing with a liberal concept of right); rather the
freedom to fulfill duty is contained in the duty toward self-preservation
and the duty to achieve perfection (in one of the passages mentioned
above, Kant maintains, for example, that happiness would follow
of itself).62 For this reason, right actually belongs to the moral
pragmatic without thereby betraying Kants system of reason.
Thirdly, Kersting assumes that right of coercion toward the freedom to fulfill duty . . . must relate itself to a duty that demands the
formally perfect treatment of others . . . grounded . . . in an internal
lawgiving . . . requiring an act out of respect for the law, and for that
reason cannot depend on external assistance. Kersting shows here
that he locates the motivation for compliance to law in coercion, yet
sees the source of law as an institution entitled to coerce in the moral
duty to establish the right of coercion toward the freedom to fulfill
duty. In no case do I deny that, according to Kant, it is a part of duty
to enter into a state of laws wherever the state of nature still exists and
to contribute to the establishment of a just legal system to the best of
ones ability. Kant does not say, however, that the legal system should
originate from a humans conscious fulfillment of duty. On the contrary, he mentions misery in the state of nature, that is, natures
coercion being the drive for the establishment of a legal system, which
is opposed to the inclination of humans. Immediately after one of the
citations introduced above, we find the following remark:
Man, who is otherwise so enamoured with unrestrained freedom, is
forced 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.
41
indeed the most stringent of all forms of necessity, for it is imposed by

men upon themselves, in that their inclinations make it impossible for
them to exist side by side for long in a state of wild freedom. But once
enclosed within a precinct like that of civil union, the same inclinations
have the most beneficial effect.63
Kant repeats this conception in his Critique of the power of judgment and
in the text Toward perpetual peace:
For the problem is not the moral improvement of human beings but
only the mechanism of nature, and what the task requires one to know is
how this can be put to use in human beings in order so to arrange the
conflict of their unpeaceable dispositions within a people that they
themselves have to constrain one another to submit to coercive law
and 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, as
I believe, which one should count as being the more Kantian, that is,
which of the two is more consistent with Kants system, and which the
more convincing concept? The answer to the second question appears
clearly to be: the liberal conception of right accords not only with the
reality of the legal systems we rate as just and possibly even the reality
of 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 Kants Doctrine of right, the numerous deviations from those actions that, from the perspective of the categorical
imperative, should be held to be necessary, for instance those concerning the proscription of lying, the dictate of the fulfillment of
promises, etc. In Kants Doctrine of right, for instance, the right to
property entails the right to allow fruit to rot without regard for other
people actually needing the fruit or being able to make better use of
it. The categorical imperative does forbid such waste, whereas Kants
right does not. Furthermore, the liberal concept of right is, in contrast
to the other concept of right in Kant, not found only in one work:
standing alone opposed to the Critique of pure reason (1781), the Idea
for a universal history from a cosmopolitan point of view (1784), the Critique
of 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.
42
text Toward perpetual peace (1795) is the Metaphysics of morals (1797); all
the works listed belong to the critical section of Kants oeuvre.
That the concept of right is more comprehensively developed in the
Doctrine of right than in the other works should count less than the
repetition of the other concept in systematic contexts. Seen systematically, the liberal concept of right is also more convincing than the
others, because it reinforces the independence of right from morality,
expands the range of coercion from compliance with the law to the
establishment of a legal system, and saves the Metaphysics of morals from
dichotomization 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 Kants Doctrine of
right, and derivable from a direct application of the categorical
imperative, are shamefully, or out of respect for such an otherwise
impressive philosophy, left unconsidered. In this context, Kants conception of marriage can be named, for instance, or his rigorous
proscription of public prostitution (venus volgivaga), the island
example, the remark on bestiality, etc.65 Fortunately, such embarrassments cannot justifiably be found in the passages of the aforementioned 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 so
advances a non-liberal concept of right, goes so far as to draw such
conclusions.
Last but not least, many Kant interpreters appear to me usually to
misunderstand the Doctrine of right as a depiction of a liberal concept
of right and thereby they choose more or less consciously the interpretation that they best hold to fit into Kants 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 furthermore regret that Kant in the Doctrine of right did not derive the
corresponding 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 arise
from both of the interpretations. A further reason compels me to such
an investigation. The argument from } 49e in the Doctrine of right in
65. RL Ak vi:27780; RL Ak vi:325; RL Ak vi:333; RL Ak vi:363. Practical philosophy, ed.
Gregor, pp. 42630; p. 467; p. 474; p. 498.
66. Vittorio Hosle, Was darf und was soll der Staat bestrafen? in Hosle, Rechtsphilosophie
des deutschen Idealismus (Hamburg: F. Meiner, 1989), pp. 545.
43
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 of
right elements67 with which the liberal concept of right is incompatible. If Kant speaks, for example, of a proportionality of punishment
to his [the criminals] inner wickedness he obviously means an
internal wickedness as opposed to an external wickedness, that is, to
a mere contrariness to right of external freedom.
It is completely possible that such elements actually do not belong
to the liberal concept of right, though yet being compatible with it,
and that we are dealing with two heterogeneous yet consistent theories of penal law. In this case we would have to observe that Kant does
not keep to the division into the Doctrine of right and the Doctrine of
virtue; a Kantian retributivist penal theory would still, however, remain
tenable, although in this case it would have to be based on a moral
argument. It would then simply reinforce the non-liberal interpretation of Kants concept of right. Therefore, after we have first investigated the arguments that are neutral to the moral theory as well
as those arguments that are influenced by moral theory in support
of retributivism, we can count Kants retributivist penal theory as
being refuted.
In what follows, I will accordingly test initially the arguments independent of the moral theory in favor of retributivism (the second
chapter of this book) and then the arguments influenced by moral
theory (the third chapter). Both the liberal concept of right and
Kants moral theory will prove to be incompatible with his retributivist
penal theory.
67. Cf. Jean-Christophe Merle, Il punto di vista educativo e religioso dei Contributi
dentinati a rettificare il giudizio del pubblico sulla Rivoluzione francese: la
dimensione politica del Saggio di una critica di ogni rivelazione, in Aldo Masullo
and Marco Ivaldo (eds.), Filosofia trascendentale e destinazione etica (Milan: Guerini,
1995), pp. 30325.
2
KANTS LEGAL
JUSTIFICATION OF PUNISHMENT
Utilitarianism and deontological ethics have traditionally dominated

the debate about the justification of punishments and have thereby
been associated respectively with the theories of deterrent and retributivist punishment. Nowadays, the debate about the justification of
punishment is increasingly dominated by mixed theories.1
Most of these mixed theories represent the attempt of deontological philosophers, that is, mainly those influenced by Kant, to break
with the traditional conception of the deontological, that is, the rather
Kantian justification of punishment, as a purely retributive theory.
Such theories were actually with good reason suspected of resting
more on private morality than on legal principles.
In what follows, I will assess the success of these attempts to
strengthen the Kantian retributivist theory through the integration
into the justification of punishment of an element of some form of
deterrence or prevention. First, I hope to show that these attempts are
only superficially mixed, whereas in reality they are completely based
on a retributive foundation. Secondly, I will inquire into the arguments that Kant employs to justify the principle of retribution and
the type of retribution he adopts, and attempt to determine the
weaknesses of this argument. Finally, I will propose a principle of
punishment that, in my opinion, fits in better with Kants principle
1. Cf. Don E. Scheid, Kants retributivism, Ethics, 93 (1983), 26282; Sharon Byrd, Kants
theory of punishment: deterrence in its threat, retribution in its execution, Law and
Philosophy, 8, no. 2 (1989), 151200; Thomas E. Hill, Kant on punishment: a coherent
mix of deterrence and retribution, Annual Review of Law and Ethics, 5 (1997), 291314;
Sarah Holtman, Toward social reform: Kants penal theory reinterpreted, Utilitas,
9 (1997), 321; Otfried Hoffe, Vom Straf- und Begnadigungsrecht, in Hoffe (ed.),
Metaphysische Anfangsgrunde der Rechtslehre, pp. 21333.
44
kants legal justification of punishment
45
of right than his own conception of punishment. In so doing, I will

consciously and consistently employ Kants principle of right as a
criterion for judging the correctness of that principle of punishment.
The Introduction to the doctrine of right formulates the principle of
right in the following manner in } B: Right is therefore the sum of the
conditions under which the choice of one can be united with the
choice of another in accordance with a universal law of freedom.2
Upon this foundation, I will criticize Kants theory of punishment, and
attempt to construct one anew.
2.1. The interpretation of Kants penal

theory as a mixed theory
For a long time the philosophical debate about penal theory was
dominated by the dichotomy between retribution, exemplarily illustrated by the Kantian school, and general deterrence. By the latter
term, I mean deterring the commission of future crimes by deterring
not only the criminal but also other citizens. Recently, however, the
situation has radically changed: the discussion is now dominated by
mixed theories, which conceive of the theories of retributivism and of
deterrence as reciprocally restricting and complementing each other.
On the one hand, general deterrence theorists clearly profess culpability as being a prerequisite to every punishment and endorse some
kind of proportionality between a crime and its punishment. On the
other hand, retributivists generally agree with the deterrence theorists
on the observation that although the principle of retribution in itself
indeed forms a moral principle it does not qualify as a principle of
right in the cases in which punishment demanded by it obviously
would not deter from the commission of any crime.
Each of these positions complements the other, because the element that is stronger in one is weaker in the other, and vice versa. The
general deterrence theories remedy a weakness in the principle of
retribution by providing a rationale for punishment that clearly connects it not with a private morality, but instead with a system of public
law. Retributivism improves the theory of deterrence by furnishing a
principle that gives a simple answer to the question of proportionality
of punishment and seems to be unobjectionable with relation to
justice toward individuals. Sharon Byrds important article about
46
Kants conception of penal law3 brilliantly contrasted this combination with the traditional view of Kants penal theory according to
which it is through and through a retributivist penal theory. Byrd
argues that legal systems must be secured against the inclination of
citizens toward breach of the law and that they obtain this guarantee
through threatening punishment. Hence, according to Byrd, deterrence constitutes the fulfillment of the public right to coerce. But once
the aim of punishment is determined in that way, the execution that
is, the type and the degree of the punishment no longer follows
the principle of deterrence, but instead the principle of retribution.
The main reason Byrd proposes for this shift from deterrence to
retribution is that only the latter treats human beings not merely as
means but also as ends in themselves. Thus, Byrd points out, in Kants
Doctrine of right deterrence is found in the threat of executing the
punishment, whereas the retributive principle is found in the actual
execution of the punishment, so that deterrence and retribution
reciprocally limit each other.
Even though Byrd is of the opinion that threatening and retribution reciprocally limit each other in this relation, there is, however, an
obvious asymmetry between them, which, in my opinion, privileges
retribution and explains the appeal of Kants retributivism in recent
years. Indeed, even though in Byrds reconstruction general deterrence constitutes the aim of punishment, it is retribution that determines the amount of punishment. I would like to express this in the
following manner. Citizens should be punished for their crimes if and
only if:
1 the threat of punishment could deter them (deterrence condition);
and
2 the punishment punishes the crime (retributivist condition); and
3 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: if
they are not satisfied then both the punishment and the threat of it
are prohibited. They do not, however, positively determine the type
and amount of punishment. For clarification of this point I would
like to look at the strongest example upon which Byrd grounds her
new interpretation of Kant, that is, Kants expounding on right of
3. Byrd, Kants theory of punishment.
47
necessity (Ius necessitatis) by way of what Kant terms the plank of

Carneades:
There can be no penal law that would assign the death penalty to
someone in a shipwreck who, in order to save his own life, shoves
another, whose life is equally in danger, off a plank on which he had
saved himself. For the punishment threatened by the law could not be
greater than the loss of his own life. A penal law of this sort could not
have 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 is
certain (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 the
degree of punishment, cf. the third point above). The second premise
is formulated indirectly, that is to say, negatively: A penal law of this
sort could not have the effect intended. The logical opposite to this is
our requirement that penal law prevent some crimes, which means
that the threat cannot be proven to have no effect. The logical opposite does not mean that the effect of deterrence must be either proven
or certain or maximal. This prerequisite is so weak that I have to ask
myself if there is any imaginable punishment that could ever fail to
satisfy it. Is there truly no one who would prefer a certain drowning to
a possible death by a judicial verdict? Provided that the risk of being
caught (for example because there were witnesses to the event or for
some other reason) is not negligible: is there really no person who
prefers drowning and being mourned because of this tragic accident
over the public dishonor of a judicial sentence possibly or probably
leading to death? To stress this point, let us assume that the court will
probably accept mitigating circumstances and sentence the criminal
to a prison sentence of twenty years, so that the person who has the
choice between murder and drowning could be sure that the death
sentence would not be imposed. Can we exclude the possibility that
some persons might prefer to die tragically rather than face a life in
prison without honor and career chances?
If that is so, then the threat of the death penalty (even of a prison
sentence) really does prevent some murders even in such a case of
necessity. I admit that a not uncertain death by judicial verdict
would 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. Practical
philosophy, ed. Gregor, pp. 3912.
48
murders upon which the threat of an uncertain death or prison

sentence by a judicial verdict would have no effect. At least in some
cases, 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 to
the expression could not have the effect intended requires only some
deterrent effect, but by no means the maximal deterrent effect. That is
why there can be a penal law that also imposes a death sentence (or a
lesser sentence) for murder in such a case of necessity. In more
general terms, I do not see any punishment, any imposition of pain
on guilty persons, of which one can claim that it does not have in one
way or another a deterrent effect (the punishment of innocent
people, on the contrary, would probably completely fail, because
such a punishment would not be related to a crime). Even if death
was sanctioned with something as ridiculous as either a one-month
period in jail or a 10,000 dollar fine, these punishments would still
deter some potential murderers; therefore, they would still have a
deterrent effect.
Because the principle of retribution prescribes a punishment for all
crimes, it seems that the first condition listed above (that the punishment must have a deterrent effect) will always be satisfied. Thus, the
condition of deterrence does not in any way limit the principle of
retribution. Nor does the principle of retribution limit the principle
of deterrence, because the latter only requires some deterrent effect.
However if we (in opposition to Byrds interpretation of Kant) were to
understand the deterrence principle in such a way that it also determined the degree of punishment, then it would aim to maximize the
deterrence effect. In this case, the principle of retribution would
actually limit the principle of deterrence. In our case, of necessity,
the maximization of the deterrence effect would require not only the
death penalty, but also death after a lengthy and varied torture.
One might object by contending that the prohibition of torture
only means that a punishment should not surpass a certain threshold,
which, for instance, would be established by the prohibition of a cruel
or inhumane treatment, though below this threshold the principle
of deterrence can still require the adoption of the punishment that
is the most deterrent. This could be an alternative approach, but it
is certainly not Kants or Byrds, because the principle of retribution
does not allow any latitude to apply any sort of leximin. Rather, the
principle of retribution allows only one solution, for instance the
death penalty for murder as in the example given above.
49
If, in Byrds essay, no incompatibility is present between the retributive and the deterrent elements, then it is due only to the weakness of
the criterion that is adopted for the latter: the punishment must at
least deter some future crimes. The retributive principle satisfies this
criterion and is thus not really limited by it. Hence, all that this mixed
theory adds to the classical theory of retribution is the condition that
classical retributivism would not be justified if the prescribed threat of
punishment would not actually prevent at least one crime. The claim
that the aim of punishment for Kant rests in deterrence shows just one
way in which the case for retributivism could be made stronger. For
the rest, the mixed theory prescribes the same degree of punishment
as the classical theory of retribution.
In order to express more clearly my point that mixed theory displays
no actual relationship between retributivism and deterrence, I would
like 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 crimes
amongst themselves should correspond to the relation of the punishments amongst themselves. By that I mean that a serious crime
should be punished more harshly than a minor crime, and that two
equally serious crimes should be punished with equivalent harshness.
4 The degree of punishment must be equivalent to the crime.
The third thesis compares two classes of relations, that is, the relation
between different crimes and the relation between different punishments. In contrast to that, the fourth thesis directly links a crime to
a punishment without consideration of proportionality. The third
thesis obviously does not imply the fourth thesis, it merely prohibits
punishing a shoplifter more harshly than a murderer.
Now consider the following possibilities. The first possibility is that
the thief is sentenced to a week of community service, and the murderer to twenty years behind bars. The second possibility is that the
thief is sentenced to one year in jail, and the murderer is sentenced to
death. The third possibility is that the thief is sentenced to one week of
community service, and the murderer is sentenced to death. All three
examples comply with the third thesis. However, they are not just
50
representing very different approaches to penal legislation. The first

possibility definitely does not comply with the fourth thesis, because
the murderer is not sentenced to death. The criterion for the fourth
thesis appears to me to be better fulfilled by the third possibility than
by the second, but I do admit that this might be controversial. I do not
wish to discuss here the question whether a different penal law could
be imagined that complies with the fourth thesis without complying
with the third thesis, even though this appears to me not to be
impossible. The only reason for my differentiation between the third
and the fourth thesis is to reject the fourth thesis as being incompatible with Kants concept of right, even though Kant defends this thesis.
Yet, I admit that the third thesis is statistically true (I will explain later
what I mean by statistically).5
I observe that theses one through four remain untouched by Byrds
mixed theory. I would like to distinguish now the possible theses
concerning the theories of deterrence.
1 Future crimes are prevented by the punishment of actual criminals
(in contrast to the following two theses, this descriptive thesis
belongs to no normative theory of deterrence).
2 Future crimes should be prevented by the punishment of actual
criminals.
3 Citizens should be punished in such a manner as to provide the
most effective deterrent to future crimes.
4 Criminals and only criminals should be punished, and only in such
manner that provides the most effective deterrent to future crimes.
I observe that the third and fourth theses are not supported by mixed
theory. Because the first thesis is a purely descriptive thesis, the only
general deterrent thesis within the mixed theory that remains is the
second thesis.
Therefore, I will now concentrate on the principle of retribution.
As a first step, I will inquire into Kants justification and attempt to
show that his justification is clearly insufficient. As the second step,
I will show that Kants principle of retribution contradicts his principle
of right.
5. I take statistically to mean that the seriousness of the offense correlates with the
degree of punishment actually carried out, even though there will also be cases where
more serious crimes are punished more mildly than less grave ones. This will be
explained at the end of this chapter as well as being illustrated in Chapter 7, using
the example of punishment of crimes against humanity.
51
2.2. Retributivism as a justification of punishment

Before we inquire into Kants justification of the principle of retribution we will have to distinguish between two meanings of retribution, a
stronger one and a weaker one: these are Vergeltung (the second retributivism thesis, or mere retribution) and Wiedervergeltung (the fourth
retributivism thesis, or ius talionis).6 Most of the current Kantian retributivists argue for Vergeltung and reject Wiedervergeltung.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, he
attempts to justify Wiedervergeltung, or ius talionis, as a principle for
deciding the degree of punishment. In the first step, Kant rejects the
utilitarian conception of penal law:
Punishment by a court . . . can never be inflicted merely as a means to
promote some other good for the criminal himself or for civil society . . .
The law of punishment is a categorical imperative, and woe to him who
crawls through the windings of eudaimonism.8
Most interpreters implicitly derive the principle of retribution (Vergeltung) from this rebuttal. Yet neither the word retribution (Vergeltung)
nor its concept is present before the second step. Even then, one
cannot find mere retribution (Vergeltung) but only equality in retribution (Wiedervergeltung). Therefore, I suspect that the interpretation of
the first step as a justification of mere retribution comes from the
interpretation of the second step, that is, from the justification of
retaliation. Illustratively, the second step contains yet again the same
rejection of the utilitarian theory as the first step: according to Kant,
all principles except ius talionis are fluctuating and unsuited for a
sentence of pure and strict justice because extraneous considerations
are mixed into them.9 In part five of the Appendix, Kant makes a plea
for rejecting the utilitarian concept of punishment in favor of ius
talionis.10
6. The German terms Vergeltung (that is, requiring comparable recompense for a wrong)
and Wiedervergeltung (that is, demanding wholly equal recompense) correspond roughly
to the English terms retribution and retaliation (etymologically: ius talionis),
respectively.
7. Cf. Schied, Kants retributivism; Jeffrie G. Murphy, Does Kant have a theory of
punishment? Columbia Law Review, 87 (1987), 50932; and Hoffe, Vom Straf- und
Begnadigungsrecht.
10. RL Ak vi:363. Practical philosophy, ed. Gregor, pp. 497f.
52
Kants formulation of retaliation, that is, ius talionis, confirms my

suspicion:
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, you
strike yourself; if you kill him, you kill yourself.11
The first sentence cites an undeserved evil and therefore does not
base itself on degree of punishment; thus it cannot concern retaliation, that is, equality in retribution. Furthermore, whoever wishes
to understand the passage as a plea for mere retribution will have to
derive this concept from the second sentence. In fact, there are no
examples to be found in the second sentence for the like-for-like
principle of punishment. For instance, the following interpretation of
the first sentence would be completely sufficient: if you commit a
crime, you throw society back to the state of nature in which you are
not 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 second
sentence.
Therefore, the interpretation of the first step as the justification of
retribution is based merely on retaliation. This interpretation, though,
bases itself singularly and alone on the rejection of the utilitarian
position in which it is tacitly assumed that there cannot be a third
alternative 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 allows
exceptions with regard to retaliation. The first exception rests in the
alleged 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 someone
in a shipwreck who, in order to save his own life, shoves another, whose
life is equally in danger, off a plank on which he had saved himself.
For the punishment threatened by the law could not be greater than
the loss of his own life . . . Hence the deed of saving ones life by
violence is not judged inculpable (inculpabile), but only unpunishable
(inpunibile).12

53
This case is supported by neither the second, the third or the fourth
thesis of retributivism. There is a further exception, if the number of
a murderers accomplices
is so great that the state, in order to have no such criminals in it, could
soon find itself without subjects . . . then the sovereign must also
have it in his power . . . [to] pronounce a judgment that decrees
for the criminals a sentence other than capital punishment, such as
deportation.13
The first retributivism thesis must therefore be modified to: 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.
Or, formulated differently: All criminals, and only criminals, should
be punished, presupposing that the state can ubiquitously enforce the
law from the time before the deed till after the execution of the
punishment. This means: In a stable state, all criminals and only
criminals should be punished.
From this we can conclude that for Kant even the weakest retribution thesis, that is, the first thesis and therefore also the strongest
thesis, that is, the fourth retribution thesis (the principle of retaliation, or Wiedervergeltung) possesses no absolute validity, but instead
is subject to the realization of his principle of right.
2.3. The ambiguity of the concepts

of retribution and of right
Now I would like to explain why the principle of retribution has
nothing to do with Kants principle of right, and furthermore why
even its application in the Doctrine of right can stand in the way of his
principle of right.
It has already been demonstrated how the principle of retaliation
suddenly arises in } 49e, without Kant attempting to deduce it from
his principle of right. If we search for an argument for this principle in
the Doctrine of right, then we find the following in the Preliminary
concepts of the metaphysics of morals (philosophia practica universalis).
If someone does the duty he can be constrained by law to do
54

he does what is owed (debitum) . . . if what he does is less than the law
requires, it is morally culpable (demeritum). The rightful effect of what is
culpable is punishment (poena) . . . conduct in keeping with what is owed
has no rightful effect at all.14
The debt (debitum) results from the criminal not fulfilling what the law
requires. It is here that the ambiguity of the German term Schuld
(guilt) comes into play. It can mean either debt (debitum) or culpability (culpa). Accordingly, retributivism can be schematically described in
the following manner:
1 Whoever makes him- or herself culpable of a crime has not paid his
or 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 is
misleading and leads to fallacies in the case of crimes that cause
damage for which no compensation is possible, that is, for crimes
for which there is no reparation. All infringements that are not irreparable are private crimes not falling under the jurisdiction of criminal
courts, but rather under that of civil courts. Kant gives the example of
embezzlement, that is misappropriation of money or goods entrusted
for commerce, and fraud in buying and selling when committed in
such a way that the other could detect it.15 Such crimes endanger the
existence not of the commonwealth, but only of the individuals who
trusted and freely entered into a contract with the criminal. On the
other hand, public crimes do not damage private legal contracts, but
instead the commonwealth itself. For such crimes, as in the case of the
exercise of possibly deadly force upon a victim, there is no possible
compensation. How can, for instance, a prison or death sentence ever
provide compensation for the victim of such a crime? What could ever
provide compensation for the insecurity of all citizens caused by a
murder? The debts (debita) to the commonwealth can only be paid off
when there is abstention from committing crimes in the future: as
soon as a crime occurs, though, the criminal is no longer able to pay
off his or her debts. Once the criminal has lost civil personality, then
this person has lost even the possibility of paying off the debt to the
community in the future, that is, to abide by the law in the future.
14. RL Ak vi:227f. Practical philosophy, ed. Gregor, p. 382.
55
The objection could be raised that there is no real possible compensation between the commonwealth and the criminal, but at least
an internal compensation is possible. An internal compensation of
this sort stands possibly connected with the problem of the highest
good that is, with the exact correspondence of happiness with
morality.16 Kant seems to support himself on such an idea when he
writes:
This fitting of punishment to the crime, which can occur only by a judge
imposing the death sentence in accordance with the strict law of
retribution, is shown by the fact that only by this is a sentence of death
pronounced on every criminal in proportion to his inner wickedness.17
There are two kinds of objections that can be raised against such a
justification of retributivism.18 The first and most obvious objection is
that, according to Kant, the state exercises control in a state of laws,
that is to say, it regulates the relation between the external freedom of
the individuals to one another. The highest good is, on the contrary,
part of an internal and therefore individual relation.19 Secondly, even
if procurement of the highest good falls under the purview of the
state, then the requirements for the highest good would never justify a
retributivist penal theory, but would rather refute it. The highest good
would require that the relative relation between immorality and
unhappiness, or pain (or remorse), should be the same. Actually,
Kant defines punishment as the right a ruler has against a subject
to inflict pain upon him because of his having committed a crime.20
What should the pain consist of so that it could be comparable with
the amount of immorality? Let us rely on Kants definition from the
Critique of practical reason:
Happiness is the state of a rational being in the world in the whole of
whose existence everything goes according to his wish and will, and rests,
therefore, on the harmony of nature with his whole end as well as with
the essential determining ground of his will.21
16. KpV Ak v:125. Practical philosophy, ed. Gregor, p. 240.
18. For a more detailed refutation of theories of punishment as restoration of harmony,
cf. Jean-Claude Wolf, Strafe als Widerherstellung eines Gleichgewichts, Jahrbuch fur
Recht und Ethik, 11 (2003), 199216.
19. Cf. Thomas E. Hill, Kant on wrongdoing, desert, and punishment, Law and
Philosophy, 18 (1999), 40741 (p. 429); see also below, Chapter 3.
56
If pain is the opposite of happiness, then pain must be the state of a

rational being to whom, in the whole of existence, nothing occurs
according to his or her wish and will. Therefore, pain should consist
in the whole absence of harmony between the persons nature and his
or her end, as well as between the end and the essential determining
ground of his will. Owing to the punishment, the criminal, who is
forced to undergo stringent discipline in order to make the wish and
will spotless again, does not profit from the criminal deed. Therefore,
one must grant that punishment for the purpose of reform brings
with it much pain. Ius talionis also does not make the criminal happy.
It does, however, allow that something in the world occurs according
to the wish and will of the criminal. The principle of punishment
would really be decided according to the criminals maxim: if
the 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 the
thief will be taken, etc. Paradoxically, the punishment reaches something like a universalization of the criminals 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 maxim
for action is invariably forbidden. This should be understood to be a
warning signal against ius talionis. It is also, coincidentally, the maxim
upon which a vendetta, that is, a personal revenge or blood-feud,
is based.
My arguments are indifferent to the debate as to whether ius talionis
is to be taken literally regarding Kant. Everyone knows Hegels ironic
commentary against the literal application of ius talionis.22 However,
Kant declines punishing sex offenders with rape he rather demands
that they be castrated.23 He sentences the bandit to slave labor, but
surely not in order to allow the bandit to fall victim to the same
violence felt by the victims of banditry.24 Because enough Kant interpreters have emphasized this point,25 it is not necessary to explain any
further why my argument does not have to distinguish between the
third thesis and the fourth thesis, but instead allows me to reject these
theses along with the second thesis.
The only retribution taking place in the punishment is purely a
negative type. Kant writes: Whoever steals makes the property of
22.
23.
24.
25.
Cf. GPhR } 101 Anm.

RL Ak vi:333. Practical philosophy, ed. Gregor, pp. 4745.
For example, Hoffe, Von Straf- und Begnadigungsrecht, p. 227.
57
everyone insecure and therefore deprives himself (by the principle of

retribution [Wiedervergeltung]) of security in any possible property.26
This means nothing other than the criminal being shut out of the
commonwealth and sent back into the state of nature. Therefore, only
criminals can and should be punished, under the condition given by
the first modified retributivism thesis27 that a stable state exists. Retributivism is correct then only if we understand it in that way. The use of
the word retribution in this case is indeed misleading. In fact, both
the reciprocal limiting of rights in the commonwealth and reciprocal
protection of the citizens of a state could be called reciprocal retribution. Recognition of the rights of another and the protection of them
are actually a type of reciprocal contribution or tribute. In contrast
to that, a crime is merely negative and in no respect a contribution.
For this reason, the terms retribution and retaliation appear to be
ill suited to describe the modified first thesis.
The simplicity and strictness of ius talionis may be fascinating for
some people. The strictness of a punishment ought not to be mistaken, however, for strictness regarding a concept. Ius talionis cannot
be deduced from Kants concept of right, but rather contradicts that
very concept; it does not even originate from common sense for it is
disputed in equal measure between experts on penal law and between
citizens. Above all else, it is possible to reconstruct Kants theory of
punishment 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 deterrence 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 criminals
reform, which could be called specific deterrence, is given much
attention in todays debate about punishment, because both ideas
are alleged to replace the punishments true meaning with social
considerations that have nothing to do with the crime. I regard such
an objection as being completely baseless, and believe that this type of
specific deterrence could provide rationale for the punishment itself.
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.
58
In the following, I will describe the theory of punishment that Kant

should have built upon his principle of right, in order then to criticize
the theory he actually constructed upon it, which, in my opinion,
contradicts this principle. I will refer as far as possible, though, to
individual passages in Kants Doctrine of right so long as they appear to
me to be compatible with his principle of right. I will argue for the
following theses:
1 Criminals ought to be punished in order to prevent them from
carrying out further crimes. This ought to be accomplished by
measures (2) and (3).
2 Punishment ought to deprive the criminal of the status of a citizen
until (3) is accomplished.
3 The punishment ought to reeducate the criminal so that the criminals status as a citizen can be recognized again.
In the following, I will connect the principle of right with the categorical imperative in order to reconstruct a Kantian theory of punishment.
Besides the rejection of utilitarianism, Kant names two other arguments in his first step (see Section 2.2), two affirmative arguments:
first, the criminal must be punished because of his having committed
a crime; secondly, the penal law is a categorical imperative.28 Both
these points contain a descriptive and a normative element: on
account of the crime committed by the criminal, the said criminal
must be punished for the sake of the categorical imperative. There is,
however, in the Doctrine of right no mention of an imperative that
specially relates to the penal law. In the proper meaning of the term,
only one single categorical imperative can exist, for which the Groundwork of the metaphysics of morals offers different formulations. The
Doctrine of right develops an imperative of right as a particularization
of the categorical imperative by factoring in the circumstance that a
large number of people share the same limited world. I propose to
conceive of penal laws categorical imperative as being a particularization of the categorical imperative of right, under the premise that a
crime has been committed.
By definition, because of the crime, a state of nature prevails over
the relationship between the criminal and the rest of the commonwealth. Kant regards the state of law and the state of nature as a
dichotomy that excludes any third possibility. In fact, Kant terms the
59
situation after the crimes occurrence thrice as state of nature: in the

case of a murder with many accomplices, in the case of a duel and in
the case of infanticide.29 In an original state of nature before the
establishment of a commonwealth, as well as during periods of regression back to the state of nature, the categorical imperative demands
the creation of a legal system. Should the establishment of the institutions constitutive of a state of law not be immediately possible, then
permissive law30 demands achievement of it for everyone therefore,
for the criminal as well in the quickest way possible. I am arguing
that nothing other than punishment can rehabilitate and make the
criminal a member of the commonwealth again, because one must
consider that the criminal is not immediately able to reenter the
commonwealth.
By committing a crime the criminal has obviously accepted the
regression back to a state of nature. In a state of nature, nobody has
any rights and external freedom is continually threatened, whereas,
however, what the categorical imperative demands is that the latter be
respected. The criminal accepts the disappearance of his or her external freedom as well. This means that the will of the criminal commits a
type of suicide, which makes this person unfit to be a citizen.31 The
criminal cannot be treated any longer by the state as a free person, as a
rational being. The type and the degree of punishment that Kant
requires demonstrate this in a clear way. The most famous example
is surely the death sentence for murder. Hermann Cohen and some
modern interpreters32 have raised a convincing objection against it,
saying that this punishment contradicts the Kantian principle of
morality because the punishment irrevocably destroys a rational
being. 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. 4767.
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 Kants
Rechtslehre, in R. Brandt (ed.), Rechtsphilosophie der Aufklarung: Symposium Wolfenbuttel
1981 (Berlin: De Gruyter, 1982), pp. 23385.
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),
150123 (p. 1516); Steven S. Schwarzschild, Kantianism and the death penalty,
Archiv fur Rechts- und Sozialphilosophie, 70 (1985), 34377; Attila Ataner, Kant on
capital punishment and suicide, Kant-Studien, 97, no. 4 (2006), 45282.
60
philosopher who dares to endorse this part of Kants penal law. (An
interesting exception would be Tom Sorell, who argues in favor of
both Kantian retributivism and the death penalty.)33
If one accepts the Kantian theory of penal law, as Cohen does, then
such objections are, however, not convincing. They become so only
when one reconstructs an alternative theory of penal law from Kants
concept of right, which is explicitly counter to Kants theory of penal
law.34 In fact, Kant would deal with these objections in the following
way. First, a rational being ceases to be such as soon as a crime has
been committed, and not only after he or she has been punished. The
punishment only derives the consequence from the factor that
through the crime a rational being denies its rational character.
Secondly, the degree of punishment that Kant imposes for all other
crimes he mentions shows clearly that he no longer sees the criminal
as a rational being. In this vein, I wish to provide the following
example:
But what does it mean to say, If you steal from someone, you steal from
yourself? Whoever steals makes the property of everyone else insecure
and therefore deprives himself (by the principle of retribution) of
security in any possible property. He has nothing and can also acquire
nothing; but he still wants to live, and this is now possible only if others
provide for him. But since the state will not provide for him free of
charge, 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 for
a certain time, or permanently if the state sees fit.35
At this point, Kant fails to differentiate between two issues: (a) that
someone without property has to work for his or her livelihood and
(b) that the work for ones own livelihood means enslavement, instead
of, for example, working as a day laborer. Actually, there is a commonality between this enslavement and the death penalty, just as between
every other punishment mentioned in the Doctrine of right, for example
deportation,36 permanent expulsion from civil society, or castration.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.
34.
35.
36.
37.
Tom Sorell, Moral theory and capital punishment (Oxford: Blackwell, 1987), p. 162.
Cf. Section 2.5.
RL Ak vi:363, Appendix 5. Practical philosophy, ed. Gregor, p. 498.
61
are ways of partially murdering oneself.38 Even if we leave aside the

question whether suicide is punishable or not, then castration by
another person still remains a type of partial death.
With all of these kinds of punishment, the criminal is neither seen
as being a constituent of the commonwealth nor treated as being
such. Kant himself speaks against slavery, just as he does against
voluntary slavery as well as against selling ones own children or
against inherited slavery. Yet he makes a single exception expressly
against a person who has forfeited his personality by a crime.39
The exact consequence of this exclusion appears to have been
expressed most clearly by Fichte: either the criminal is deported like
an outlaw into the desert, where an even more horrible death awaits
than the legal execution of a death penalty, or the criminal becomes
someone who anyone may kill, as dangerous animals are killed. This
extreme conclusion resulting from exclusion from civil society has
only a heuristic function, however, since just as it is nowhere near
Fichtes last word on the subject, so little shall it be the last word in our
reconstruction of Kants theory of punishment.
2.5. Rehabilitation by discipline

Even though the criminal may have lost his or her personality, it is
forbidden from the Kantian standpoint to treat the criminal as one
would a thing or a (possibly dangerous) animal, for the following
reasons.
The most important provided by Kant in } 49e of the Doctrine of right
seems to rely on the following formulation of the categorical imperative: So act that you use humanity, whether in your own person or in the
person of any other, always at the same time as an end, never merely as a
means.40 At first glance, the Doctrine of right seems to apply this
formulation to the instance of the criminal:
For a human being can never be treated merely as a means to the
purposes of the other or be put among the objects of rights to things:
his innate personality protects him from this, even though he can be
condemned to lose his civil personality.41
38.
39.
40.
41.
TL Ak vi:423. Practical philosophy, ed. Gregor, p. 547.

GMS Ak iv:429. Practical philosophy, ed. Gregor, p. 80.
62
Although these sentences resemble one another, they do not appear

to use exactly the same dichotomy: whereas the Groundwork of the
metaphysics of morals speaks of humanity and personality, the Doctrine of right uses the words innate personality and civil personality.
However, if we inquire into the meaning of the latter dichotomy, the
formulation of the Groundwork would then appear relevant for
explaining 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 (philosophia
practica 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 rational
being under moral laws (whereas psychological personality is merely
the ability to be conscious of ones identity in different conditions
of ones existence). From this it follows that a person is subject to
no other laws than those he gives to himself (either alone or at least
along with others).
A thing is that to which nothing can be imputed. Any object of free
choice which itself lacks freedom is therefore called a thing (res
corporalis).42
Thomas Pogge reads this passage in my opinion, correctly as

consisting of two definitions:
By italicizing moral, Kant flags that moral personality is more
specific than personhood. The most plausible specification, which
would also vindicate the therefore, is this: Having moral personality
means being a subject whose inner actions are capable of imputation,
a subject with (transcendental) freedom of the will . . . Persons in
the wider, weaker sense are then subjects whose external actions can
be imputed to them as expressive of their will, choice, or intentions.43
I shall term the latter a subject with freedom of action, in contrast

to the former, which, in addition, possesses freedom of the will.
According to Pogges view, a moral personality presupposes the other,
weaker personality and, hence, freedom of action as well, which
can be guaranteed only by a legal system. This means that moral
personality presupposes civil personality. On the other hand, one

43. Thomas Pogge, Is Kants Rechtslehre comprehensive? Southern Journal of Philosophy, 36
supplement (1997), 16187 (p. 163).
63
can construct a legal system that simultaneously ensures freedom

of action and neither presupposes nor demands that one dispose of
freedom 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 either
the moral personality or a third type of personality. But since the
moral personality presupposes the civil personality, but the innate
personality is opposed to the civil personality, the innate personality cannot then be the moral personality. From the term innate
personality we can derive that this personality is original (that is,
not acquired) and inalienable. From the term personality we can
draw that it is somehow dependent on or relates to a type of imputability. Since Kant, while distinguishing persons from things, identifies
imputation with free choice, that is, identifies it as being freedom
of action, one could conclude that the innate personality, as well as
the civil and indirectly the moral personality, should be guaranteed by
a legal system. Actually, the innate personality receives the protection
of penal law. However, the innate personality is not the same as the
civil personality, and one can be the bearer of the former without
being 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 reason
consists in the humanity in either my person or the person of someone else. In this context, the concept of humanity can now
be understood in a misleading way, that is to say, by linking it to a
dichotomy that only plays a role where the differentiation between
morality and right is at stake.
Now the human being as a natural being that has reason (homo
phaenomenon) can be determined by his reason, as a cause, to actions in
the sensible world, and so far the concept of obligation does not come into
consideration. But the same human being thought [of] in terms of his
personality, 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 the doctrine of duties a human being can and should be represented
in 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.
64

so too merely in terms of his humanity, his personality independent
of physical attributes (homo noumenon), as distinguished from the
same subject represented as affected by physical attributes, a human
being (homo phaenomenon).46
In both of these citations, humanity clearly means moral personality or freedom of the will. Understanding humanity in the formulation of the categorical imperative in this way would then make its
statement redundant: So act that you use [moral personality],
whether in your own person or in the person of any other, always at
the same time as an end, never merely as a means.47 The Groundwork
confirms this reading:
A human being, however, is not a thing and hence not something that
can be used merely as a means, but must in all his actions always be
regarded as an end in itself. I cannot, therefore, dispose of a human
being in my own person by maiming, damaging or killing him.48
In the specific context of the Doctrine of right, the following reading

prevails:
Every human being has a legitimate claim to respect from his fellow
human beings and is in turn bound to respect every other. Humanity
itself is a dignity; for a human being cannot be used merely as a means
by any human being (either by others or even by himself) but must
always 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 beings
in the world that are not human beings and yet can be used, and so over
all things.49
Forbidding that human beings be used can be read in two different

ways. Seen from the standpoint of the categorical imperative, it is
surely justified, since absolutely every human is able to develop autonomy, that is, free will. From the standpoint of the legal system, it
means that this system in no way protects only those citizens who
abide by it and limit their freedom of action in such a way as to be
compatible with the freedom of action of every other citizen. Rather,
the legal system protects all humans, even those who are actually
unable to abide by it, for example children and criminals. By analogy
46.
47.
48.
49.

TL Ak iv:462. Practical philosophy, ed. Gregor, p. 579.
65
to the manner in which Kant deals with children, I propose that

criminals 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 to
chance, since they have brought not merely a worldly being but a citizen
of the world into a condition which cannot now be indifferent to them
even 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 of
others. Yet both of them are capable of acting; therefore, they are
endowed with freedom, though at the same time lacking the ability to
develop, in a way compatible with the freedom of all the other persons
(borrowing Pogges 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 obtaining
this freedom, as soon as and for as long as they are in the position to
respect the freedom of others. In my opinion, that is the meaning of
the term citizen of the world in the aforementioned passage.
At the very least, I see two important confirmations of my interpretation. 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 of
autonomy or freedom of the will, but instead to be seen as being
a faculty which enables the creature to extend far beyond the limits
of natural instinct the rules and intentions it follows in using its
various powers, and the range of its projects is unbounded.52 Thus,
according to Kant, the innate character belonging to all agents
requires their own development toward the establishment of and
membership in a commonwealth, which is a notion that finds further
development in the Doctrine of right. Other writings first and foremost
Toward perpetual peace contain this perspective.

66
Secondly, even criminals sentenced to slavery are expressly not

subjected to regulation by the right of things (or the law of property). For Kant, the difference between a slave and any animal consists
of the slave being able to fulfill duty, even when he or she possesses no
rights whatsoever. In the Division of the metaphysics of morals as a
whole, Kant defines the legal status of things and animals as being
the relation in terms of rights of human beings toward beings that
have neither rights nor duties. Kant says, with regard to the relation
in terms of rights of human beings toward beings that have only duties
but no rights, that it is missing (vacat) in the doctrine of right for
these things would be human beings without personality (serfs,
slaves).53 Though these human beings lack status as citizens, they
cannot, 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 is
supported by metaphysical assumptions. Indeed, Kant obviously
makes metaphysical assumptions in the Doctrine of right, for example
relating to child-rearing: the offspring is a person, and it is impossible
to form a concept of the production of a being endowed with freedom
through a physical operation.55 Yet one need not accept such
assumptions to recognize agents as potential citizens or as actual
citizens, as soon as and for as long as they are able to respect the
rights of others for whatever reasons they choose, be it due to the
observance of the categorical imperative or be it simply due to selfinterest for having their own rights recognized by the commonwealth.
Every citizen of the world has a claim on the state maintaining the
possibility for him or her to become an actual citizen of the said state.
Moreover, the establishment of a commonwealth amongst all human
beings is a postulate or a prerequisite. It is a prerequisite not only from
the standpoint of the categorical imperative, but also when we accept
with Pogges one-way thesis that as soon as we accept Kants categorical imperative we also have to accept his system of right, though not
the other way around. Actually, Kants doctrine of right is not only
descriptive, but also normative it clearly contains the criterion of
universalizability and applies it to all subjects of external freedom, for
example agents. Therefore, Kants system of right does not create
67
agency, but instead presupposes it, and ought to incorporate as much

as possible the principle of the reciprocal restriction of freedom. In
my opinion, this requires that the legal system promote and develop
the ability to follow the law.
In fact, Kant explicitly deals with the development of this ability in
children and in savages. Now, there is no difference between criminals and children in respect to their legal status. Therefore, I will view
the treatment of criminals as being analogous to Kants treatment of
children.
In Pedagogy, Kant conceives of child-rearing as a twofold task:
discipline and culture.56 This goal is to be followed regarding
children as well as savages. Kants concept of right consists in the
reciprocal coercion of citizens guaranteed through public coercion
exercised by the state. In contrast thereto, the criminal is like a child
and a savage and an object of unilateral coercion, which is termed
by Kant in Pedagogy discipline or breeding. Discipline submits the
human being to the laws of humanity and begins to let him feel the
coercion of the laws. Kant conceives of discipline as being a prerequisite for culture, that is, for the positive part of education. Respect
for the law must first be acquired through an external force applied
against human beings, before the ability is gained to obey the law
voluntarily and most importantly to be ones own legislator, which
constitutes 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 to
obey our rule. With a human being, however, we ought to try to
promote humanity as being the goal, that is, encourage external
freedom under the law. Punishment ought to promote the criminals
reform.
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, or
permanently if the state sees fit.57 Unfortunately, however, this timelimit is found nowhere else in the Doctrine of right. The reason for this
might be that Kant assumes that the education of savages is much
more difficult than educating children. He explains: But the human
has from nature such a large tendency to freedom that if he has grown
used to it for a while he sacrifices everything for it.58 However
56. Pad, Ak ix:449. (Translation mine)
58. Pad, Ak ix:442. (Translation mine)
68
nowhere does Kant explicitly exclude for a criminal the possibility of

rehabilitation. In Kants Doctrine of right, a murderer does not always
have to die. The cases listed above (see Section 2.2) are not the sole
exceptions: not only in the case in which a murderer has too many
accomplices, for example, but also in the case in which a sovereign
exercises his right of majesty to grant clemency . . . only in individual cases.59 Therefore, the murderer can eventually be rehabilitated
and released, unless his or her release would threaten the security of
the other citizens, i.e. unless he or she is not yet disciplined.
Now, the possibility of rehabilitation is completely incompatible
with the death penalty. The alternative to retributivism that I am
presenting here is in no way utilitarian. It absolutely does not view
punishment as being merely a means to promote some other good
for the criminal himself or for civil society.60 Admittedly, the murderer certainly profits more from his or her rehabilitation than from
the death penalty. Yet the rehabilitation model that I have sketched
out here supports itself solely on the goal of restoring the civil state
destroyed by the crime in such a way that the criminal can be reintroduced into that very state. The only available means to this end
is punishment with unilateral coercion. During the time when the
discipline is being carried out, the commonwealth is protected from
the risk of the criminal repeating the crime. In the model of punishment I have sketched out, punishment therefore carries out an intention that is not only inherent in the concept of right but also inherent
in the goal of preserving the humanity in every person. I do concede
that the theory of general deterrence also follows a goal inherent in
the concept of right, in so far as the punishment motivates the other
citizens to respect the law; however, a theory that requires deterring
the rest of the population from committing a crime does not really
take the goal of preserving the humanity in every person seriously. In
order to maximize the deterrent effect, the theory of general deterrence could end up increasing the punishment in duration and in
degree beyond what is necessary to rehabilitate the criminal back into
a full-fledged citizen.
One may object to specific deterrence, which I propose as being
an alternative to the principle of retribution, on the grounds that it
infringes the fairness principle, according to which a punishment is
69
not allowed to be lengthened or formulated anew once the criminal

has already been sentenced. Such an objection would, however, rely
on the false assumption that the punishments unrevisability implies
that it is set to an exact duration, that is, to a sentence of a certain
amount of either months or years. But the duration is clearly defined
in my proposal: the punishment ends with the criminals rehabilitation, that is, with the reestablishment of his or her ability to abide
by the law like other citizens, so that the criminal poses no more of
a threat to the law than they do. A weightier objection might lie in
the difficulty in determining exactly when the criminal is able to
reenter the commonwealth as a full-fledged citizen. In order to assess
whether a criminal is rehabilitated one certainly needs, in most cases,
a complex decision in which there always exists a danger of error. Yet
this danger is also present in every court sentence. Moreover, todays
judicial system does actually apply to a certain extent specific deterrence: most prisoners are released on parole, which reduces the
effective term of imprisonment, and recidivists are often punished
more harshly than first-time offenders, and so on.
Let us imagine for a moment how a mechanism might look with
which one could eliminate the risk of arbitrariness in assessing the
degree of rehabilitation. One could imagine an institutional system
that would gradually remove the burden of proof from the criminal
and place it on competent judges. After a while, the burden of proof
would shift and it would become increasingly more difficult for the
appropriate judges to prove that the punishment should continue.
Aside from that, such measures as parole, community service and
more severe sentences against recidivists could constitute a sort of
second test. The second test would be found in real life and would
constitute at the same time both a threat to and an incentive for the
prisoner to reintegrate sincerely into society and respect the legal
system. Finally, the punishment could also include specific measures
of specific deterrence. The judge Marianna Pfaelzer not only sentenced the hacker Kevin Mitnick to exactly four years in jail, but also
forbade him from using computers and mobile telephones without
written approval of his probation officer.
Certainly specific deterrence and rehabilitation the latter meaning strict discipline prevent numerous crimes from being committed
by other citizens. Therefore, both theories always include a certain
dimension of general deterrence, even though they do not target
its maximization. Both specific deterrence and rehabilitation, in
70
accordance with Kants definition of punishment, inflict pain upon

criminals61 and threatening pain for a breach of the law certainly is an
important component of deterring potential criminals from crimes.
Yet general deterrence does not become the rationale for punishment
since we would still punish criminals even when not a single other
citizen would be deterred from committing a crime by the punishment. Nonetheless, the effect of general deterrence caused by specific
deterrence certainly does constitute a strong and very fortunate side
effect of the punishment. I would, however, like to make it clear that
the concept of punishment as being rehabilitation is in no way expressive of overly compassionate feelings of an affected humanity, for
which reason Kant reproaches Beccarias theory of punishment.62
Discipline, breeding, enslavement and unilateral coercion certainly
appear to criminals as being neither attractive nor enjoyable ways to
be punished.
Specific deterrence and rehabilitation may also frequently coincide
with the proportionality of punishment required by the retributivist
thesis formulated above on p. 49. Although one might not be able to
determine in principle that a thief probably might be rehabilitated
more quickly than a hostage-taker or even a murderer, one can do it
empirically. However, there are certainly some cases in which just the
opposite may be true.
Last but not least, retributivism and general deterrence do not
exclude the death penalty. In fact, Kant even explicitly prescribes it
even though many Kantians either deplore this or maintain a careful
silence on the subject. For a long time, there was a debate about the
usefulness of the death penalty as a means of general deterrence.63
Yet all the participants in this debate agreed that it would be justified if
it could be proven that the death penalty was effective in preventing
crimes by other citizens. In opposition to both retributivism and
general deterrence, my alternative radically excludes the death penalty. It also excludes sentences condemning the criminal to life in
prison without the opportunity for a reduction, although it does not
forbid keeping a criminal behind bars for life if all attempts at
rehabilitation should fail. The solution that I have proposed appears

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).
71
to me to reach in the best possible way what the twofold requirement

of all theories of penal law ought to be: ensuring the safety of all
citizens and respecting the humanity of all criminals.
2.6. Summary
The mixed theories of penal law fail in their attempt truly to link
retributivism and general deterrence. They necessarily fail because
both theories clearly differentiate themselves from each other as
justification for punishment. I have therefore proposed we undertake
a twofold modification of the terms in the debate between deontological ethics and deterrence theory. First, I have attempted to show
that Kants retributivist theory of punishment does not draw on his
concept of right. Secondly, I have consciously avoided attempting to
mix the Kantian concept of right with a principle of general deterrence, but instead have argued that the Kantian concept of right is
completely compatible with a theory of specific deterrence, which
encompasses both incapacitation and the rehabilitation of criminals.
This theory is not a mixed theory, and nor does it have to be one.
Therefore, if Kants theory of punishment is justifiable on the basis
of the Kantian system, then this justification could only occur while
attempting a moral argument from which legal consequences arise. In
the following chapter, I wish to show that Kants moral theory cannot
afford this justification. Quite to the contrary, it will confirm the
end result of this chapter, which had as its point of departure the
theory of right.
3
KANTS MORAL
JUSTIFICATION OF PUNISHMENT
Now that I have rejected Kants theory of retribution on the basis of

a liberal interpretation of his concept of right in favor of a combination of specific deterrence and rehabilitation, I would like to investigate whether the alternative, stronger or moral interpretation of his
concept of right can substantiate Kants theory of retaliation. As was
shown in the first chapter, this interpretation consists of an implementation of the categorical imperative in the legal system so far as
the commands and prohibitions of the categorical imperative are able
to be implemented by force.
3.1. Which type of moral proportionality is at issue?

The current retributivist theories of punishment often radically differ
from Kants own theory because they assume that punishment is
justified by a goal, however modest this goal may be. The position
held by expressivists, for instance, states that punishment should be
seen retributively so that societys moral judgment about the punishable action can be expressed.1 Another theory of punishment, such as
the one held by Jean Hampton and Jeffrie Murphy, regards the
punishable action as comprised of the criminal putting his or her
value over the worth of the victim.2 The retributive punishment that
is then stipulated is a punishment that restores the proper relation
between 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), 291314 (pp. 3201).
2. Cf. Jean Hampton and Jeffrie Murphy, Forgiveness and mercy (Cambridge: Cambridge
University Press, 1988), pp. 4553.
72
kants moral justification of punishment
73
In opposition to such theories as well as similar ones, and also

against the legal tradition following Senecas dictum (as I have mentioned before), nemo prudens punit quia peccatum est sed ne peccetur,3 is Kants view, according to which the punishment should not
be allowed to have a goal (Zweck) but rather should be inflicted only
because of his having committed a crime.4 Kant regards this requirement as a moral justification of punishment, by which he defines the
criminals worthiness of punishment. In the following, I would like to
conduct an interpretation of Kants moral retributivism that satisfies
this requirement. I shall attempt to show that although one can derive
from Kants moral theory a retributivist conception of a criminal
deserving of punishment, one can, however, justify not a retributivist
punishment in the legal system, but rather that combination of specific deterrence and rehabilitation that already ensues from the liberal
concept of right (see Section 2.5).
At this point, I would like to introduce a classification through an
additional differentiation, in which I will enumerate several possible
retributive theses. H. L. A. Hart divides the retributive position into
three theses:
First, that a person may be punished if, and only if, he has voluntarily
done something morally wrong; secondly, that his punishment must in
some way match, or be equivalent of, the wickedness of his offence; and
thirdly, that the justification for punishing men under such conditions
is that the return of suffering for moral evil voluntarily done, is itself just
or morally good.5
3. A sensible person does not punish a man because he has sinned, but in order to keep
him from sin. Seneca, De ira, i.19.7. Seneca uses as a basis Platos Laws 11.933e934b,
where it is said that every thief should not only provide compensation for the item that
was taken, but also suffer a more lenient or a severer punishment depending on his
motives: This additional penalty is to be inflicted not because of the crime (whats
done cant be undone), but for the sake of the future: we hope that the offender
himself and those that observe his punishment will either be brought to loathe injustice
unreservedly 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. 13181616 (pp. 15856). Therefore,
Platos as well as Senecas intention is clearly directed, not only toward the
compensation of the victim, but toward the reform of the criminal as well as toward
general deterrence, which is conceived of as a collective reform. Seneca, De ira, in Moral
essays, ed. John W. Basore (3 vols., London: Heinemann/Cambridge, Mass.: Harvard
University Press, 1928), vol. 1, pp. 106355 (pp. 1589).
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.
74
The first thesis does not constitute a specific characteristic of retributivism, because, in my opinion, no alternative theory of criminal
justice ever allowed the punishment of the innocent.6 The third thesis
can correspond to as many as the four following subtheses:
1 All criminals and only criminals should be punished.
2 The punishment of criminals acts as retribution for a past criminal act.
3 The degree of punishment should be proportional (in an ordinal, not
in a cardinal way) to the crime, that is, the proportion of punishments
to one another should correspond to the proportion of the crimes to
one another.
4 The degree of punishment must be equal to the crime.
The third thesis connects a legal descriptive observation (the descriptive observation of a punishable action) to a legal consequence (the
punishment of an action). In contrast, the second thesis connects a
legal consequence (the punishment of an action) to a descriptive
moral observation (the wickedness of the perpetrator). Thus, retributivisms moral justification pertains to the second thesis; however, it
can support or contradict the third thesis. In what follows, I will show
that it contradicts the third thesis. Even the point of reference is
lacking for proportionality between the moral wickedness and the
punishment. Kant distinguishes, in fact, actions in conformity with
duty from actions contrary to duty, and with the latter he distinguishes
actions that are merely in conformity with duty from actions that are
done out of it. However, neither a gradation of how much an action
stands in conformity with duty nor a gradation of contrariness to duty
can be found in Kants works. Kant merely states:
If someone does more in the way of duty than he can be constrained by
law to do, what he does is meritorious (meritum); if what he does is just
exactly what the law requires, he does what is owed (debitum); finally if what
he does is less than the law requires, it is morally culpable (demeritum).7
Virtue a is opposed to negative lack of virtue (moral weakness) 0
as its logical opposite (contradictorie oppositum); but it is opposed to vice a
as its real opposite (contrarie s. realiter oppositum).8
6. Cf. Fred Rosen, Utilitarianism and the punishment of the innocent: the origins of a
false doctrine, Utilitas, 9, no. 1 (March 1997), 2337.
8. TL Ak vi:384. Practical philosophy, ed. Gregor, p. 516. (There is a small error in the
translation that I have corrected.)
75
Indeed, the gravity of the punishable action should not be confused

with the correspondence between the intention and the duty.
Whereas a punishable action is more or less serious, any injury of duty
counts as being contrary to duty, and Kant does not differentiate any
further than that. Kant even emphasizes that an exception from the
moral law or from the fulfillment of duty should never be allowed to
be trivialized, but should be considered as really representing a complete misconduct.9 The individual instances of misconduct can be
traced back to a supreme maxim, which does not obey the categorical
imperative. Because the assertion of a gradation of evil is problematic
with Kant, I will then, in what follows, confine Kants retributive moral
theory to the assertion stating that the punishment is not imposed in
order to fulfill an aim, but instead merely because the criminal did
injury to the law.10
3.2. The highest good and the necessary

relationship between guilt and punishment
The requirement for a relationship between the moral evil of the
individual who acts contrary to duty and his or her punishment is
derived from the postulate of the highest good as it is formulated in
the Critique of practical reason. It is generally known that this postulate
requires a necessary connection between the two components of the
idea of the highest good, that is, between virtue and happiness (that is,
between Tugend and Gluckseligkeit); however, Kant further refines this
connection into a subordination of happiness beneath virtue, which
serves as a prerequisite for happiness. Kant writes:
Now, inasmuch as virtue and happiness together constitute possession
of the highest good in a person, and happiness distributed in exact
proportion 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 no
further condition above it, whereas happiness is something that,
though always pleasant to the possessor of it, is not of itself absolutely
and in all respects good but always presupposes morally lawful conduct
as 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), 60120, shows that it is also problematic in general.
11. KpV Ak v:110f. Practical philosophy, ed. Gregor, p. 229.
76
The highest good means postulating the reign of the principle of right
over the physical world, which usually abides by the laws of physics.
The highest good can only be realized under the reign of the
principle of right. Retaliation against evil, for the time being, plays
no role in this; instead, it is only the reward to the moral disposition
that plays a role. How can the relation between having wicked dispositions and the worthiness of punishment be derived using such a
definition of the highest good?
Next, we must refer to the status and the substance of the highest
good, whose negative variant is the worthiness to be punished. Being a
postulate of the practical reason, the highest good has the status of a
noumenal, synthetic and necessary connection between virtue and
happiness; in the sensible world this connection wherever it might
happen 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 modern
times could nevertheless have found happiness in precise proportion to
virtue 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 inverse
of the highest good, and therefore have the same status as the highest
good, then the relation must belong exclusively to the noumenal
world. In the Doctrine of virtue, we find an acknowledgment of this very
assumption:
Punishment is not an act that the injured party can undertake on his
own 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 to
it; and when (as we must in ethics) we regard human beings as in a
rightful condition but in accordance only with the laws of reason (not civil
laws), then no one is authorized to inflict punishment and to avenge the
wrongs sustained by them except him who is also the supreme principle
of 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 between

1775 and 1780), deterrent punishment inheres in the worldly ruler
and retributive punishment inheres in the moral ruler:
12. Cf. KpV Ak v:114f. Practical philosophy, ed. Gregor, pp. 231ff.
77
All punishments are either deterrent or retributive . . . All punishments

by authority are deterrent, either to deter the transgressor himself, or to
warn others by his example. But the punishments of a being who
chastises actions in accordance with morality are retributive.15
There are two reasons for the highest goods noumenal status. First,
for us human beings, actions done out of duty and actions that are
merely in conformity with duty can never be discerned from one
another.16 Secondly, we will never find a worldly ruler of whom we
can be completely sure that he or she will always abide by the categorical
imperative.17
One could argue against assigning this exclusively noumenal status
to the inverse of the highest good, that is, to the relation between guilt
and punishment, by saying that with an action that is contrary to the
legal law one is able to suggest wholly and with absolute certitude
the existence of wicked dispositions; and with that, one dispenses with
the first reason against a retributive punishment imposed by the
earthly authorities. The second reason still persists, however, because
there is no individual to be found, who acts out of duty, to be the
highest judge. In what follows, I would like to show that, even when
this second reason is disregarded, the earthly authority is still not
authorized to impose retributive punishments, principally for reasons
that result from the substance of the highest good.
The highest good consists of the necessary connection between
virtue and happiness, where virtue is the prerequisite for happiness.
Therefore, the converse of happiness should be derivable from the
converse of virtue, that is, from wicked dispositions. This appears prima
facie to suggest that the converse of happiness is unhappiness, in the
sense of an unhappy condition. However, this would be a fallacy,
because the logical converse of happiness is the lack of happiness as a
simple privatio. Now, Kant defines happiness in the following fashion:
Happiness is the state of a rational being in the world in the whole of
whose existence everything goes according to his wish and will, and rests,
therefore, on the harmony of nature with his whole end as well as with
the essential determining ground of his will.18
15. VE Ak xxvii:286. Immanuel Kant, Lectures on ethics, ed. Peter Heath and
J. 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. 467.
78
According to this, the converse to happiness is a rational beings

condition in the world for whom nothing in his entire existence goes
according to his wish and will, and therefore rests on the complete
lack of correspondence between nature and his aim, that is, the
essential determining ground of his will. In short, the will of the
wicked human being should never be able to realize itself. The complete lack of effect of the evil will is something other than inflicting
some sort of harm on the evil individual. In the clearest possible
fashion, Kant rejects the latter as immoral and stringently distinguishes
it from actual punishment:
It is, therefore, a duty of virtue not only to refrain from repaying
anothers enmity with hatred out of mere revenge but also not even to
call upon the judge of the world for vengeance, partly because a human
being has enough guilt of his own to be greatly in need of pardon and
partly, and indeed especially, because no punishment, no matter from
whom it comes, may be inflicted out of hatred.19
From the Kantian perspective, the punishment should be derived

from the crime itself according to a moral law that reigns over the
world. This means a necessity similar to that of the physical laws of
nature. Kant himself uses the analogy:
Before reason awoke, there were no commandments or prohibitions, so
that violations of these were also impossible . . . From the moral point of
view, therefore, the first step beyond this state was a fall; and from the
physical point of view, this fall was a punishment, for it led to a host of
hitherto unknown evils.20
Regarding crimes, this means that the moral legislator should prevent
their perpetration, and do so, in fact, before the implementation of the
criminal intention. Kant also did not ignore the fact that it does not
usually happen like that in the real world. However, one is allowed to
postulate that the necessary connection between guilt and punishment might be compensated in the noumenal world. When one has
finally determined that, despite this postulate, the intention that is
contrary to duty led to a deed that is contrary to duty, the question
arises as to why the evil intention, whose realization could not be
prevented, 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. 22134.
79
appropriate to differentiate between two cases: either the perpetrator

remains an evil person after the punishable action or he or she
converts to the side of the good.
If the perpetrator were to remain an evil person, then he or she
would want further to realize intentions that are contrary to duty. The
necessary connection must lead to the perpetrator being prevented
from realizing those intentions. This prevention occurs not for the
sake of punishing the specific crime that he or she committed, but
instead because of the evil manifested along with the crimes
perpetration.
According to the Kantian perspective, in the case of the perpetrator
converting to the side of the good since that time, one would have to
look at this case from two different angles. In Kants work Religion
within the boundaries of mere reason, he analyzes the relation of moral
conversion with punishment as follows. In the case of conversion, the
punishment takes place within a situation in which he now leads a
new life and has become a new man.21 Kant locates punishment
in the situation of conversion itself: the conversion is an exit from
evil and an entry into goodness,22 and he sees that by means of the
very concept of moral conversion, we can think that situation as
entailing 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 simultaneous
existence between the old and the new human:
As an intellectual determination, however, this conversion is not two
moral acts separated by a temporal interval but is rather a single act,
since the abandonment of evil is possible only through the good
disposition that effects the entrance into goodness, and vice-versa. The
good principle is present, therefore, just as much in the abandonment
of the evil as in the adoption of the good disposition, and the pain that
by rights accompanies the first derives entirely from the second.24
The punishment of an evil human being consists wholly of the good

human beings birth pains; in short, it is comprised of that through
which 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. Allen Wood and George
di Giovanni (Cambridge: Cambridge University Press, 1998), pp. 31192, (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.
80
par excellence, the evil human beings lack of happiness, because in the
conversion he or she breaks with the evil intentions and new, good
intentions lead to happiness, that is, to their full effect. Kant evidently
describes punishment in a clear form with similarities to the Pauline
Epistles in the New Testament:25
The emergence from the corrupted disposition into the good is in itself
already sacrifice (as the death of the old man, the crucifying of the
flesh) and entrance into a long train of lifes ills which the new human
being undertakes in the disposition of the Son of God, that is, simply for
the sake of good, yet are still fitting punishment for someone else, namely
the old human being (who, morally, is another human being).26
What appears to be moral happiness from the new point of view is

perceived from the old point of view as ill-being and punishment.27
Here, the appeal of the usual interpretation of retributive punishment
to Kant proves to be false. Thomas Hill sees a motive for future moral
behavior in the moral human beings pangs of conscience about his or
her 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 in
the phenomenal world from this moral punishment in the noumenal
world? It is appropriate, owing to the case of the human being who
continues to remain an evil human being, to use incapacitation29 as a
means so that his or her evil intentions will not come to fruition in the
future. In the case of the human being who has converted, there is no
longer any reason for punishment provided that one can discern that
there has truly been actual reform on the part of the human being.
Now, a conversion in the phenomenal world a conversion which was
completed in the noumenal world does not lend itself to being
substantiated with absolute certitude. Admittedly, one could increasingly conjecture such a conversion in the case of a continued behavior
in conformity with duty, so that, in the course of time, and in view of
Rom. 6:2, 6; Gal. 5:24; Eph. 4:1724; Col. 3:117.
Rel Ak vi:74. Religion, p. 90.
Cf. footnote in Kant, Rel Ak vi:75. Religion, p. 91.
Cf. Hill, Kant on punishment, 35860; similar to John Deith, On the right to be
punished: some doubts, Ethics, 94 (1984), 191211 (pp. 21011).
29. Cf. Arthur Ripstein, Equality, responsibility and the law (Cambridge: Cambridge
University Press, 1999), p. 144.
25.
26.
27.
28.
81
such behavior, reintegration into society becomes more and more

appropriate. In general, specific deterrence in the form of incapacitation with subsequent rehabilitation is most appropriate, because
(at least in a first period of time) one cannot discern if one is dealing
with the first or the second case when it comes to a convicted criminal.
Neither specific deterrence nor rehabilitation is at the disposal of
the public authorities, but they are their duties. On the one hand, the
duty to provide assistance to those in need in itself also encompasses
the duty to provide assistance to any person who may be a future
victim of crime. On the other hand, however, we are not allowed to
do harm or to deny freedom to any human beings who became good
in the meantime owing to their previous transgressions, if either this
harm or this denial of freedom is not necessary to public safety. At this
point, we have reached the same point that our Kantian critique of
Kants theory of criminal justice derived from the Kantian liberal
concept of right reached.
Different from the conclusion that Kant reached in the same
year (1797) in his Doctrine of right, in the Doctrine of virtue Kant came
to draw the same conclusion that we just have:
It is, therefore, a duty of human beings to be forgiving (placabilitas). But
this must not be confused with meek toleration of wrongs (mitis iniuriarum
patientia), renunciation of rigorous means (rigorosa) for preventing the
recurrence of wrongs by others; for then a human being would be
throwing away his rights and letting others trample on them, and so
would violate his duty to himself.30
The retributivist might want to celebrate half a victory because of this

result, because while it may not be the legal punishment it is
however the case that the moral punishment is conceived of retributively. As early as in the Lecture on ethics, Kant defines Retributive
punishments as those pronounced because the evil has occurred.31
Unlike deterrent punishments, moral punishments are imposed
because a sin has been committed; they are consectaria of moral transgression.32 By celebrating half a victory, however, the retributivist
overlooks the fact that Kant, even though he does not say it explicitly,
understands moral punishments to be the worthiness of punishment, whereas he understands the other punishments to be legal
31. VE 79.
32. VE 79.
82
punishments, that is, punishments in the proper sense. Kant comments 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 so
that it shall not be committed. But every crime, in addition to this
punishment, has a property of deserving to be punished, because it has
taken place. Such punishments, which must therefore necessarily follow
upon the actions, are moral in character, and are poenae vindictivae; just as
a reward follows upon a good action, not so that further good actions
should be done, but because there has been a good action done.33
Because the worthiness of punishment which is to say, the moral

punishment belongs to the noumenal world, the other theories of
punishment do not compete with this moral retributivism for a simple
reason, that is, because the other theories are merely legal theories
of punishment, which make no claim on the noumenal world. Therefore, the alternative theories do not express themselves against the
second thesis in the manner in which it was formulated by
H. L. A. Hart (see Section 3.1). They simply ignore the subject of
retribution against moral evil and merely controvert the third thesis,
which retributivism intends to substantiate with help from the second
thesis. The retributivists want to derive the third thesis from the
retributive moral thesis, which we have just examined, as a result of
which a necessary proportionality between the punishable action and
the punishment will come to exist. We have, however, just shown with
our reconstruction of Kants moral argumentation that the third
thesis 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 has
been argued in this chapter in order to refute it.
As a provisional conclusion, we are allowed to assert that neither
the liberal (see Chapter 2) nor the moral (see Chapter 3) concept
of right, which one can derive from two different interpretations of
Kant (see Chapter 1), can justify a retributive theory of punishment.
And there is more: if both paths of argumentation were to pay attention to the consistency of their deduction of punishment from the
Kantian concept of right they would then lead in our sketch of the
reconstruction of the arguments to a combination consisting of
specific deterrence and of rehabilitation which strictly excludes
33. VE 79.
83
retributivism without arguing for general deterrence. At this point, it

finally comes into full view that the retaliative theory is not the only
alternative to general deterrence, which is a view that is contrary to
both Kant and many of the authors inspired by him. Fichte and Hegel
then reconstructed what should have been Kants theory of punishment, if it had been consistent with Kants practical philosophy.
Although this is a controversial issue, I believe that none of these
authors was a retributivist.
PART II
PUNISHMENT AS A
MEANS OF REHABILITATION
I will proceed with my reconstruction of what should have been Kants

theory of penal law, had it been derived from Kants concept of right,
through an interpretation of the Fichtean and the Hegelian theory of
punishment. Since the degree of punishment that both Fichte and
Hegel support (incapacitation and resocialization, as we shall see) is
nearly the same, interpreting both authors may seem redundant. Thus
one may ask oneself whether treating the more famous of the two, who
is at the same time the one often considered as being a retributivist,
would not suffice. However, the ways in which they come to this shared
conclusion could hardly be more divergent. As we shall see, whereas
Fichte begins with the unilateral breaking of the social contract by the
criminal and with its consequences (the exclusion from the commonwealth), and then introduces a normative obligation to realize the
concept of right, Hegel starts with the validity and the enforcement of
law that has to be manifested, without drawing on the figure of
exclusion on the contrary, the criminal does not cease to belong to
the commonwealth and to enjoy by the same token respect for his or
her 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 of
punishment on the basis of the same concept of law in spite of the very
different ways of reaching it should make it clear that it is not the
method but the adoption of this shared concept of right that matters
for reconstructing a Kantian theory of punishment. Nor do the differences between Kant, Hegel and Fichte regarding the foundation of
their respective systems and of their respective conceptions of autonomy matter for our investigation. Therefore, I shall deliberately put
them aside in the following argumentation.
85
4
FICHTES EXPIATION CONTRACT
It is well known that the liberal interpreters of Kants concept of right

(see Chapter 1) and Fichte share the same concept of right, even
though they deduce it in fundamentally different ways:1 the reciprocal
restriction of the external freedoms according to the universal law of
equal rights. Nevertheless, Fichte develops a theory of penal law
clearly incompatible with that of Kant. That is why I wish not simply
to compare both of the theories in what follows, but rather now that
I have confronted Kants theory of penal law with his concept of right
in Chapters 2 and 3 also now to compare Fichtes theory with the
concept of right common to Kant and Fichte.
In Chapter 2, I have already criticized Kants retributivism, and
indeed both the classical interpretation of this retributivism as the
sole justification of the existence of punishment and the newer interpretation of this retributivism as justification of the degree of punishment. In both interpretations, Kants retributivism cannot be derived,
or deduced, from his concept of right: moreover, his retributivism
even contradicts his concept of right. That is why I will now, as a lead
in, sketch with what rationales Fichte rejects Kantian retributivism as a
justification of the degree of punishment, even though he defines the
degree of punishment according to the criterion of talion law.
Fichte published the first part (}} 116) of his Foundations of natural
right in 1796, half a year before the publication of Kants Doctrine
of right. Half a year after the publication of Kants Doctrine of right,
the second part (}} 1724) of Fichtes Foundations of natural right
appeared, 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. 15972 (p. 161).
87
88
punishment as a means of rehabilitation
right about the right of coercion (}} 13ff.), Fichte developed this
aspect of the Kantian mixed theory, which was then neglected, but
since the 1980s it has again been receiving attention: general deterrence through threatening punishment. Fichtes actual theory, as it
was formulated in 1797 in } 20 of the Foundations of natural right,
seems, however, to be a mixed theory: Fichte interpreters have shown
that in the Foundations of natural right a combination of elements of
general deterrence through threatening punishment, elements of specific deterrence and elements of reform and of rehabilitation can
be found. In this respect, Kaufmann rightly sees a commonality
between the Kantian penal law interpreted as mixed theory and
Fichtes 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 overwhelmingly
accepted nowadays of the connection between the theories of deterrence
and retribution. The justification of punishment, on the whole, occurs
through deterrence; in the determination of who should be punished,
not only the damage plays a role, but also the extent to which a bad or
an inadequate will played a role. The same is valid for determining the
degree of punishment.2
Likewise, Lazzari observes that both pillars of the Fichtean theory

of punishment, the ideas of performing expiation and of deterrence,
are widely in harmony with the diffuse requirement of a more humane
system of punishment. Lazzari praises the mixed character of the
Fichtean theory of punishment, even though he is aware of the problems
connected with it:
A mixed attempt of that sort appears in many respects really
illuminating and befitting the multiplicity of the aspects having to be
taken into account. What is problematic in Fichtes argument is, first and
foremost, the incompatibility of the legal and theoretical assumptions,
which substantiate the deterrent part of Fichtes theory of punishment in
the first part of the Foundations of natural right and the restricting
framework in } 20.3
The objection that I raised earlier against the Kant interpretation as

a mixed theory, however, holds ipso facto against Fichtes 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. 17386 (pp. 1834).
fichtes expiation contract
89
deterrence through threatening punishment is not compatible with

the Fichtean or Kantian concept of right. Thus, it is no coincidence that one year later, in the actual chapter in the Foundations of
natural right about penal law (} 20), Fichte contradicts his chapter on
the right of coercion and, in my opinion, justifies punishment primarily with specific deterrence, even though he never explicitly renounces
his earlier mixed theory.
4.1. Retributivisms lack of justification

from the legal perspective
I will now briefly remind the reader of Kants argumentation in favor
of a retributivist grounding of penal law. In } 49e of the Doctrine of
right, Kant advocates a retributivist position, more precisely talion law:
like for like.4 Kant delivers no direct justification of talion law, but
merely claims that there are no acceptable alternatives.5 An alternative
can be acceptable only if it passes the test of the categorical imperative,
especially its third formula: So act that you use your humanity, whether in
your 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 the
purposes of another or be put among the objects of rights to things:
his innate personality protects him from this, even though he can be
condemned to lose his civil personality.7
Kant declares the retributivism of talion law as being the only theory of
penal law that passes the test of this formula; all other theories of
penal law treat the criminal merely as a means. Now, not only do the
individual punishments determined by the retributivist talion law
clearly infringe the aforementioned formula of the categorical
imperative,8 but Kants negative argumentation in favor of retributivism also clearly applies to that theory, which regards general deterrence whether by threatening punishment or by an example carried
out as being the primary justification of punishment, while leaving,
in my opinion, the thesis of specific deterrence untouched.9 It is
4.
5.
6.
7.
8.
9.
RL Ak vi:332.
See Section 2.2.
RL Ak vi:331.
See Section 2.4.
See Section 2.5.
90
exactly this neglected option that Fichte reaches as the result of his
deduction of penal law.
Fichte proceeds under the assumption of the same concept of
right: Every relation of right is determined by this proposition: each
person is to limit his freedom through the possibility of the others
freedom.10 He obviously regards Kants negative argumentation in
favor of retributivism as a crude fallacy. In the second part of his
Foundations of natural right, Fichte criticizes with sharp language Kants
Doctrine of right, which had appeared nine months earlier:
Punishment is not an absolute end. The claim that it is (whether stated
explicitly or through propositions that implicitly presuppose such a
premise, e.g. the unmodified, categorical proposition that he who
has killed, must die) makes no sense.11
Some pages later, regarding the claims he is attacking, Fichte refers

explicitly to Kant, the great, though not infallible, man12 who
accuses Beccaria of alleged overly compassionate (Empfindelei) feelings of an affected humanity.13 Fichte turns against
an absolute right of punishment, according to which judicial punishment
is regarded not as a means, but as itself an end, which is said to be
grounded on a categorical imperative that is itself not further
examinable . . . By relying on what is supposedly unexaminable, this
theory allows its proponents to exempt themselves from the need to
prove their claims and so to charge those who think differently with
sentimentality (Empfindelei) and an affected humanitarianism.14
Thereby, Fichte does not wish to contest, in any way, that retaliation is
a principle of justice. In the second edition of his Attempt at a critique of
all revelation (1793), Fichte definitely sees a motivation to moral action
in the morally necessary proportionality between virtue and happiness,
that is to say, in God as the infinite rational being whom Fichte declares
as just.15 In the Foundations of natural right, Fichte claims furthermore:
10. GNR i/3 411. Johann Gottlieb Fichte, Foundations of natural right: Grundlagen des
Naturrechts 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.
15. Cf. Jean-Christophe Merle, 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 Aldo Masullo
91
There is absolutely no dispute about whether a murderer has been

treated unjustly, if he, too, should lose his own life in a violent
manner.16 However Fichte immediately stresses that the question of
penal 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 purely
juridical question that the noble Beccaria (who was certainly not
unfamiliar with that kind of moral judgment) had in mind. Whoever
ascribes this right to a worldly sovereign will surely be required . . . to
regard . . . 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 (}} 1316), Fichte identifies this with the right of coercion exercised by the state. Now, such an
equation is by no means self-evident, as a comparison with Kants right
of coercion and penal law shows.
Numerous Kant interpreters proceed under the assumption, which
they consider as obvious, that in Kants thought penal law naturally
and directly belongs to the right of coercion, even though Kant
neither claims nor suggests this interpretation. Next, let us investigate
what exactly the right of coercion consists of for Kant. The right of
coercion exercised by the state is derived and deduced from that very
right of coercion inhering in every person as soon as his or her rights
are infringed. This individual right of coercion entitles the battered
person to curtail the freedom of a fellow human being, because,
according to Kant,
Whatever is wrong is a hindrance or resistance to freedom in accordance
with universal laws . . . Therefore, if a certain use of freedom is itself a
hindrance to freedom in accordance with universal laws (i.e., wrong),
coercion that is opposed to this . . . is consistent with freedom in
accordance with universal laws, that is, it is right.18
The consummate example of this individual right to coercion is the

right 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. 30325.
92
the equivalent of the individual right to self-defense when seeking to

prevent the commission of an infringement of the law by means of
police at the actual moment when the violation of rights is still just an
attempt. The state is also authorized to exercise a right of coercion in
the forms of nullification or annulment, seizures, and the revocations
of licenses (withdrawing a concession or a privilege, in the Hohfeldian
meaning, for example nowadays a driving license). Apart from that,
the state is endowed with a right of coercion after the commission of a
rights infringement (as with the forced restitution of stolen goods)
and with the court-ordered delivery of compensation for damages, if
this is possible (with some rights infringements typically with murder
there is no fitting substitute restitution). These cases concern retribution in the narrow sense. Where it is possible (with some rights
infringements typically, once again, with murder it is fully impossible)19 this retribution completely satisfies the legitimate individual
claims of the victim of a rights infringement. It is in no case a punishment. Whereas Kant presents the right of coercion immediately in his
Introduction to the doctrine of right20 and links each right with a
right of coercion, he first treats penal law only much later and merely
as one of the elements of public law.21
Fichte diverges from Kant not only in introducing penal law as early
as in the deduction of the right of coercion in general, but also in his
definition of the right of coercion as well as in the end he ascribes to
it. Whereas the Kantian right of coercion merely resists an existing
hindrance to freedom in accordance with universal laws, coercion,
according to Fichtes chapter about the right of coercion, already
prevents the formation of such a hindrance even before the attempt:
Now if things could be arranged so that the willing of any unrightful
end would necessarily, and in accordance with an ever-operative law,
result in the opposite of what was intended, then any will that is contrary
to right would annihilate itself . . . It was necessary to present this
proposition in its full, synthetic rigor, since all laws of coercion, or penal
laws, (the entirety of penal legislation) are grounded on it.22
According to the respective definitions of the right of coercion

delivered by Kant and Fichte, the right of coercion appears for Kant
19.
20.
21.
22.
See Section 2.3.

RL Ak vi:3317. Practical philosophy, ed. Gregor, pp. 4728.
GNR i/3 426. Foundations of natural right, p. 126.
93
initially in the perpetration of a rights violation, while for Fichte, on

the other hand, it occurs already, in its very essence, in the intent of
perpetration. Fichte states clearly:
The security of the two parties is not supposed to depend on a
contingency, but on a near-mechanical necessity that excludes every
possible exception. There can be such security only if the law of right
is the inviolable law of both parties wills.23
Incidentally, it is worth noticing that, in Kants justification of

punishment, deterrence through threatening punishment plays an
explicit role only in the passage about the right of necessity, and that
it plays this role only in an indirect and negative way; therefore, with
Fichte, 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 paragraph
about penal law Fichte speaks no longer about the threat of
punishment, but about the punishment itself,24 which presupposes
that deterrence by threatening punishment has failed at least
partly: the crime has already taken place. In view of the demanding,
declared goal of the general deterrent threat of punishment The
threat of punishment aims to suppress bad wills . . . in which case
punishment will never be necessary25 the failure is obvious. Either
the humbler task of general deterrence through example must inhere
in the punishment of a criminal or it must have another end. In the
first case, and in the absence of any other goal, the criminal would be
treated as a mere means for general deterrence, which is, however,
forbidden.26 Therefore, another primary justification of punishment
must be searched for, which would in turn substantiate the normative
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 of
will, in a meaning which remains to be defined more precisely. The threat of
punishment 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 coercion
in general, instead of its being a right of coercion meant for special cases and being a
special form of the right of coercion.
25. GNR i/4 60f. Foundations of natural right, p. 228.
26. See Section 4.1.
94
acceptability of deterrence by threatening punishment indeed no

longer 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 the
thesis and the antithesis of that deduction leading to the expiation
contract (Abbuungsvertrag) will notice, first, that in the thesis there is
no indication of the punishment to be found, and, secondly, that the
antithesis no longer draws on the same concept of coercion as the
earlier chapters about the right of coercion.
The thesis depicts the crime as a breach in the civil contract on the
part of the criminal, who excludes him- or herself ipso facto from
society. Even though the Antithesis, in the strict sense of the term
by which I mean the paragraph that begins with Antithesis, not the
following paragraph dedicated to the synthesis refers to these
punishments in contrast to other punishments,27 it obviously does
not concern a punishment in the legal sense. This becomes clear as
soon as one reads the passage about the destiny of the criminal, who can
be considered to be incorrigible and in whose case the thesis of being
excluded from the commonwealth remains valid. Fichte stresses that
the criminals death is not a form of punishment, but only a means to
ensure security. This gives us the entire theory of the death penalty. The
state as such, as judge, does not kill the criminal; it simply cancels the
contract with him . . . If, afterwards, the state also kills the criminal, then it
does so not by virtue of its judicial authority, but through the police. This
takes place, not in consequence of any positive right, but out of necessity.28
The substance of the synthesis begins with the sentence: This can be
arranged only through a contract of all with all,29 even though there
is, from the start of the Antithesis to the end of the chapter, no
heading for Synthesis. The synthesis immediately and explicitly
concerns a punishment in the legal sense. Yet, public security is
not to be reached anymore by the right of coercion, in the sense of
the chapter about the right of coercion, that is, by a 100 percent
effective right of coercion in the form of general deterrence through
threatening punishment.30 The synthesis rather deals, on the one
hand, with specific deterrence both by the criminals incapacitation
27.
28.
29.
30.
GNR i/4 59f. Foundations of natural right, p. 227.

95
(the convicts live on their own island) and by the threat of irrevocable
exclusion from the commonwealth in the absence of any reform
and, on the other hand, with general deterrence by example:
The [penal] laws first aim was to prevent the criminal from committing
a crime. Since this goal was not achieved, the states punishment of
the criminal serves another purpose: to prevent other citizens, and to
prevent the criminal in the future, from committing the same offense.31
If general deterrence whether through threatening punishment or

example were the end of the punishment, then it would be better
served by sentencing people to a death in the desert or summary
execution than by a prison sentence or even by forced labor. Because
Fichte provides for the possibility of reform and the rehabilitation of
the criminal, criminals will possibly be released one day. In view of
public security and general deterrence either by threatening punishment or by making an example of the criminal, every instance of
recidivism means that the punishment recommended by Fichte
appears to be less effective than exclusion from society, which,
according to Fichte, equally leads either to the death of the criminal
in the desert or to shooting the criminal like a wild animal. In the
Antithesis, Fichte makes it clear that the prerequisite for legal punishment to replace the exclusion of the criminal from the commonwealth is that by legal punishment the mutual security of the rights
of all in relation to all others32 is met as effectively as through
exclusion from the commonwealth. By the acceptance of a possible
incidence of recidivism, the conclusion can be drawn that public
security which in the chapter about the right of coercion still
demanded the almost mechanical, full prevention of every crime even
before the stage when it is attempted is now understood only in a
much 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 coercion
admittedly finds a similar inconsistency already there. Fichte determines a degree of punishment for general deterrence through the
threat of punishment that is in no way optimal for that goal: ius
talionis, which is supposed to countervail the criminals impetus for
crime (the opposite to A A being the criminal goal so that the

96
opposite to A means the counterpoise).33 In } 20, Fichte explicitly

refers back to } 14, in order to recommend that the punishment must
be equal to the offense: poena talionis. According to him, it should be
the material principle of positive punishment within a state, in short,
a criterion. Fichte does not fortify his claim with any argument other
than the vague image of a counterpoise, that is, of the pans of a scale
balancing that both general deterrence, whether by the threat of
punishment or by example, and specific deterrence are best achieved
by way of talion law, that is, by a retributivist degree of punishment.
Therefore, the conflict between the different theories of penal law
pertains, allegedly, only to the grounds for punishment, not to its
degree. Nevertheless, Fichte emphasizes that exclusion from the state
is the most terrible fate a human being can encounter;34 he even
rejects any form of torture before killing the criminal.35 The objection
could be raised that killing the criminal without any court sentence,
that is, as a mere police measure, provided that it is complemented by
torture before the homicide and by making public both the torture
and the killing, might deter even in cases in which talion law does not
suffice to curb the criminal impetus. In his Antithesis, Fichte himself
mentions that talion law is a much milder punishment than exclusion
from the commonwealth.36 If one were to proceed under the assumption of Fichtes psychological premise that the more severe the punishment is, the greater the criminal impetus drops off, then it follows
that exclusion from the commonwealth is a more efficient deterrent
than ius talionis. Nonetheless, Fichte ultimately rejects exclusion from
the commonwealth. Moreover, not once does he attempt to show that
this exclusion from the commonwealth would be ultimately linked to
a loss of deterrence.
Only because of Fichtes unsubstantiated and implausible assumption that general deterrence is best reached by ius talionis do we have
a paradox before us: legal punishment (retaliation as in talion law)
meant to serve general deterrence by the threat of punishment is
33.
34.
35.
36.
GNR i/4 426f. Foundations of natural right, p. 229.

Cf. GNR i/4 74. Foundations of natural right, pp. 2423.
One could argue that exclusion from the commonwealth sets a limit for the
punishment determined by general deterrence. It is indisputable that exclusion from
the commonwealth is mostly less painful than torture before execution. Nevertheless, in
most cases, exclusion from the commonwealth, which leads to a sure death in the
desert, remains a worse fate than the degree of punishment recommended by Fichtes
talion law.
97
milder than the lack of a legal punishment, that is to say, milder than
extralegal punishment or exclusion from the commonwealth. Therefore, whoever compares the legal punishment to the lack of a legal
punishment will come to see that the legal punishment does not imply
punishment in the sense of the infliction of suffering due to the
crime, but instead a mitigation of extralegal punishment. Under these
circumstances, Fichte should have argued for the weaker antithesis
according to which legal punishment, despite its lesser deterrent
effect, does not lack every deterrent effect, as it would in the case in
which the criminal were further to enjoy the status of a normal citizen,
that is, impunity. With such an antithesis, the legal punishment recommended by Fichte must exert a not insignificant deterrent effect.
However, because this legal punishment in no way optimizes general
deterrence, it is impossible for general deterrence to be punishments
primary end. If at all the case, general deterrence would rather be the
primary end of an extralegal punishment (in my opinion, it is not
even that, because exclusion from the commonwealth has a reason
the criminal is a potential danger for the commonwealth rather than
a general deterrent end). In this respect, Fichte is not able to make the
claim anymore that, since exclusion from the commonwealth is not
as effective a deterrent as legal punishment, there is in such cases . . .
no reason to exclude the offender; but admittedly . . . there would also
be no reason not to exclude him. The decision would be a matter of
free choice.37
After Fichte has in this way wrongly assumed the parity of legal and
extralegal 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 what
directly arises from the combination of the concept of right and the
concept of infringement of right. The result of this investigation
should actually not deliver merely a subsidiary criterion for the
appraisal of competing notions of degrees of punishment, but instead
the primary grounding of punishment itself. But, at first glance, this
option is not to be found. Now, the logical consequence of the reciprocal limitation of freedoms contained in the Fichtean concept of
right combined with the infringement of these limits would be the
complete exemption from legal punishment, that is, the unadorned
98
exclusion from the commonwealth. It would primarily be concerned

merely with the consequence of the infringement of rights, that is,
with the elimination of an obstruction of rights, and not be concerned
with goal-oriented measures, that is, with the prevention of obstruction of rights that has not yet occurred.
Instead of this, Fichte adopts the reciprocal usefulness of the
criminal and of the state to one another as a subsidiary criterion. It
results from this criterion that the criminal, despite the crime, continues to be a member of the commonwealth. According to the
subsidiary nature of this criterion, general deterrence by threatening
punishment and by example remains, throughout the entire text of
} 20 of the Foundations of natural right, the primary end;38 rehabilitation constitutes only a secondary justification of punishment.
Now, Fichte introduced the subsidiary criterion only because of its
false premise that exclusion from the commonwealth extralegal
punishment and legal punishment out of consideration for deterrence are equally efficient. Because this premise is false, Fichte must
have given up, de facto and tacitly, general deterrence as being the
primary end of penal law that would require exclusion from the
commonwealth. Admittedly, Fichte refers to the principle of general
deterrence not only at the beginning, but also in the entire text of
} 20. Nevertheless, he does not argue in favor of this principle, and
does not derive any consequences from it regarding the degree of
punishment. Therefore, we must draw the following conclusion: what
Fichte introduces as being a subsidiary criterion constitutes in reality
the single true, primary justification of legal punishment in } 20. It is
now time to investigate this justification.
4.4. Specific deterrence and rehabilitation

as justification of legal punishment
Fichte justifies only very briefly why he opts for an alternative to the
exclusion of the criminal from the commonwealth. His entire
grounding can be summarized in two citations:
It is just as much in the states interest to preserve its citizens . . . as it is
in each individuals interest not to suffer the loss of all rights for every
single offense.39
99
This expiation contract is useful for all (for the state as a whole) as well as
for each individual citizen. Under it, the whole obtains both the prospect
of preserving citizens whose usefulness outweighs their harmfulness, as
well as the obligation to accept their expiation; the individual citizen
obtains the perfect right to demand that some expiation be accepted in
place of the more severe punishment that he deserves.40
Whereas the interest of the citizens of a state to be legally instead of

extralegally punished is easily comprehensible, the usefulness of the
criminal to society deserves at least some explanation.
It would be a mistake to suspect that there is a utilitarian element in
this usefulness. It would be in equal measure wrong to believe that this
appeal to usefulness does not represent any legal rationale and that
the legal concept of right remains unaffected in the exclusion of the
criminal from the commonwealth only for the reason that the commonwealth persists amongst the other, innocent citizens of the state.
Already in } 4, two instances of deduction of the concept of right attest
to the fact that, for Fichte, exclusion from the commonwealth, owing
to the denial of mutual recognition, cannot be a final and absolute
exclusion. Exclusion should rather remain temporary for the duration
of 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 to
the recognition that has occurred and must judge him in accordance
with it . . . Assuming that his action is indeed determined by the sensible
predicates of his prior actions . . . but not determined by his having
recognized 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 an
object . . . Thus, in this case, I am able, with perfect consistency (which is
my only law here), to treat him as a merely sensible being, until both
sensibility and rationality are once again united in the concept of his
action.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 the
present case . . . But, in so far as I appeal to that common law in my
opposition to him, I invite him to be a judge along with me; and I demand
that in this case he must find my action against him consistent and must
40. GNR i/4 60. Foundations of natural right, pp. 2278.
41. GNR i/3 3556. Foundations of natural right, pp. 456.
100
approve of it, compelled by the laws of thought . . . The source of this

obligation is certainly not the moral law: rather it is the law of thought;
and what emerges here is the syllogisms practical validity.42
In the lines that follow that (heading C), Fichte universalizes this
requirement for consistent behavior: it holds true for relations to
every individual.
2. In } 4, Fichte in general attempts to show what his headings
already clearly formulate: (i) I can expect a particular rational being to
recognize me as a rational being, only if I myself treat him as one.43 (ii) But
in every possible case, I must expect that all rational beings outside me recognize
me as a rational being.44 (III) . . . I must in all cases recognize the free being
outside me as a free being, i.e. I must limit my freedom through the concept of
the possibility of his freedom.45 I will not address the reasons why,
according to Fichte, the satisfaction of these requirements is a prerequisite of the possibility of self-consciousness in finite, rational
beings; even less will I treat the question here as to whether Fichte
was right about this issue. Here I only have to observe that, according
to Fichte, the requirement that every finite, rational being should
belong to the commonwealth and be treated as a member of it is a
component of the concept of right, so that the decision about who
belongs to the commonwealth is not even at the disposal of the
commonwealth itself. The laws of thought demand that every finite,
rational being is recognized as a member of the commonwealth, so
long as this being recognizes other rational beings and treats them as
such, or as soon as the being is again ready to do so after having denied
this recognition for a time to others.
Thus Fichte makes it clear in } 4 that exclusion from the commonwealth cannot be final. It is merely a suspension that lasts as long as it
takes for the denier-of-recognition to recognize others. It becomes
clear that Fichtes use of the word usefulness is absolutely not
applied in the utilitarian sense. Fichtes background is neither the
Scottish school nor the utilitarianism of the eighteenth century, but
instead modern natural right, which is inspired more from Cicero
than from Aristotle. According to Cicero, that same usefulness (utilitas), which is not a merely apparent usefulness, is defined as being
42.
43.
44.
45.

101
nothing other than that which contributes to right in the sense of a

just ordering of the community.46 Fichtes following explanation
should 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 each
individual citizen. Under it, the whole obtains . . . the prospect of
preserving citizens whose usefulness outweighs their harmfulness.47
Usefulness for state and citizens consists of the punishment contributing to what is just, that is, contributing to the establishment of a
legal system. If the state of law and the state of nature amount to a
dichotomy without a third option, which is the case with Kant and also
with Fichte, then the crime poses a transition from a state of law to a
state of nature between the criminal and other people. The concepts
of right and usefulness require, however, that all humans are part of
the commonwealth. The usefulness of legal punishment (in the twofold meaning of the word usefulness: in both the ancient and the
modern senses of the term) therefore consists in reestablishing
the state of law between the criminal and the community, and, indeed,
as quickly as possible. The concept of right requires simultaneously
the suspension and the fastest possible rehabilitation of the criminal
as a member of the commonwealth, provided that he or she is again
just as able as others to adhere to the laws.
That is why Fichte quickly abandons ius talionis, which he had
initially applied for the degree of punishment, in favor of specific
deterrence and the theory of reform. At the beginning of } 20, Fichte
still 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 that
which is labelled beneficial. Custom has stumbled over this word and strayed from
the path, gradually sinking to the point where she has severed honourableness from
benefit, decreeing that something can be honourable which is not beneficial, and
beneficial which is not honourable. Nothing more destructive than this custom could
have been introduced into human life . . . For they [the philosophers] hold that
whatever 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 do
not see this clearly often admire shrewd and crafty men and mistake wickedness for
wisdom. Theirs is an error that must be uprooted; and their fancy must be wholly
converted to that hope which consists of the understanding that they will achieve what
they 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:
102
Four pages later, Fichte says of the criminal that, He must forfeit his
freedom until it is clear that he has reformed; or else he must be
excluded 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 according
to 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 persons
have 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 must
be free. Therefore, a chief maxim is: such people must be free within
necessary limits and must live in society among themselves.50
Therefore, what remains of this Thesis is the exclusion from the

commonwealth, which is, however, still only a provisional exclusion.
What remains of the Antithesis is the criminals affiliation to the
commonwealth, which, however, is provisionally suspended and only
fully enters into force after the criminals reform. The synthesis is the
expiation contract. The expiation contract consists in offering a
chance to the convicted criminal to escape exclusion from the commonwealth and from that status as a human being who, like a savage
outlaw, can be freely shot like a harmful, wild animal. Instead of being
excluded, the criminal can decide to be punished by spending time in
a prison or upon an island with other convicts. Depending on the
criminals behavior in the penal camp, after a specific period of time
he or she will then be either readmitted into the commonwealth or
permanently excluded from it.51 Fichtes contract of expiation entails
a double originality: first, it is not a contract between the ordinary
citizens designed to legitimate the states 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, Fichtes conception diverges, for example, from
the non-retributive Kantian approach to punishment based on an
appeal to a hypothetical social contract.52
49.
50.
51.
52.

Michael Clark, A non-retributive approach to punishment, Ratio, 17, no. 1 (2004),
1227.
103
What soon proves to be Fichtes justification of penal law (specific

deterrence and rehabilitation) takes on two essential elements of
Kants philosophy of right. First, the decision about who shall belong
to the commonwealth is a decision that does not fall to the commonwealth, because every human should belong to it. Secondly, the state
of law should be reached by the shortest possible path; detours
are allowed and dictated only along this path (Kants permissive
law).53
4.5. Is general deterrence really necessary for

a justification of punishment?
Naturally, the objection can be raised against my interpretation of } 20
that nowhere in this entire paragraph does Fichte explicitly abandon
the justification of penal law with that sort of deterrence that consists
of threatening punishment. Nevertheless, in the course of } 20, Fichte
abandons general deterrence and talion law; in the end, the criminals
reform and not ius talionis determines the date of release. This also
means that sometimes the punishment should be harsher than what
retaliation would require:
It would be very prudent if the criminal himself were allowed to
determine, in accordance with the degree of his depravity, the length
of time of his reform but with the proviso that he would later be free to
extend it in accordance with a certain standard. But each criminal must
be given a peremptory term for reform, in accordance with his
particular crime.54
After this term expires, the criminal should either in the case of
reform be reassumed into the commonwealth, or in the case that
reform did not occur be permanently excluded from it.55
As a result of the occurrence of multiple justifications of penal law
by Fichte, which exclude each other from being primary justifications
53. Cf. Section 2.4 above and Reinhard Brandt, Das Erlaubnisgesetz, oder: Vernunft und
Geschichte in Kants Rechtslehre, in Brandt (ed.), Rechtsphilosophie der Aufklarung
(Berlin: De Gruyter, 1982), pp. 23385.
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 person
does not pose any danger for instance, for the prison guards. Now, we cannot
exclude the possibility that a criminal, after a certain period of time, still has not
reformed, but without having posed an additional security problem ipso facto by
continuing to remain in the correctional facility.
104
and also cannot be combined into a mixed theory because they lead to
different notions of degree of punishment, one must either declare
Fichtes penal law as being hopelessly contradictory or investigate the
individual arguments, as well as their role and relevance in Foundations
of 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 false
premises, according to which legal and extralegal punishments would
have an equally deterrent effect, then only Fichtes argument for specific deterrence by incapacitation would remain as a primary justification for penal law. Secondly, that argument is the only one that is
compatible with the degree of punishment, which Fichte conclusively
adopts at the end of } 20, for this reason. That is why I am of the opinion
that Fichte admittedly only after lengthy comings and goings
substantiates penal law with specific deterrence and rehabilitation.
Naturally, legal punishment exerts a deterrent effect by being
openly disseminated (general deterrence by threatening punishment)
and by the execution of punishment not being a secret (general
deterrence by example). This effect certainly contributes to public
security and, for that reason, is thoroughly welcomed by the commonwealth. Nevertheless, threatening punishment and public execution
of punishment are only then allowed if the infliction of punishment is
itself justified, because punishment exceeds the reciprocal limitation
of freedoms contained in the concept of right. If general deterrence
by threatening punishment were 100 percent effective and the punishment, ipso facto, never needed to be executed as Fichte wishes
in the chapter about the right of coercion then the infliction of
punishment would never be justified; then, however, the whole of } 20
of the Foundations would be superfluous.
The chapter about the right of coercion clearly stipulated a 100
percent effective general deterrence by threatening punishment: the
security of any two given persons
is not supposed to depend on a contingency, but on a near-mechanical
necessity that excludes every possible exception. There can be such
security 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, is
potentially infinite; it is always able to transgress limits and even
105
normatively inviolable limits as well. Solely the laws of nature are

unable to be infringed. Coercion is surely an effective yet not 100
percent effective means to obtain public security. As Hegel later wrote,
Only he who wills to be coerced can be coerced into anything.57 Many
a culprit accepts the risk of punishment, even if he or she later regrets
it because of fear and may come to be aware of the fact that the
punishment actually was too high a price to pay. Therefore, not only
threatening punishment but also inflicting punishment needs to be
justified. The threat is not first in line for justification.
Hegel attains specifications of penal law, similar to Fichtes, without
needing to take a detour via general deterrence and via its shortcomings. Furthermore, Hegel proceeds under the assumption of the
necessary validity of right and, therefore, does not need the thought
experiment of exclusion from the commonwealth. In this respect,
Hegel offers a systematically more stringent representation of the
same view about penal law, as I would like to show in the following.
An enlightening comparison to the conceptions of Fichte and
Hegel is offered by Karl Christian Friedrich Krause (17811832) in
his philosophy of law, which is based upon similar Kantianidealist
foundations and was published in the period between the publication
of both of their works. Admittedly, only the first section (1803) and
the posthumous second section about cosmopolitan law (1890) of
Krauses own Grundlage des Naturrechts (The foundation of natural right)
have been published, whereas the planned section about public law
and a subsection about penal law within this section remained stuck in
the planning stages.58 But Krauses Latin habilitation thesis of 1802,
which he translated into German and upon which he made comments,59 provides an informative sketch of the philosophy of law.
According to Krause, penal law should cause it to appear to the citizen
that 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 ihr
geistesgeschichtlicher Hintergrund (Ebelsbach: Aktiv Druck und Verlag, 2000), pp. 21016.
59. K. C. F. Krause, Abhandlung uber die Idee und die Eintheilung der Philosophie und der
Mathematik und den innigen Zusammenhang beider, in Paul Hohlfield and August
Wunsche (eds.), Philosophische Abhandlungen aus dem handschriftlichen Nachlasse (Leipzig:
Schulze, 1889), pp. 540.
60. Krause, Abhandlung uber die Idee, pp. 345.
106
reforming the criminal thoroughly and in a rational way, and toward

also ensuring that punishment does not affect his other rights.61
Krause remarkably proceeds in the determination of punishment
from the assumption of the inalienable rights of the perpetrator,
which, as a formula, indicates at the same time the limits of punishment.62 Krause clearly breaks away from retributivism. As the other
published parts of Grundlage des Naturrechts63 as well as unpublished
references in his letters show, Krause derives from that axiom,
amongst others, the absolute inviolability of the body, and, along with
that, a proscription of death as well as of all corporal punishment.64
As in other areas,65 the visionary Krause ironically comes somewhat
closer in the consequences which he draws from his theory to the
conclusions which one should expect from a liberal interpretation of
the Kantian idealist concept of right than the more famous authors
Fichte and Hegel.66
Krause, Abhandlung uber die Idee, pp. 345 (footnote).
Krause, Abhandlung uber die Idee, p. 35.
Part i in Krause, Abhandlung uber die Idee, pp. 2378.
Cf. Forster, Krauses fruhe Rechtsphilosophie, pp. 1089, 303.
For instance, in the theory of marriage, in K. C. F. Krause, Grundlage des Naturrechts oder
philosophischer Grundri 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
(17811832): Studien zu seiner Philosophie und zum Krausismo (Hamburg: Meiner,
1985), pp. 8092.
61.
62.
63.
64.
65.
5
HEGELS NEGATION OF CRIME
5.1. The controversial classification of the

Hegelian theory of punishment
Unlike in the secondary literature about Kants theory of punishment,
which for a long time served as the model for all retributivist or
absolutist theories of penal law, the divergence from the traditional
and often held as being self-evident1 retributivist interpretation
has quite a long history with regard to the interpreters of Hegels
penal law. Whereas the interpretation of Kants penal law as being a
mixed theory has only come about since the 1980s,2 Hegels penal
law was very early on regarded by Christian Reinhard Kostlin as a
mixed theory,3 by British Neo-Hegelianism (for instance by Bernard
Bosanquet)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. Haldane
and J. Kemp, seventh edition (4 vols., London: Kegan Paul, Trench, Trubner, 1909?),
vol. 1, pp. 43052. Nowadays, cf., for instance, Ulrich Klug, Abschied von Kant und
Hegel, in Jurgen Baumann (ed.), Programm fur ein neues Strafgesetzbuch: der AlternativEntwurf der Strafrechtslehrer (Frankfurt a.M.: Fischer, 1968), pp. 3641; 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. 2834; Jean-Claude Wolf, Verhutung oder Vergeltung? Einfuhrung in ethische
Straftheorien (Freiburg i.Br.: Alber, 1992), p. 50; Otfried Hoffe, Gerechtigkeit: eine
politische 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 seine
Strafrechtstheorie, 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 University
Press, 1918), chapter 5.
107
108
as being a reform theory. Admittedly, this last interpretation has not

prevailed. Even today, there is controversy about whether Hegel is
really a retributivist6 or whether he possibly also adopts general or
specific deterrent elements,7 or whether at the opposite extreme, he is
only a theorist of pure deterrence.8
Just as in Kantian studies, at this time mixed theories are also
predominant in Hegelian studies. Some of the best interpreters
Allen Wood and Georg Mohr, for instance combine the retributivist
interpretation with that of deterrence. According to Mohr, Hegels
theory of punishment delivers a type of grounding, which was new for
that time, for the theory of retribution, which was up to then rather
poor in its arguments. According to Mohr, Hegel is able to attain this
through allotting to the punishment as its inherent goal the task of a
positive general deterrence.9 Likewise, Wood initially contends that
6. Cf. especially Mark Tunick, Hegels political philosophy: interpreting the practice of legal
punishment (Princeton: Princeton University Press, 1992); in addition, for instance,
Ossip Kurt Flechtheim, Die Funktion der Strafe in der Rechtstheorie Hegels, in
Flechtheim, Von Hegel zu Kelsen: rechtstheoretische Aufsatze (Berlin: Duncker and
Humblot, 1963), pp. 920 (p. 13); Flechtheim, Zur Kritik der Hegelschen
Straftrechtstheorie, Archiv fur Rechts- und Sozialphilosophie, 54 (1968), 53948 (p. 542);
Peter G. Stillman, Hegels idea of punishment, Journal of the History of Philosophy, 14
(1976), 16982 (p. 173); Klaus Scala, Hegels Begriff der Strafe und die moderne
Strafvollzugsproblematik, Hegel-Jahrbuch (1987), 16470; Igor Primorac, Punishment as
the criminal right, Hegel-Studien, 15 (1980), 18798 (p. 193); and Igor Primoratz (these
are two spellings of the same name), Banquos Geist: Hegels Theorie der Strafe, HegelStudien, supplement 29 (1986), 53; Kurt Seelmann, Wechselseitige Anerkennung und
Unrecht: Strafe als Postulat der Gerechtigkeit?, Archiv fur Rechts- und Sozialphilosophie, 79,
no. 2 (1993), 22836; Seelmann, Versuch einer Legitimation von Strafe durch das
Argument selbstwiederspruchlichen Verhaltens des Straftaters, Jahrbuch fur Recht und
Ethik, 1 (1993), 31526; Matthias Kaufmann, Rechtsphilosophie (Freiburg i.Br.: K. Alber,
1996), p. 308; footnotes in G. W. F. Hegel, Principes de la philosophie du droit, trans. JeanFrancois Kervegan (Paris: Presses Universitaires de France, 1998), pp. 1801; Arthur
Ripstein, Equality, responsibility and the law (Cambridge: Cambridge University Press,
1999), p. 93; Otfried Hoffe,Konigliche Volker: zu Kants kosmopolitischer Rechts- und
Friedenstheorie (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 des
Staatsrechts, Archiv fur Rechts- und Sozialphilosophie, 70 (1984), 71112; Allen W. Wood,
Hegels ethical thought (Cambridge: Cambridge University Press, 1990), p. 110; Matthias
Kaufmann, Zwangsrecht (}} 1316), in Jean-Christophe Merle (ed.), Johann Gottlieb
Fichte : Grundlage des Naturrechts (Berlin: Akademie Verlag, 2001), pp. 12537.
8. Cf., for instance, Stanley Benn, An approach to the problems of punishment, Philosophy,
33 (1958), 32141; Anthony M. Quinton, On punishment, in Harry Burrows Acton
(ed.), The philosophy of punishment (London: Macmillan, 1969), pp. 5564.
9. Georg Mohr, Unrecht und Strafe, in Ludwig Siep (ed.), G. W. F. Hegel: Grundlinien der
Philosophie des Rechts (Berlin: De Gruyter, 1997), pp. 95124 (pp. 95, 105).
hegels negation of crime
109
Hegel is a genuine retributivist. He rejects as superficial all theories

that try to justify punishment by the good which is supposed to come
of it. On the next page, Wood observes: but the states intention to
reassert the validity of right in the face of wrong looks like an intention
not to do justice as such, but to promote a good end, namely the public
recognition of the validity of right. Wood asks:
Why is it important for the state to assert the validity of right, to express its
disapproval of crime? Is there any reason for it to do this apart from its
devotion to such consequentialist ends as preventing future crimes and
reassuring people that their rights are being protected . . .?10
These very interpreters highlight in this way that Hegel argues

more convincingly for retributivism than Kant. According to Mohr,
compared to Kantian retribution theory . . . Hegels theory of punishment is the argumentatively better executed foundation of this type of
theory.11 Wood says of Hegels theory of punishment: This position
agrees with Kant, whose commitment to retributivism is clear, but
whose defense of it remains at best embryonic.12
Just as for Kants penal law, for Hegels penal law as well it is fair to
assume a mixed theory of punishment that ascribes to every argument
a part of the task of justification. Thus, Mohr wishes to recognize
Hegelian penal laws merit of building up a theoretic instrument
that does profitable preliminary work underlying the current claims
to a mixed theory.13 Just like the currently prevailing interpretation
of Kants theory, Mohr also sees a mixed theory of punishment in
Hegel:
Taking all the relevant passages from the Elements as the basis for the
interpretation of Hegels theory of punishment yields the diversified
picture of a modern mixed theory. It differentiates between justice in
punishment as an institution rooted in retributive theory and justice in
degree of punishment specified by deterrent theory.14
According to Mohrs interpretation, the relationship between a

certain element of retributive theory and a certain element of deterrent theory is formulated by Hegel the reverse way of that by Kant.
10. Allen W. Wood, Hegels ethical thought (Cambridge: Cambridge University Press, 1990),
pp. 10910.
11. Mohr, Unrecht und Strafe, p. 121.
12. Wood, Hegels ethical thought, p. 109.
110
According to the interpretation of the mixed theory, in Kant deterrent

theory justifies punishment as an institution, and retributive theory
justifies the degree of punishment. In contrast, Heiko H. Lesch sees in
Hegel the same relation between both elements of the mixed theory
as in Kant.15
The way in which Hegels penal law is interpreted as a mixed
theory, however, only indirectly correlates to the process of argumentation in Hegels text. Indeed, the connection made by such a mixed
theory between each of the arguments offered respectively by the
theories of deterrence and retributivism does not concern the relationship of the argumentation based on Abstract right to the argumentation based on Ethical life.16 The internal relationship of the
individual arguments based on Abstract right to one another is not
inquired into by the mixed theory.
5.2. The twofold justifications of punishment

It may astound one to see that the interpretations of Hegels theory of
punishment can contradict one another so deeply that some interpreters seem to vacillate between multiple interpretations. These different interpretations, however, draw on similarly different passages in
Hegels Elements of the philosophy of right (1821).
Elements of the Hegelian theory of punishment can be found
not only in the section C. Coercion and crime within Abstract right,17
but also in Ethical life, under the heading B. The administration
of justice, specifically in b. The existence (Dasein) of the law and
c. The court of law.18 The deterrence theory interpretations of
Hegel often base themselves on Ethical life, whereas the retributivist
interpretations either simply forget }} 21820 or declare them to
be unimportant. The latter interpretations categorize these paragraphs mainly as being a mere indication that a punishment must
be tailored to the respective societal conditions without the rationale
for punishment and the principle determining its degree being
affected by }} 21820.
15. Heiko Hartmut Lesch, Der Verbrechensbegriff: Grundlinien einer funktionalen Revision
(Cologne: Heymanns, 1999), p. 97.
16. GPhR, Parts i and iii, respectively.
17. GPhR }} 90103.
18. See GPhR }} 218 and 220, respectively.
111
5.2.1. An objective and subjective justification of punishment. The reverse

of those interpretations that wish to ignore or neglect }} 21820 is an
overestimation of the second part of the Elements of the philosophy of
right (Morality). In Morality, there are indeed elements that clearly
have to do with crime, especially in the section entitled Purpose and
responsibility. Yet in the Morality section, the word punishment
is found only three times. Moreover, it is not to be found either in
the main body of the text or in the Remarks but instead in the
Additions and there it is only an aside. Punishment is only treated
in the part of right as the universal or right in itself and in the part
of the subjective disposition, but of that right which has being in
itself.19 Punishment clearly belongs to the objective aspect of the idea
of right. Yet in the passage out of Abstract right, many interpreters
attempt to demonstrate the presence of two arguments that are both
in favor of retributivism: an objective one and a subjective one. The
clearest presentation of this twofold argumentation is provided by
Ossip K. Flechtheim and by Igor Primorac, and in a divergent form
by Kurt Seelmann.20
Flechtheim differentiates an objective justification from a subjective justification, since the objective way to legitimate punishment
appears to Hegel . . . as still being insufficient. He wishes to demonstrate that the criminal, by the criminal deed, also subjectively consents to the punishment.21
In the objective justification, there is an argument that draws on the
nature 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, Hegels deduction
has been based on the axiom that the relation between deed and
punishment is inseparable; that the deed already contains the
19. GPhR } 141, Remarks. G. W. F. Hegel, Elements of the philosophy of right, ed. Allen
W. Wood, trans. H. B. Nisbet (Cambridge: Cambridge University Press, 1991), p. 186.
20. Flechtheim, Die Funktion der Strafe, pp. 920; Flechtheim, Zur Kritik der
Hegelschen Straftrechtstheorie, 53948; and Flechtheim, Hegels Straftheorie
(Berlin: Duncker und Humblot, 1975). Primorac, Punishment as the criminal right,
18798; and Primoratz, Banquos Geist; Seelmann, Wechselseitige Anerkennung,
22836; Seelmann, Versuch einer Legitimation, 31526. Further authors assert
the existence of a twofold argumentation in Hegels theory of punishment, for
instance, Bertrand Guillarme, Penser la peine (Paris: Presses Universitaires de France,
2003), pp. 634.
21. Flechtheim, Von Hegel zu Kelsen, pp. 1718.
22. Primoratz, Banquos Geist, 39.
112
punishment in itself; and that the punishment of the deed necessarily

proceeds from the deed.23
I shall name this thesis the thesis of negation of the negation. According to
Flechtheim, 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, as
a second coercion which cancels an initial coercion.25
In } 99, Hegel equates the cancellation (Aufheben) of the crime, which

would otherwise be regarded as valid with the restoration of right.26
Flechtheim identifies the restoration (Wiederherstellung) explicitly
with the classical thesis of retaliation (Wiedervergeltung), which is
discussed in } 101.27
According to Flechtheims interpretation, the classical retributivist
thesis forms the link between the objective and the subjective justification. That is because, unlike right in itself, the principle of retaliation
draws not on the total legal system, but instead on the individual and
on his or her individual deed.
In the subjective justification, there is an argument that draws on the
relation 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 the
action of a rational being, that the crime is something universal and
that it lays down a law. The lawbreaker recognizes through his or her
deed that this law applies to him- or herself, so that the lawbreaker is
subsumable under that law, which is his or her own right. The criminal
is honored as a rational being by his or her action being regarded in
just such a way, that is, as if the action itself placed a claim on the
punishment by virtue of it being its own right.28
Flechtheim comments that:

From this conception of the legal offense, Hegel obtains the importance
that he attributes to the guilty plea and to the jury court. If the wrongdoer
23.
24.
25.
26.
27.
28.
Flechtheim, Von Hegel zu Kelsen, p. 12.

Coercion can also be regarded as being force.
GPhR } 93. Elements, ed. Wood, p. 120.
113
consents to his or her punishment, not only in so far as the wrongdoer is

the personification of an objective rational will but also as he or she is the
manifestation of his or her subjective particular will, hence every court
and every legal proceeding becomes superfluous since the criminal can
pass his own sentence him- or herself. But even Hegel does not dare to
construe reality in terms of this idea, and contents himself with the
requirement that what the judges pronounce does not differ from what
is in the consciousness from the subjective self-consciousness of the
wrongdoer.29
What Flechtheim quotes without providing a page reference is

derived, however, from the Additions in } 227, which is expressly
concerned with confessing to the elements of a crime therefore, it
is not, as Flechtheim assumes, concerned with the criminals consent
to penal law. Then, according to Flechtheims conception, } 101 is
seen as being a qualification of } 100. According to Flechtheim, the
criminal should have recognized talion law. Meaning, he continues,
that this criminal should recognize that he or she deserves to be
punished and that what the criminal has done should also happen to
him.30 This thesis is realized in the principle of equality between
crime and punishment. In turn, this principle is specified not as being
a specific equality, that is, as being in natura, but instead as being an
equality in terms of its value.31
It remains unclear, however, why punishment should require such
a twofold justification. The context in Flechtheim suggests that this
twofold justification is meant to bring about the criminals reconciliation with society. An obvious objection arises, though, against this
rationale for a twofold justification. Reconciliation was indeed Hegels
goal in the writings of his youth, but punishment was seen by him at
that time as something that is alien to the criminal as a human being.
That is because it was not something subjective, but instead only
something objective. At the outset of the section with the title
C. Coercion and crime, the concern is about something objective: the
cancellation (Aufhebung) of crime and the restoration (Wiederherstellung)
of right. Admittedly, Hegel writes: The injury . . . which is inflicted
on the criminal is not only just in itself [objective aspect] . . . it is also
a 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.
114
will [subjective aspect].32 Yet this implies neither that this subjective
aspect would be necessary for the justification of the punishment as a
complement to the objective justification, nor that the criminal would
have to consent to penal law in order that the punishment might be
legitimate. Hegels formulation in } 100 completely allows that, after
first being sentenced and then having served a punishment, the
criminal gains the insight that his or her deed must imply the punishment, even though the deed implied the punishment in itself already
while he or she was committing the deed. In such a case, the criminals
consent would not be unlike the case with the cancellation of the
crime and the restoration of right a (normative) prerequisite, but
instead a result of the crime.
Naturally, this does not mean that the subjective insight of the
criminal into the justification of his or her punishment has no systematic locus in the Elements of the philosophy of right. Obviously, the process
by which the criminal gains insight into the necessity of penal law
makes possible the Transition from right to morality. In the part
about morality that truly does constitute a further advance in the
inner conceptual determination of the will, punishment is not at
issue the word punishment itself practically never appears.33 It is
in the part about Ethical life, in which the opposition between the
general will in itself and the general will for itself is cancelled, that
punishment is discussed again for the first time. There it is handled
from the perspective that dangerousness is a criterion for punishment
in general as well as for legal punishment in particular, that is, for the
opposite of private revenge. }} 218 and 220 clearly do not make the
claim, however, to contributing to the justification of the existence of
punishment. To this extent, the justification of punishment rests
exclusively 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 Hegels handwritten marginalia in GPhR } 118 as well as in the
Remark on } 132, though Allen W. Wood, for instance in } 218, renders the
German 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 following
contexts: Suffering in general including punishment as a consequence in general of
the action, punishment, infringement, and unhappiness merely as such has no
ethical interest just as injustice and punishment are not mere ills (translation
mine); The sphere in which the above circumstances come into consideration as
grounds for relaxing the punishment is not the sphere of right, but the sphere of
clemency, in Elements, ed. Wood, p. 161.
115
it rests on the objective part. My conclusion is that the subjective

aspect of punishment provides no justification of punishment as being
the justification of a legal institution, even if there happen to be
elements of the concept of right in that justification.
In the subsection Coercion and crime, Hegel clearly formulates
the affiliation of the justification of punishment to the objective
aspect of the concept of right. In his Remark on } 99, Hegel speaks
of the objective consideration of justice, which is the primary and
substantial point of view in relation to crime.34 While the supposed
subjective justification comes under the heading the criminals own
right, its validity expressly remains conditional to the validity of the
objective justification: The injury . . . which is inflicted on the criminal is not only just in itself (and since it is just [emphasis mine], it is at
the same time his will as it is in itself, an existence . . . of his freedom, his
right).35 And while the alleged subjective justification appears in the
context of classical retaliation theory, then it is again expressly conditional to the validity of the objective justification: The cancellation . . .
of crime is retribution (Wiedervergeltung) in so far [emphasis mine] as the
latter, by its concept, is an infringement of an infringement.36 Hegel
criticizes the classical theory of retaliation by opposing it with
the argument of the nullity of crime: Yet the concept itself must
always 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 the
objective justification alone or the subjective justification alone is
sufficient. If neither of the two would suffice then one would have to
ask one more question to find to what extent they reciprocally complement each other and they together would be sufficient. What can
be found regarding this point in the secondary literature tends not
really to be convincing. Primoratz attempts to substantiate the twofold
justification in the following way:
This aspiration of Hegel to justify punishment both in the objective and
in the subjective sense is in no way contingent. Nor is it an expression of
his doubts about the rightness and convincingness of the one or the
other justification and about the consequent need for a complementing
argument . . . His fundamental objection that he brings forward against
34.
35.
36.
37.
GPhR } 99, Remark. Elements, ed. Wood, p. 125.

116
Plato consists in Plato not recognizing the principle of subjectivity, that

is, 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 criminals

humanity. It still remains for us to consider the alternative interpretation of the twofold justification of punishment given by Seelmann.
He sees in Hegel two different constructions for legitimating punishment and penal law: the argument from law and the argument
from recognition.39 Both link the objective with the subjective
dimension.
The argument from law is the thesis previously mentioned, that is, the
thesis of the criminals own right or subsumption under the criminals own law.40 I will come back to this later. In what follows, I will
address the question whether Hegels (objective) argument from law
is linked to the (subjective) argument from recognition in a twofold
justification.
The argument from recognition relies on } 97 of the Elements of the
philosophy 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. The
manifestation of its nullity is that the nullification of the infringement
likewise comes into existence; this is the actuality of right, as its necessity
which mediates itself with itself through the cancellation . . . of its
infringement.41
According to Seelmann this argument can be understood only if

one takes as a point of departure Hegels understanding of right as
being a universal relation of reciprocal recognition of free and equal
persons . . . an understanding that Hegel borrowed from Fichte, while
introducing some modifications.42 Seelmann himself formulates the
argument along the following lines:
Since . . . recognition is a reciprocally conditioned relation [according
to Hegel], the wrongdoer himself withdraws the recognition from himor 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. 1267.
42. Seelmann, Wechselseitige Anerkennung, 230.
117
manifestation of this situation that the wrongdoer him- or herself has

brought about, whereby punishment lowers the physically observable
legal status of the wrongdoer in order to create a legal relationship
based on equality.43
The argument from recognition is thus not to be understood as if

punishment were justified by the crime being a refusal of recognition
on the part of the criminal, because based on this denial of recognition one could just as well shoot the criminal as if he or she was a
dangerous animal that anyone should be allowed to kill without any
particular procedure that is, without a legal sentence, as is the case in
Fichtes thought experiment, to which Hegel alludes in another passage.44 Rather, the argument from recognition signifies that the aim
of punishment is situated in compelling the recognition of the victim
as an equal by the wrongdoer: the aim of punishment is the creation
of (in Seelmanns words) a legal relationship based on equality.45
This means that the humiliation of the criminal, which goes along
with the punishment, should only occur temporarily. Therefore, the
punishment is justified not in itself, but merely by its aim. If retaliation
is to play a role here, then it should only play a role as an instrument
of rehabilitation.
Before one adopts the argument from recognition that has been
reconstructed in such a way, one must note that it lacks any basis in this
paragraphs text; the relation of the argument to } 97 is not substantiated by Seelmann.46 The argument from recognition sees disruption of
a reciprocally conditioned relation in the nullity of the infringement
of 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 in
civil society, crime is no longer an injury . . . merely to a subjective infinite,
but to the universal cause . . . whose existence . . . is inherently . . . stable
and strong.48
43. Seelmann, Versuch einer Legitimation, 320.
44. Fichtes thought experiment is found in GNR i/4 59. Fichte, Foundations of natural
right, p. 226. For more on Fichtes thought experiment, see Lazzari, Eine Fessel,
pp. 17386. Hegel alludes to this in GPhR } 100. Elements, ed. Wood, p. 126.
46. It is substantiated neither in Versuch einer Legitimation nor in Wechselseitige
Anerkennung.
118
In the state, right, getting legal validity, is no longer dependent on

recognition on the part of individuals. If the crime would not be
cancelled by a second coercion then, indeed, it would constitute a
precedent that might call into question the validity of right.49 In such
a case, it would be not only the criminals legal status as in Fichte
that would be at risk, but also as in Fichte the legal relationship of
all persons with one another. Moreover, the crime does not necessarily imply the non-recognition of a legal person by the criminal. In his
Remark on } 95, Hegel draws attention to the negation of the legal
capacity of a person as only representing shapes of the crime. Hegel
makes 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 single
person but rather against the state. Furthermore:
The substantial element within these forms is the universal, which
remains the same in its further development and in the further shapes
it assumes; thus its infringement, i.e. crime, also remains the same, in
conformity with its concept.50
First and foremost, the crimes nullity can be explained in another

way. According to Hegel, force or coercion immediately destroys itself
in its concept, since it is the expression of a will which cancels . . . the
expression or existence of a will.51 The crimes nullity relies on the
will itself being destroyed in the crime. Thus, according to Hegel,
the crime consists in a self-contradiction rather than in an infringement upon reciprocity.
5.2.3. A mixed theory of punishment. Another twofold objective and
subjective justification of punishment is also presented on the model
of Kants penal law: a mixed theory (see Section 2.1). Such a mixed
theory of punishment consists of ascribing to retributivism and to the
theory of general deterrence each a part of the task of justification. In
this way, Mohr wishes to recognize a theoretic instrument [that] does
profitable preliminary work underlying the current claims to a mixed
theory.52 According to Mohr, the Elements combine a retributivist
49.
50.
51.
52.

Mohr, Unrecht und Strafe, p. 122.
119
rationale of punishment as an institution with a general deterrent

specification 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 retributivist
element in Abstract right. Therefore, it inverts that interpretation
of Kants penal law as a mixed theory according to which deterrence
theory justifies punishment as being an institution and retaliation
theory justifies the degree of punishment. If Kant and Hegel are really
mixed theorists, then Hegel is thus the only one of the two who
justifies 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 mind
again the results of Chapter 2: with Kants conception, as it is generally
known, one cannot avoid a collision between general deterrence
in the execution of punishment and the categorical imperatives
application dictates that the humanity in the criminal should not
be treated as a means, but always only as an end. General deterrence,
which is mixed with retributivism, does not solely concern the threat
of punishment. Kants (controversial) implicit assumption presupposes that the mere threat of punishment treats no one as a mere
means, since, merely by threat, no harm is inflicted on anyone so long
as this person does not commit a crime. Whoever commits a crime
would then not be punished according to the degree dictated
by general deterrence but rather according to the degree dictated
by retributivism. Hence only the threat of punishment not the
punishment itself is justified using general deterrence. Thus, both
the existence of punishment and the degree of punishment remain
grounded in a purely retributive way. That is due to Kant not allowing
the degree of punishment to be defined other than retributively.
In Kant, the threat of a non-retributivist degree of punishment is
not allowed to lead to a non-retributivist degree of punishment
during the execution of the punishment, which would contradict
the categorical imperative. Mixed theorys interpretation, hence, is
only in appearance a mixed theory in reality it remains a retributivist
justification of punishment.
Now how does all of this look in Hegels alleged mixed theory? To
Mohr, } 218, with its character of general deterrence, is a further
relevant angle for the apportionment of punishment appended by
53. Mohr, Unrecht und Strafe, p. 122; see also Section 5.1.
120
Hegel to his conception of punishment first in the theory of civil

society.54 The possibility of a mixed theory presupposes that this
relevant 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 and
strong existence that the law should have, that is, in relation to
compliance with the law.55 The more the crime endangers general
compliance with the law, the more strictly the crime must be punished. If no state exists, the danger is, by definition, infinite and the
revenge must then be in equal measure to the danger.56 If the power
of society has now become sure of itself then the punishments could,
and should, become much more lenient. Such a definition of punishment prescribes deterring further crimes as the aim of punishment.
One cannot see, in Hegels view, any retributivism unless one were
to conceive so broadly of the definition of retributivism that the
proportionality, or the parity, between crime and punishment would
end up amounting to the parity between danger and punishment.
By so doing, though, the retributivist thesis actually would become
indistinguishable 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 supposed
Kantian mixed theory. A Kantian mixed theory remains fundamentally retributive; a Hegelian mixed theory ultimately remains a theory
of deterrence.
5.3. Objections to the retributivist interpretations

If one can find in Hegels theory of punishment neither a twofold
justification nor a mixed theory, but instead only an objective interpretation, which draws either on retributivism or on a deterrent
theory, then the retributivist reading seems likely. Therefore, it is
the traditional and most supported reading. Now, however, compatibility of the classical principle of retaliation with the Hegelian concept
of right proves itself to be problematic in at least two ways.

56. GPhR } 218, Remarks. Elements, ed. Wood, pp. 250f.
121
5.3.1. Hegels 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 of
the negation57 as well as in relation to taking into account the crimes
quantity and quality in the determination of the negation of the
negation:58
The cancellation . . . of crime is retribution in so far as the latter, by its
concept, is an infringement of an infringement, and in so far as crime,
by its existence . . . has a determinate qualitative and quantitative
magnitude, so that its negation, as existent, also has a determinate
magnitude.
Immediately thereafter, Hegel expressly rejects classical retaliation as

being equality in the specific character of the infringement. Hegel
writes: this identity [of crime and retribution], which is based on the
concept, is not an equality in the specific character of the infringement, 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 the
specific character of the infringement can possibly mean is elucidated by Hegel in the Remarks:
It is only in terms of this specific shape [specific equality] that theft
and robbery . . . and fines and imprisonment etc. . . . are completely
unequal, whereas in terms of their value, i.e. their universal character as
injuries . . . they are comparable.60
Therefore, the only retaliation (Wiedervergeltung) that Hegel

announces himself as supporting is nothing other than the infringement of the infringement.
Some interpreters are of the opinion that Hegels critique is
directed only against a literal application of the classical theory of
retaliation. According to these interpreters, Hegels critique can be
reduced to these lines: the sphere of externality, in which no absolute
determination is in any case possible; in the realm of finite things, the
absolute determination remains only a requirement, on which the
understanding must impose increasing restrictions . . . [which admits]
only an approximate fulfilment. This aspect of Hegels critique surely
explains the insuperable difficulty for the thesis of specific equality
57.
58.
59.
60.
GPhR }} 93 and 97. Elements, ed. Wood, pp. 1201 and p. 123.
GPhR } 96. Elements, ed. Wood, pp. 1223.
GPhR } 101, Remarks. Elements, ed. Wood, p. 129.
122
to determine punishments (especially if psychology also invokes

the strength of sensuous motives . . . and, as a corollary, either the
correspondingly greater strength of the evil will or if we prefer the
correspondingly lesser strength and freedom of the will in general).61
If, however, one reads Hegels critique of classical retaliation in an

unabridged form then one must particularly take into account the
second half of Hegels argument:
If we not only overlook this nature of the finite realm but [emphasis
mine] also proceed no further than abstract and specific equality, an
insuperable difficulty arises when we come to determine punishments
. . . Furthermore [emphasis mine], it is very easy to portray the retributive
aspect (Wiedervergeltung) of punishment as an absurdity (theft as
retribution for theft, robbery for robbery, an eye for an eye, and a
tooth for a tooth, so that one can even imagine the miscreant as
one-eyed or toothless); but the concept has nothing to do with this
absurdity, 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, but

rather in the axiom that what the criminal has done should also happen to
him, and therefore in the specific equality or in talion law in contrast
to 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 a
philosophical determination of the concept of right. He introduces
the remarks on } 101 as follows:
It is usual in science for a determination in this case, that of
punishment to be defined in terms of the universal representations . . .
of conscious psychological experience. In the present case, this
experience would indicate that the universal feeling of peoples and
individuals towards crime is, and always has been, that it deserves to be
punished, and that what the criminal has done should also happen to him.64
Primoratz maintains that this attitude of reciprocity is the very

reciprocity of a man towards punishment. He annotates Hegel: On
this point, commonsense is correct, Hegel believes, and philosophy has
61.
62.
63.
64.

123
to abide by the judgment of common sense.65 This is contradicted not

only by the fact that Hegel has time and again either strictly differentiated between common sense philosophy and the mere science of
understanding or expressly wanted to differentiate between the two.
One of the many points of Hegels critique of common sense and of the
science of understanding is that they become entangled in contradictions from which they attempt in vain to loosen themselves, for they
cannot cancel these very contradictions philosophically. Again, in the
passage about crime Hegel continues:
It is incomprehensible how those sciences which derive their
determinations from universal representations . . . should on other
occasions accept propositions which contradict such so-called
universal facts of consciousness.66
Hegel presumably means by this those various theories of punishment as prevention, as a deterrent, a threat, a corrective, etc. which
presuppose, in the punishments rationale, the superficial character
of an evil.67 Hegel sharply criticizes them:
As a result of these superficial points of view, however, the objective
consideration of justice, which is the primary and substantial point of
view in relation to crime, is set aside; it automatically follows that the
essential consideration is now the moral point of view, i.e. the subjective
aspect of crime, intermixed with trivial psychological ideas.68
I proceed under the assumption that the contradictory propositions that Hegel criticizes in the Remark on } 101 are those very
theories I have described above, because both these and retributivism
were already then the two dominant competing theories of punishment: punishment was imposed either because the crime was committed
(retributivism), or in order that something occur (rehabilitation, either
general 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 raises
against the various consequentialist theories of punishment, that is,
that they rely on common sense, that they are subjective rationales
and that they contradict one another.
65.
66.
67.
68.
Primoratz, Banquos Geist, 37.

124
It is well known that, whenever the theories of understanding

contradict one another and the dispute seems to be irresolvable,
Hegels science of reason consists in canceling (aufheben) these
contradictions.69 In this cancellation, the opposing theories prove
themselves to be merely partial insights into the relevant concept.
In the question of penal law as well, Hegel proceeds exactly according
to this method in the Elements of the philosophy of right. Both retributivism and consequentialism are categorized expressly as partial
insights into the concept of punishment by Hegel so that both, by
their claim that they provide a rationale of punishment, distort
that very concept of punishment. Hegel says of the consequentialist
theories of punishment that
the various considerations which are relevant to punishment as a
phenomenon . . . 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. In
the present discussion, we are solely concerned with the need to cancel
(aufheben) crime not as a source of evil, but as an infringement of right
as right and also with the kind of existence which crime possesses, which
must also be cancelled.70
About retributivism, Hegel states: It is this inner identity which, for the
understanding, 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 opposition
between both of these theories in the concept of punishment, there is
a point that contradicts the assumption of some interpreters, stating
that the classical theory of retaliation and the Hegelian penal theory
are 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 the
negation of an opposing element, as in the case that A is the negation
of :A, and therefore it is ::A. For Hegel, the negation of the negation
is rather the cancellation of the negation, that is, the cancellation of
69. Cf., for example, Hegels Encyclopedia of the philosophical sciences, } 36, Addition.
125
coercion. It is not ::A, but rather A. Yet it is not the absolute A, but
instead 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 negation
is on the whole negated. Now, right requires that the negation itself
becomes negated, and therefore that no violence results from the
punishment. The punishment is the special negation actually bringing
to 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 has
negated the personality (A) of his or her victim: :A. Happening to the
criminal what he or she him- or herself has done to the victim means
the negation of the personality (B) of the criminal: :B. The relationship between :A and :B is naturally that B executed the action :A.
Through :B, :A was indeed negated but B was also negated. Therefore, :B must be negated. In other words: whoever is responsible for
making sure that what the criminal has done happens to the criminal
him- 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 negation
instead of canceling it. Hegel expresses this in the following way:
In the sphere of the immediacy of right, the cancellation . . . of crime is
primarily revenge, and its content is just so far as it constitutes retribution
. . . Thus revenge, as the positive action of a particular will, becomes
a new infringement; because of this contradiction, it becomes part of an
infinite progression and is inherited indefinitely from generation to
generation.73
Hegel clearly draws the consequence from this opposition between

the negation of the negation and the classical theory of retaliation.
To require that this contradiction [between the negation of the crime and
its multiplication] . . . should be resolved . . . is to require a . . . punitive
rather than an avenging justice. Primarily, this constitutes a requirement
for a will which, as a particular and subjective will, also wills the universal
as such.74
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 uncivilized
peoples that propagates itself as revenge and then revenge for that revenge from
generation to generation.
126
A common way of interpreting this passage involves reading it as

merely being a requirement for impartiality of the judge against the
risk of an interested and thus partial private justice. However, this
implies an inadmissible abridgment. For even if this private justice
committed itself to impartiality, the problem that Hegel addressed
would still exist. The problem consists namely in that the victim place[s]
[his or her] infinity in any infringement . . . which occurs, and whose
justice is therefore altogether contingent. In this sentence, Hegel
does not deny that the revenger acts justly. The content of revenge
is, he writes, just so far as it constitutes retribution, and Hegel speaks
of [its] justice.75 Therefore, this sentence is not concerned with the
risk of the misuse of punishment. It is concerned with the right that is
to 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 and
therefore are allowed to make a claim on the right of personality, in
the state every person is subject to the law, in which inheres the
objective right of universality. Before the law, the person has only
the subjective right of personality, which implies, for instance, that
the state is not allowed to handle him or her arbitrarily. The consequence is that that which is to be restored as right is, in the case of
revenge, the personality of the criminal but in the case of punishment
the validity of the law.76
Now the victims personality does not directly contain the personality of the criminal, whereas the law, by definition, recognizes the
personality of all members of the commonwealth. Hegel says of the
punitive justice of the state: Instead of the injured party, the injured
universal now makes its appearance. Hegel places the subjective and
contingent revenge in contrast with the punishment:
Objectively, this reconciliation applies to the law, which restores and
thereby actualizes itself as valid through the cancellation (Aufheben) of the
crime; and subjectively, it applies to the criminal in that his law, . . . which
is known by him and is valid for him and for his protection, is enforced upon
him in such a way that he himself finds in it the satisfaction of justice
and merely the enactment of what is proper to him.77
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.
127
What Hegel calls retribution (Wiedervergeltung), which is an infringement of an infringement, should not be confused, therefore, with the
retaliation (Wiedervergeltung) of classical retributivism.78
In the degree of punishment as well, a difference now arises
between the restoration of the personality of the individual or of a
group of individuals and the restoration of the law. Restoring the
law, that is, to reaffirm its validity, is completely feasible. Restoring the
existence (Dasein) of the will in an individual person in which it was
found before the crime is, on the contrary, only possible in some
cases. These cases are treated in } 98 of the Elements of the philosophy
of right: The cancellation . . . of the infringement, where the latter has
caused damage, is civil satisfaction in the form of compensation (in so
far as any compensation is possible) [second emphasis mine].79 In most
of the cases, though, there is no compensation to be found. And in
those cases as well, wherein compensation is possible, the positive
existence of the injury, that is, the particular will of the criminal, remains
unaffected: the criminal can provide the whole compensation while
keeping his or her criminal intentions.80
Some authors attempt to justify the classical principle of retaliation
not merely by appeal to reciprocity in general, but also in other ways.
Jean Hampton mentions, for example, the explanation that talion law
has the aim of destroying the criminals claim permanently to acquire
supremacy over his or her victim. Poking out the criminals eye is,
according to this conception, a way for the victim, who was made
monocular by the criminal, to revoke that superiority unjustly gained
by the criminal. The question of the rightness of this and of similar
explanations remains to be addressed (Hampton herself rejects it).
Bear in mind, though, that Hegel does not enter into such interpretations of classical retaliation theory.81 In its place, Hegel adopts a
symbolic interpretation: It is this inner identity which, for the understanding, is reflected in external existence . . . as equality. In revenge,
the restoration of right cannot really occur: therefore, it occurs symbolically that is, as talion law, as a specific equality: robbery for
robbery, an eye for an eye, and a tooth for a tooth.82
Cf. John Kleinig, Punishment and moral seriousness, Israel Law Review, 25, no. 3
(1991), 40121 (p. 416).
78.
79.
80.
81.
128
The absurdity of such punishments consists of their having

nothing to do with the restoration of right, with the termination of
the coercion and with reconciliation a reconciliation with the criminal as well. On the contrary, such punishments do not treat the
criminal as an existence of freedom. Only in so far as the criminal
is a member of a commonwealth is the preservation of the wills
existence in the criminals person necessary and required for the
restoration of a commonwealth to which he or she belongs and by
which his or her rights are protected as well. The criminals forced
amputation is not necessary for such a restoration of the validity
of law. Rather, amputations contradict the idea of a commonwealth
of free persons. For that reason, amputations are absurdit[ies].83
5.4. Reconstruction of the Hegelian legal

justification of punishment
The grounding of punishment in Hegel consists of the verification
that the punishment is a necessity of right. The keystone of the
normative construction of punishment, namely the second coercion
or that which cancels an initial coercion (we will see later how this
might be understood), is not only conditionally right but necessary
for right.84 Punishment, therefore, is not only in compliance with
right by itself; rather, without it there would be no right. In contrast
to that, the initial coercion is contrary to right. Hegels grounding of
punishment does not separate the authority to punish from the duty
to punish; initially, it does not attempt to demonstrate first that
punishment may occur and then, after that, that punishment should
occur, but instead it fulfills both tasks together.
Hegel substantiates this necessity in the following way: Because
coercion destroys itself in its concept, it has its real expression . . . in
the fact that coercion is cancelled . . . by coercion.85 This thesis of the
conceptual self-destruction of coercion is not to be taken as a physical
thesis. It does not concern, for instance, the physical self-destruction
of coercion; instead, what matters is that this coercion is contrary to
right.86 Contrariness to right is the same as the concept of legal
83.
84.
85.
86.
GPhR } 101. Remarks. Elements, ed. Wood, p. 128.

129
impossibility.87 The coercion destroying itself means that coercion is

rightly impossible or incompatible with the concept of right. In
the self-destruction two elements must be distinguished: (1) coercion
is destroyed; (2) it itself carries this destruction through: it is a selfdestruction. The first coercion is canceled by the second coercion;
important here is to realize that it is canceled by nothing else other than
coercion.
In the second regard, coercion contrary to right differentiates itself
from the infringements it creates, that is, from the evil or the
damage, because when an infringement of right occurs, it does
have a positive external existence . . . but this existence within itself is
null and void.88 Even if the infringement being a consequence of
the coercion has a physical existence that is, as damage it actually
does not have any existence in right at all. This infringement is not
destroyed by itself.
The difference between the self-destruction of coercion and the
extrinsic destruction of damage comes to expression in the form of
the differentiation between the cancellation (in right) of coercion
and the nullity (in right) of damage.89 As is generally known, the
Aufhebung (cancellation) of contradictions, which is a central concept in the Hegelian dialectic, means not only destroying whatever
may contain a contradiction, but also simultaneously preserving something of it. Therefore, something is preserved from coercion by its
destruction. We must take note of this point for later.
Many interpreters do not know how to make sense of selfdestruction of coercion or regard it as being nonsensical. This incomprehension often consists of overlooking the premises upon which
the thesis of self-destruction rests. I shall provide the original text
from Hegels Elements in order to make reference to the grammatical
structures in the German original, which I also emphasize:
[Because] the will is Idea or [is] actually free only in so far as it has
existence . . . and the existence in which it has embodied itself is the
being of freedom, [then] force or coercion immediately destroys itself
in its concept.
87. There is more at stake here than either the credibility of the legal system or the honesty
of the civil authorities. David Cooper believes this to be Hegels core argument. David
Cooper, Hegels theory of punishment, in Zbigniew A. Pelczynski (ed.), Hegels political
philosophy (Cambridge: Cambridge University Press, 1971), pp. 15167.
88. GPhR } 97. Remarks. Elements, ed. Wood, p. 123.
89. GPhR }} 93 and 97, respectively. Elements, ed. Wood, pp. 120 and 123.
130
Weil der Wille, nur insofern er Dasein hat, Idee oder wirklich frei und
das Dasein, in welches er sich gelegt hat, Sein der Freiheit ist, so zerstort
Gewalt oder Zwang in ihrem Begriff sich unmittelbar selbst.90
The fundamental premise here as in the Elements of the philosophy of

right on the whole, actually is that free will must develop; and this
development is what Hegel terms right. I shall here content myself
with calling to mind a few central statements from the introductory
paragraphs of the Elements: The subject-matter of the philosophical science
of 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 precise
location and point of departure is the will; the will is free, so that
freedom constitutes its substance and destiny . . . and the system of
right is the realm of actualized freedom.92
And: Right is any existence . . . in general which is the existence of the

free will. Right is therefore in general freedom, as Idea.93 Right having
to be plays the role for Hegel that is played by the categorical imperative for Kant: the establishment of a state of law is a necessity.
The existence of freedom required by the idea of right consists in
the abstract right of the individual body and life as well as of
personal property. Hegel notes that
free will in and for itself cannot be coerced . . . except in so far as it fails to
withdraw itself from the external dimension in which it is caught up, or from
its idea . . . of the latter . . . Only he who wills to be coerced can be coerced
into anything.94
It should be noted that free will must want to allow itself to be

coerced by the criminal because, according to the idea of right, it must
have an existence: in terms of right it cannot withdraw from externality. Right forbids the victim from a stoic retreat into internality in
order to escape the criminals coercion.
In Ethical life it is not the individual but the state that displays the
highest existence of free will. However, the grounding of penal law
in 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.
94. GPhR } 91. Elements, ed. Wood, pp. 11920. Regarding property, cf. GPhR } 90
(property) (p. 119) and } 91 (his physical side and other external attributes).
131
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 of
the free will. Therefore, the body, life and property of the criminal
are just as much the existence of the free will as the body and life of
the victim. Therefore, self-destruction in } 92 ought to be understood
in the following ways:
1 One must understand self-destruction as the thesis of the necessary
destruction of that which contradicts right. Since the realization of
right and along with it as well, the existence of the free will in the
body, in the life and in the property of the victim is a necessity, the
crime, being an attempt to infringe on this existence, can only fail
and lead to the exercising of coercion against the criminal.
2 Self-destruction must also be understood as the thesis of the identity
of the criminals free will with the victims free will. The criminal
coerces or destroys the existence of free will in the person of the
victim and thereby coerces or destroys him- or herself, since the
criminals own existence is also the existence of free will.
The two theses are linked to each other, since right must have its
existence in every person, so that, for right, the personality of the
criminal and that of the victim are equal. In Hegel, the refusal to
recognize the personality of all the other humans implies a discrepancy on the part of that person who denies the recognition: on the
one hand, the universal element in the persons will is one with the
universal element in the wills of the other persons; on the other hand,
this universality is alien to it. Its particular will and its general will
diverge. This implies that the human who does not recognize the
status of the personality in other humans lacks the foundation upon
which he or she can base his or her own status as a person. According
to Hegel, those societies in which the personality of only some
humans 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 of
coercion. Now, this requirement is still taken in the abstract (that
is, 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 be
considered in the following paragraph also applies to the particular and further
determined content [of crime], e.g. in perjury, treason, counterfeiting, forgery, etc.
132
mine]. According to Hegel, punishment is the real expression of

the conceptual destruction of coercion.96 The necessity of right
meaning, the requirement that right realizes itself because it must
realize itself demands that coercion is cancelled . . . by coercion.97 Hegel
also terms the realization or the implementation of the requirement
of right the manifestation of this requirement; the punishment is
. . . the manifestation of the crime.98
How can coercion cancel coercion? Hegel gives the following definition of coercion: in coercions externality, the will may either experience force in general, or it may be forced to sacrifice or do something
as a condition of retaining some possession or positive being, thereby
suffering 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 by making 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 coercion
and to mere force, and initial coercion, that is, the coercion that
is contrary to right (the crime), can either be coercion or force,100
Hegel in } 93 merely mentions the cancellation of coercion by
coercion, and, with the mention of coercion in the Remarks to } 93,
the cancellation of force by coercion. Nowhere does Hegel speak of
the cancellation of either coercion or force by force. In the light of the
definition of force as mere suffering, cancellation by force is explicitly
excluded by Hegel by his criticism of Ernst Ferdinand Kleins conception of punishment as being an arbitrary cancellation of an evil
through an evil:
If the crime and its cancellation . . . which is further determined as
punishment, are regarded only as evils . . . in general, one may well
consider it unreasonable to will an evil merely because another evil is
already present.101
96.
97.
98.
99.
100.
101.

GPhR } 97. Elements, ed. Wood, p. 123; and } 101, Addition. Elements, p. 129.
Cf. GPhR } 93, Remarks. Elements, ed. Wood, p. 120: initial coercion, or at least force.
GPhR } 99. Elements, ed. Wood, p. 124. Regarding Klein: Hegel is referring to Ernst
Ferdinand Klein, Grundlinien des gemeinen deutschen und preuischen peinlichen Rechts
(Halle: Hemmerde und Schwetschke, 1796), }} 910.
133
At least three consequences result from this.

1 It begs the question of the applicability of the cancellation of
coercion by a coercion exercised upon the criminal, if the criminal
does not allow him- or herself to be forced. We will come back to
this 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 but
exercises mere force against him or her. Hegels position on
the question of the death penalty appears to be inconsistent. On
the one hand, he maintains that murder necessarily incurs the
death penalty.102 Admittedly he writes that it is to Beccarias credit
that the death penalty was repealed for most crimes. Hegel himself
wants to keep the death penalty for special cases, especially with
murder in mind:
Even if neither Joseph II nor the French have ever managed to secure
its complete abolition, people have begun to appreciate which crimes
deserve the death penalty and which do not. The death penalty has
consequently become less frequent, as indeed this ultimate form of
punishment deserves to be.103
On the other hand, Hegel really sees alternatives to the death penalty. Because punishment must coerce the criminal, whereas the
death penalty does perfectly fulfill the personal wishes of some to
be sentenced to death, then Hegel must actually allow some alternative punishments, even for murder. A handwritten remark in his copy
of the Elements on } 99 notes that punishment has to be sensitive . . .
It even has occurred that murders [were committed] in order to lose
[ones] life, therefore, [they are] insensitive to the death penalty;
then it came to be, that it was changed to imprisonment. In the
Lecture on the philosophy of right (1824/5) he says in a somewhat more
detailed way:
There have been cases in which a murder occurred in order [for
one] to be sentenced to death. The murderer did it out of world
weariness, contempt for life, though especially in a religious sense . . .
Thus the death penalty does not affect him he already wills to leave
this life consequently, a sentence of life has been replaced with
prison in order to affect the will of the criminal.104
102. GPhR } 101, Addition. Elements, ed. Wood, p. 129.
104. PhR 285.
134
In the Elements of the philosophy of right, there is no argument to be

found against substituting alternative punishments for the death
penalty for all crimes. On the contrary, an adherence to the death
penalty would lead to another definition of punishment, namely to
cancellation of coercion or (for example, in the case of murder) of
force by (mere) force. Moreover, this new definition would contradict Hegels theory in some ways. I have already mentioned his
criticism of Klein. Also there is his criticism of a penal theory that
Fichte once took into consideration. A cancellation of coercion or
of force by force, according to Hegel, amounts to the conception of
punishment in which a criminal is regarded simply as a harmful
animal which must be rendered harmless.105 This theory of punishment 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 cancellation of coercion by coercion seem to imply the same conceptual selfdestruction and real self-cancellation. It is an obvious temptation to
conclude that there is a self-cancellation of coercion, which
could possibly be an argument in favor of retributivism. But Hegel
does not actually do this; instead, he refers to a conceptual selfdestruction of force.106 Apart from this, we have just seen that a selfcancellation of mere force by mere force is out of the question, so that
only a cancellation of mere force by coercion is conceivable.
Notwithstanding the question whether the crime was mere force or
coercion, what matters is that the cancellation of the crime happens
through coercion. As we have seen, coercion is (1) the exercise of
force upon a person, (2) in order to make the person act differently or
sacrifice something, (3) by making this action or this sacrifice a precondition for keeping his or her body, life or property. If this definition is applied to the second coercion to the coercion against the
criminal then from these three elements result three implications.
The cancellation of the crime is: (1) an evil or force that is
exercised against the criminal, (2) so that the criminal acts differently
or sacrifices something, (3) by making this altered deed or this sacrifice
to be a precondition for keeping his or her body, life or property.
106. Cf. GPhR } 92. Elements, ed. Wood, p. 129: Consequently, force or coercion
immediately destroys itself in its concept.
135
5.5. Integration of the other penal theories

A threefold objection can be made against such an interpretation:
first, Hegels criticism of Kleins conception of punishment as being
the infliction of an evil; secondly, Hegels criticism of the penal reform
theory;107 thirdly, the fact that Hegel nowhere at least nowhere in
Abstract right makes the criminals rehabilitation an issue.
The first two objections misunderstand Hegels criticism of Klein
and of the reform theory. Hegels criticism is directed exclusively
against absolutizing evils and reform. Hegel writes:
If the crime and its cancellation, which is further determined as
punishment, are regarded only as evils in general, one may well
consider it unreasonable to will an evil merely [emphasis mine] because
another evil is already present.108
Hegels critique only maintains that the aim of punishment, which

serves to ground punishment, cannot be to inflict an evil on the
criminal, although the criminal has actually inflicted an evil on his
or her victim. It is for this reason that Hegel speaks of a superficial
character of an evil: the evil is not a primary but at most a
derivative aim of punishment. As regards the reform theory, it is
equally discarded only as an absolutization of reform. Hegel writes:
The various considerations which are relevant to punishment as a
phenomenon . . . 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 of
punishment. But they take it for granted that punishment in and for
itself is just.109
As regards the theory of rehabilitation, it is expressly found in Ethical

life. Retribution through revenge is contrasted by Hegel to genuine reconciliation of right: subjectively, it applies to the criminal in
that his law, which is known to him and is valid for him and for his
protection.110 As early as in Abstract right, Hegel is saying of the
cancellation of crime: it is at the same time his will as it is in itself,
107.
108.
109.
110.
For both points one and two cf. the Remarks in GPhR }} 99 and 100.
136
an existence of his freedom, his right.111 The punishment, being as it

is just, guarantees the right, which is the existence (Dasein) of the
criminals free will; thus, punishment is the precondition of the full
membership of the criminal. It should be noted once more that
the criminals rehabilitation as a full member of the commonwealth
presupposes that punishment is just. Therefore, the rehabilitation
aspect is also a secondary, derivative aim of punishment.
Hegels theory is not absolutist, that is, he does not support the
thesis that punishment is in itself good and already justified because
the criminal has deserved the punishment through his or her crime.
In an absolutist theory of punishment, punishment is only grounded
upon the criminal and what the criminal has merited; punishment is
then only seen as being a merited ill. We find good evidence in Kants
example of the island in which the sole justification of punishment
cited is: so that each has done to him what his deeds deserve.112 The
role of this example in the Kantian theory of right will not be discussed here.113 Here I content myself with drawing attention to the
fact that one searches in vain for a similar example in Hegel.
Hegels theory of punishment thoroughly assigns an aim to the
punishment: the cancellation of the crime. The cancellation of
the crime by coercion has two unequal dimensions: the past and the
future.
The cancellation of the crimes past dimension clearly does not mean
that the crime would be made undone, that is, that everything were as
if nothing had happened. At the most, some of the consequences of
the crime can be eliminated by civil satisfaction in the form of
compensation (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 take
the place of its specific qualitative character.114 To the extent to
which the criminal is thus coerced to make civil amends, the amends
can be placed in the category of cancellation of coercion by coercion.
Hegel, however, finds this aspect of cancellation of crime to be of
marginal 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.
112.
113.
114.

Cf., though, Section 1.3 above.
GPhR } 98 and Remarks. Elements, ed. Wood, p. 124.
137
compensation as well.115 The criminal aspect is not situated in the evil

or the harm. As I said before, Hegel indeed ascertains a positive,
external existence in harm; it is, however, within itself . . . null and
void, for the positive existence comprises not an existence of will but,
instead, a mere product thereof.116 The criminal did not primarily
intend harm. The criminal primarily intended the realization of his or
her freedom in a specific existence in an alien good, in a larger
freedom through the elimination of other humans, etc. by means of
the exercise of a coercion on another will. At most, the criminal
recklessly disregarded the harm it might cause. The compensation
here is not to be confused with, for instance, the restitution of stolen
goods to their owner. The question arises here of the appraisal
of vandalism and of other crimes whose exclusive intent is destruction.
A possible answer to this objection would be to see in these cases the
criminals disappointed will that does not receive the existence it
designed. No matter how it stands with this objection, in any case for
Hegel the crime definitely belongs to the past: the harm corresponds
neither to something in the will of the victim who ceased to be
coerced by the criminal nor to something in the criminal who had
another primary intention.
The future dimension of the cancellation of the crime is the specific
dimension for the crime. Hegel observes: The positive existence of the
injury consists solely in the particular will of the criminal.117 The cancellation of the crime by coercion can now occur either during the
commission of the crime or after its commission.
The cancellation of the crime during its commission pertains to the
right of coercion belonging to the classical ius strictum (the abstract
right or right in the strict sense).118 First and foremost, self-defense
pertains to this immediate cancellation of the crime during its commission. By self-defense the criminal is coerced. Hegel mentions this
type of coercion admittedly only briefly which, in the strict sense, is
neither 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 crimes commission, but also
115.
116.
117.
118.
119.
GPhR }} 846. Elements, ed. Wood, p. 117.

GPhR }} 97 and 99. Elements, ed. Wood, pp. 123, 124.
Cf. GPhR } 94. Elements, ed. Wood, p. 121.
138
as the present and past will of the criminal. Hegels presumption is that
the particular will of the criminal continues to be a criminal will
should it remain unpunished. In his lecture Natural right and the science
of state (Naturrecht und Staatswissenschaft) from 1818/19, Hegel
expressly says:
What the will does is, according to appearance, a single, but also a
universal something temporal and something intemporal. Thus the
consequences of the deed remain even if the deed is itself hidden.
Thus whoever steals remains for ever a thief, not merely in memory but
actually.120
And a little later he adds: the deed of the criminal is not merely
something transient, single, but instead something at the same time
universal, a law stating that infringement against somebody is allowed.121
Therefore, if the punishment does not occur then the criminal again
commits a crime. According to this, in order that right should not be
harmed 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 the
cancellation . . . 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 necessarily belongs to the primary aim of the cancellation of the crime. General
deterrence, however, is not only an indirect side effect of punishment
grounded in such a way, but also necessarily itself an aim of punishment. It means: If society is still inwardly unstable, punishments must
be made to set an example, for punishment is itself a counter-example
to the example of crime.123
This general deterrent aim of the Hegelian theory of punishment is
also indirectly derived from the main aim of the cancellation of the
crime. With the deed, the criminal accepts all consequences that he or
she can foresee. These consequences that are internal to the criminal
action, under which there is the risk that the deed moves other
citizens to commit a crime, pertain to his or her criminal intent. The
relationship between the criminals will and these consequences,
120.
121.
122.
123.
NRSW 2756.
NRSW 2767.
GPhR } 218, Addition. Elements, ed. Wood, p. 251.
139
therefore, is closer than the relationship between the criminals will

and the harm for which he or she is held responsible. The harm is
not necessarily foreseeable by the criminal: it partly pertains to the
consequences that are external to the criminal action. It follows from
the criminal not only being held liable for the harm caused, but also
being punished for consequences internal to the criminal action, that
one must coerce the criminal, to whom belong exclusively the consequences internal to the criminal action and not those external to the
criminal action. The consequences internal to the deed must not
necessarily occur; they do, however, pose a risk (Hegel speaks of
danger)124 that the criminal accepts. Hegel explains this through
an example in his lecture Philosophy of right:
A crime is dangerous, that is, it is primarily a crime, but it has a further
effect, a further context. The single action is in itself so more far
reaching, more universal than it appears according to its external
reality. If I hold a light to a piece of wood then the flame only touches
a 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 wood
that together constitutes a house this one with other houses that
constitute a city, and thus the fire can destroy a city. This universality
is the danger of crime.125
This danger internal to the criminal action must be strictly differentiated

from danger external to the criminal action, if one wishes to avoid a
consequentialist 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 other
consequences rather than in terms of the concept of the thing . . . Thus,
the crime which is more dangerous in itself . . . in its immediate
character, is a more serious infringement in its extent or quality.126
Belonging to the particular will of the criminal is the risk that the
crime comes to be an example, therefore a precedent for a third party.
The coercion exercised by punishment of the will of the criminal also
cancels 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. 2289.
125. PhR 279.
126. GPhR } 96, Remarks. Elements, ed. Wood, pp. 1223.
140
The various amounts of danger the criminal recklessly disregards

explain the various severities of punishment, depending on the era in
which the crime was committed. The severity of punishment conforms
to how secure the commonwealth is or what the risk is for destabilization of the commonwealth by imitation of the crime. A prestate society, in which the punishment of crimes is still a private matter, or a
weak commonwealth must both punish severely in order to preserve
themselves; a secure commonwealth does not have to: it should
punish more mildly. Of one thing Hegel is sure, however: the commonwealth can never be so sure of itself that it can allow itself to leave
a crime unpunished. Hegel writes:
But whereas it would be impossible for society to leave a crime
unpunished since the crime would then be posited as right the
fact that society is sure of itself means that crime, in comparison, is
always of a purely individual character, an unstable and isolated
phenomenon.127
That is because even if no other citizen imitates the criminal, the

criminal still will commit further crimes for as long as he or she is
not coerced to do otherwise.
Let attention be drawn to the fact that the aim of specific deterrence is not itself some undetermined sort of specific deterrence. Hegels
specific deterrence is always formulated negatively: cancellation . . . of
the crime, which would otherwise be regarded as valid; it would be
impossible for society to leave a crime unpunished since the crime
would then be posited as right.128 Another conception of specific
deterrence would consist in first assuming that the more severe a
punishment is, the more deterrent it is, and thus in demanding the
harshest punishment possible, in order to maximize public safety.
This conception is contradicted by Hegels specific deterrence.
On the contrary, the criminal should be sentenced to a milder punishment; punishments must simply be precisely so severe as not
to contradict the cohesion of the commonwealth or directly to endanger it. In this respect, Hegels theory of punishment is, in my view,

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 would
otherwise be regarded as valid as follows: that is, would have general existence, because
here a single Being is universal for everyone. (Translation mine.)
141
primarily a minimal specific deterrence and secondarily an equally

minimal 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 the
sole, essential addressee of the punishment, whereas, in the various
theories of general deterrence that operate either by threatening
punishment or by making an example of the criminal, all of society
is the addressee, including the criminal him- or herself, but only as
one among many. This reflexive structure of specific deterrence
explains why Hegel in } 101 speaks of retribution (Wiedervergeltung)
(more properly, though, of retaliation). We have seen above in Section 5.3,
however, that this retribution should not be confused with the theory
of retaliation, that is, with retributivism, which is sharply criticized
by Hegel.
The question arises, though, how } 100, which is considered in the
retributivist interpretations to be a subjective justification of retributivism and which is called an argument from law by Seelmann, is
compatible with the interpretation that I have suggested here.130
Hegels } 100 is actually occupied with the subjective aspect of
crime. This can be easily explained, however, through a specific
deterrence interpretation. The lecture Philosophy of right (1824/5)
deduces the subjective perspective expressly from the cancellation of
crime in the specific will of the criminal:
This existent will of the criminal . . . must be affected . . . This point is
linked to the fact that a punishment must be impressive on the criminal.
If punishment is not impressive then his existent will is not infringed
upon by it.131
Therefore, the subjective aspect does not occur, for instance as a subjective justification without which punishment would lack a rationale,
but instead the justification is deduced from the punishments aim of
specific 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.
142
Unlike Anselm Feuerbach, Hegel understands sensitivity to be not

only that which moves, for instance, the appetites and fear, but also
the criminals moral judgments.
Now, talion law is, according to Hegel, what everyone, and therefore the criminal as well, can sense the most: Jus talionis is a natural
concept in perception and was seen from time immemorial as being
an element of punishment.132 This aspect of crime that belongs to a
subjective moral validity is not to be confused with either a psychology
of the appetites or Hegels justification of the right of punishment.
The remark made by Hegel on } 99 makes this clear. In it Hegel clearly
speaks of three aspects:
1 Coercion exercised against the appetites by an evil or by a source
of evil (this correlates to Kleins theory of punishment that was
criticized by Hegel; in the Additions to } 99, Hegel also rejects
Feuerbachs 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 and
substantial point of view in relation to crime; [the rationale that]
punishment in and for itself is just; and we are solely concerned
with the need to cancel . . . crime not as a source of evil, but as an
infringement of right).
3 The moral feeling of the criminal (the moral point of view, i.e. the
subjective aspect of crime).
Hegel differentiates (1) from (3) by criticizing the consequences of
Kleins theory of punishment, which considers punishment as an evil:
it automatically follows that the essential consideration is now the moral
point of view, i.e. the subjective aspect of crime, intermixed with trivial
psychological ideas . . . of stimuli and the strength of sensuous motives . . .
as opposed to reason, of psychological coercion and of psychological
influences on representational thought . . . (as if such influences were
not themselves reduced by freedom to something purely contingent).133
The aspect of punishment that belongs to subjective morality is

defined through reason and freedom. Precisely, it is both of the latter
that make up both of the foundational elements of } 100. This
explains why Hegel writes:
132. NRSW 277.

143
But it is neither a question merely of an evil [which should move the

appetites] 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 objective
consideration of justice, which is the primary and substantial point of
view in relation to crime, is set aside; it automatically follows that the
essential consideration is now the moral point of view, i.e. the subjective
aspect of crime, intermixed with trivial psychological ideas.134
Hegel does indeed not recognize the retributivist dimension as being

the primary and substantial point of view, in other words as the main
justification of punishment, but he sees in it a necessary means to the
end 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 justification, this justification surely could not be found in the principle of
retaliation (Wiedervergeltung). We have already seen how Hegel attributes it to common sense that is entangled in contradictions: that
[the crime] deserves to be punished, and that what the criminal has done
should 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 conjunction
with the principle of retaliation in Abstract right:
It is incomprehensible how those sciences which derive their determinations from universal representations . . . should on other occasions
accept propositions which contradict such so-called universal facts of
consciousness.137
We find an example of that straight away in the first paragraph of

Morality, Purpose and responsibility, which leads to the differentiation 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 from
the consequences external to the action only indeterminate[ly],
because the inner necessity [consequences internal to the action]
136. Regarding morality: GPhR }} 105ff. Elements, ed. Wood, pp. 135ff. Cf. Hegels
handwritten comment on GPhR } 118: Superficial[:] it should go well for the
good ill for the evil. (Translation mine.)
144
comes into existence in the finite realm as external necessity,138 that

is, inner necessity appears as consequences external to the action.
Hegel writes:
The maxim . . . which enjoins us to disregard the consequences of our
actions, and the other which enjoins us to judge actions by their
consequences and make the latter the yardstick of what is right and
good, 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 has
less adverse consequences, just as the good action must accept that it
may have no consequences or relatively few; and it also follows that,
once the consequences of the crime have developed more fully, the
crime itself is made responsible for them.140
If one separates deed and action in an abstract way then this leads one
to break down (Zersplitterung) the consequences [for penal law]
minutely.141 With this, though, the possibility of retaliation for the
criminal action disappears in the judgment of common sense, and
only the deed can be retaliated for: Guilt or innocence in relation to
evils and of evils to guilt. Suffering in general, punishment amongst
it 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 cannot
perceive the subjective aspect and abstracts the criminal will, without
which there is no crime, but instead only an evil.
The principle of retaliation cannot be regarded, therefore, as being
justification of punishment, but as in Fichte143 is rather justified
only as a means to the end of punishment. As an external, superficial
principle, the principle of retaliation is especially suited, though, to be
understandable to the criminal: retaliation places an external, immediate identity between the crime and the punishment; this identity is
immediate because it does not operate through the concepts of right
and of law as well as through their necessary validity. Instead, this
immediate identity only relates to the material consequences of the
criminals deed for the victim and to the material consequences for
138.
139.
140.
141.
142.
143.
GPhR } 118, Remarks.

GPhR } 118, Remarks. Elements, ed. Wood, pp. 1456.
Handwritten comments on } GPhR 118. (Translation mine.)
GNR i/4 61. Fichte, Foundations of natural right, p. 228. Cf. above, Section 4.3.
145
the criminal in short, to the evil. Nonetheless, the principle of

retaliation suggests to the criminal a regular, non-physical identity
between his or her deed and punishment. Thereby, a first stage of
that conscious process is reached that will lead to rehabilitation.
Hegels rationale of punishment cannot be reduced to a mere
theory of deterrence, because he justifies penal law primarily neither
with specific deterrence nor with general deterrence nor with rehabilitation nor with the principle of retaliation, but instead with the
necessary self-destruction of the coercion or from its self-cancellation.
However, if one thinks in the usual categories, one must still observe
that Hegel regards specific deterrence together with rehabilitation as
the primary aim and as the primary justification of punishment,
whereas general deterrence represents a secondary aim, though a
nonetheless important one. Retributivism, on the contrary, delivers
neither an aim nor a justification of punishment.
In conclusion of this investigation of the Hegelian justification of
punishment, I must observe, as in my investigation of both the Kantian
and the Fichtean justifications, that if one proceeds not from the
departure point of the criminals intention or deed, but instead from
the departure point of the concept of right, which all three authors
have in common, retributivism cannot then be obtained. Since the
concept of right in all three authors (and correctly so) stands on a
higher level than their penal law, any exponent of this concept of
right will have to renounce retributivism.
PART III
RETRIBUTIVIST INHUMANITY
The theory of retaliation is not only incompatible with the concept of

right of Kantian philosophy (see Part i) and of German idealism (see
Part ii). Contrary to Kants claims, it even proves itself to be a theory
that respects the dignity in the criminals person less than the alternatives of deterrence. Most radically, Nietzsche excludes the concern
for human dignity both in the citizens and in the criminals person
from being a motive and a plausible aim of retributive punishment.
According to Nietzsche, the idea of retaliation allows no room for the
necessary internalization and reflection out of which alone bad conscience and remorse in the criminal as well as humane interaction
with society could arise (see Chapter 6).
In a deliberate move away from retributivism, I have made a plea in
Part ii from the standpoint of a combination of specific deterrence
and rehabilitation for a type of interaction that respects human dignity even with the most hardened and inhumane of criminals, that is,
with those who have perpetrated crimes against humanity. Claus
Roxin believes he raises a decisive objection against the conception
of specific deterrence by accusing it of not wishing to accept its final
consequences:
The decisive example is at the moment the concentration camp
murderer by whom some innumerable number of innocent human
beings were killed due to sadistic motives. These murderers are living
today mostly inconspicuously and socially integrated, so they are not in
need of rehabilitation. Also, the danger of recidivism from which they
must be deterred and from which we must be secured, does not exist
with them. Should they really remain unpunished because of that?1
1. Claus Roxin, Strafrechtliche Grundlagenprobleme (Berlin: De Gruyter, 1973), p. 7.
147
148
retributivist inhumanity
Roxin is implicitly operating under the assumption that there

is nobody who would answer yes to this question. But precisely
this answer appears to me to be the right answer, as I will attempt
to show by way of the punishment of crimes against humanity
(see Chapter 7).
6
NIETZSCHE AND PUNISHMENT
WITHOUT REMORSE
Kant does not consider that the aim of punishment is to arouse bad
conscience, and thus the criminals remorse as well. If punishment
were to be imposed in the service of such an aim, it would not occur
as Kant requires merely because the criminal broke the law, but
instead in order that a certain situation might come about. Since the
arousing of bad conscience or remorse seems to be a minimalistic aim,
which has moreover a clear reference to the criminals dignity, some
interpreters of Kant develop a conception of retributive punishment
directed toward this aim.1 If one proceeds under the assumption that
retributive punishment especially respects the criminals human dignity and that it can generate as Hegel points out (see Section 5.5)
even in the simplest mind a connection between crime and punishment, then one can expect the criminals remorse from retributive
punishment.
It is Nietzsche who best shows us that such a justification is not
self-evident but instead very problematic. Nietzsches polemic
writing On the genealogy of morality (1887) opposes such a moralizing
conception of retaliation and of punishment. From Nietzsches perspective, retaliation does not stem from concern for human dignity
in the criminals person but instead in the victims and other human
beings active Schadenfreude. Also, the expiation of punishment cannot
possibly be an inducement to moral conversion, but rather
strengthens the criminals felonious will. It is punishment that is
directed toward deterrence which proves to be the more humane
option of penal law.
1. Cf. chapters 9 and 11 of Thomas E. Hill, Human welfare and moral worth: Kantian
perspectives (Oxford: Clarendon Press, 2002).
149
150
In Nietzsches On the genealogy of morality, Essays ii 815 belong to

the context of the second tractate whose aim is an inquiry into the
genealogy of the feeling of guilt. Essay ii 8 begins with taking stock
of the paragraphs regarding the genesis of promises in GdM ii 13
as well as about the history of the genesis of Schuld (that is, guilt
or debt)2 and its acquitment in the form of making the insolvent
debtor suffer:
The feeling of guilt, of personal obligation, to pursue our train of
inquiry again, originated, as we saw, in the oldest and most primitive
personal relationship there is, in the relationship of buyer and seller,
creditor and debtor: here person met property for the first time, and
measured himself person against person.3
The further course of the investigation throughout these aphorisms

consists of three steps. First, Nietzsche gives a description of the
genesis of public punishment according to the canon of morals
relating to justice.4 Secondly, he then delineates the genesis of public
punishment in his rejection of Karl Eugen Duhrings conception of
justice and punishment as either ressentiment or revenge.5 Thirdly and
finally, Nietzsche takes it a step further: he radically separates public
punishment, whose genesis he has hitherto explained without reference to the criminals feeling of guilt, from this alleged feeling
of guilt.6 In the following I will mainly devote myself to the first part,
that is, aphorisms 810, because these sections depict the core of
Nietzsches conception of punishment.
6.1. The genealogy of public punishment

Out of the oldest relationship between human beings (according
to Nietzsche), that is, out of the relationship between buyer and
seller, 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 mean
either a financial debt or moral guilt. Nietzsche is, of course, keenly aware of its
polysemous 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 810, cited here is section 8. Genealogy of morality, ed. Ansell-Pearson, p. 50.
5. GdM ii 1113. Genealogy of morality, ed. Ansell-Pearson, pp. 528.
7. GdM ii 8. Genealogy of morality, ed. Ansell-Pearson, p. 49.
nietzsche and punishment without remorse
151
explained by Nietzsche in The philosophers book,8 the habit becomes a

generality. In this way, the universalization arises, which constitutes
the first level of justice or its oldest, most nave canon of morals:
Every thing has its price: everything can be compensated for.9
One reason why this canon of morals is nave is because it is still not
internalized, intellectualized (vergeistigt) to make use of Nietzsches
own vocabulary. In this passage, which is based on the aforementioned
principle of justice, guilt and punishment, moreover, are still investigated not in their internalized dimension, but instead only in the
external dimension of a harm and of punishing the one who caused it.
Not only the feeling of guilt, but also the internal guilt remains unconsidered in this explanation of the establishment of public punishment.
Because this canon of morals is still not internalized, it is also nave
in the sense that it openly compares power with power and defines
itself by power relations:
Justice at the first level is the good will, between those who are roughly
equal, to come to terms with each other, to come to an understanding
again by means of a settlement and, in connection with those who are
less powerful, to force them to reach a settlement amongst themselves.10
In short: justice should be comprehended merely in a positivistic

sense. It is not the just exchange that is defined through the equivalence of exchanged goods, but instead the equivalence that is defined
through the actual exchange of goods, which is decided, in turn,
through a power relation.11 Nietzsche, regarding this point, is following
Hobbes, who raises the following objection against those authors who
regard fair exchanges as an arithmetical ratio:
As if it were Injustice to sell dearer than we buy; or to give more to a man
than he merits. The value of all things contracted for, is measured by
the Appetite of the Contractors: and therefore the just value, is that
which they be contented to give.12
The conclusion from this, we will have to remember later on, is that
impartiality is impossible.
8. WuL i, 880. Daniel Breazeale, Philosophy and truth: selections from Nietzsches Notebooks of
the early 1870s (Atlantic Highlands, NJ: Humanities Press, 1979), pp. 7997.
11. Cf. Henning Ottmann, Philosophie und Politik bei Nietzsche (Berlin: De Gruyter, 1987), p. 131.
12. Part 1, chapter 15 of Thomas Hobbes, Leviathan, ed. A. R. Walter (Cambridge:
152
Just as in Hobbess Leviathan justice is defined by the Leviathan

alone, with Nietzsche the commonwealth is limited to the tasks of
protection, of peace and of ensuring social trust. According to this, the
criminal is an attacker (he also actually assaults the creditor), a man
without peace and somebody who has broken his contract and his
word therefore, someone who is not allowed to promise anymore.13
Nietzsches explanation of the genesis of punishment in Essay ii
910 of his Genealogy of morality shows us prima facie public punishments 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 of
a community grows, its penal law becomes more lenient)
4 impunity in times that are not impossible to imagine when
societys conscious[ness] of its power is especially developed.15
When we examine these four periods more closely then it becomes
apparent not only that a public punishment exists solely in the second
and third periods, but also that a legal and civil order can only exist in
those two periods. In the following, I will attempt to show why this is so
and to what function this fact gives rise for public punishment.
Essay ii 9, in a first period, takes the logical consequence out of the
determination that the criminal is an attacker, a person without peace
and somebody who has broken his or her word and contract. Punished the least is the immediate danger; instead it is, quite apart
from this, the word-breaking that is punished, the logical result of
which ought to be that the criminal should not be allowed to promise.
The criminal is namely more than merely an insolvent debtor: he or
she attacks the creditor, thus not recognizing what is owed. Therefore,
the question poses itself whether the creditor can reach compensation
in spite of this with him or her, as is the case with the simply insolvent
debtor, upon whom the creditor can inflict suffering because this
gives him or her pleasure. I will return again to this question.
Nietzsche prefers, however, prima facie another manner of thinking. According to Nietzsche, every member of the community is a
debtor who still has to acquit the debt: he or she must contribute to
153
domestic peace, to mutual protection and trust. Because the criminal

cannot do this, he or she cannot be a member of the community any
longer. The criminal is elend, that is, expelled, exiled and outlawed.
The same provisory conclusion was already reached, for example as
I mentioned previously (see Section 4.3) by Fichte: the criminal
would be expelled into the desert and everyone would be allowed to
shoot him like a dangerous wild animal. The death of the criminal,
and everything resulting from that exile, is no longer the concern of
the commonwealth.
At the same time, there are also elements of another logic to be
found in the same section:
1 elements of compensation: The community, the cheated creditor,
will make you pay up as best it can16
2 elements of mnemotechnic: the criminal from now on, as is fair . . . is
not only deprived of all these valued benefits, he is now also
reminded how important these benefits are17
3 elements of infliction of suffering, which in Essay ii 6 counted as
compensation for the unpaid debts.
Nietzsche summarizes:
The anger of the injured creditor, the community, makes him return
to the savage and outlawed state . . . Punishment at this level of
civilization is simply a copy, a mimus, of normal behaviour towards a
hated, disarmed enemy who has been defeated . . . in fact, the rules
of war and the victory celebration of vae victis! in all their mercilessness
and cruelty.18
The suffering is indeed compensation for the simply insolvent debtor.

If suffering is compensation for the criminal as well, then insolvent
debtors and criminals are treated equally, even though the criminals
debts fundamentally are qualitatively greater. This would point to a
relative powerlessness of the commonwealth and the polity in respect
to the criminal as well as to the fact that the compensation and the
infliction of suffering constitute not the punishment, but instead if at
all the mnemotechnic.19 The contrast between the anger of . . . the
community and its reaction is obvious: first, the community cast[s]
16.
17.
18.
19.
GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 50. Also, cf. GdM ii 5f (pp. 436).
GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, p. 50. Also, cf. GdM ii 3 (pp. 412).
GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, pp. 501.
Cf. especially GdM ii 3. Genealogy of morality, ed. Ansell-Pearson, pp. 412.
154
the criminal away from itself; only thereafter can any kind of
hostile act . . . be perpetrated on him.20
It is still unclear whose anger is directed with cruelty toward the
criminal. 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 be
perpetrated on him (the logic of infliction of suffering). The latter
formulation is vague, just like the following and last lines of Essay ii 9.
Does it concern two logical moments of public punishment or two actors
of punishment?
Either, the commonwealth initially expels the criminal (the first
moment) and then handles him or her with cruelty like a defenseless
enemy, whereupon one can imagine all kinds of dishonour and
torture (second moment). Or, the state (the first actor) expels the criminal and the individual citizens (the second actor) perpetrate all manner
of cruelties on the criminal one might name this as a modern
equivalent of lynch mob justice.
The more plausible answer certainly lies in the middle of these. The
commonwealth perpetrates these cruelties because the criminal has
been excluded and because the individual citizens are angry and
demand cruelties.
The difference between the commonwealth as such and the sum of
its members reaches its clear expression at first in the second period of
punishment. There Nietzsche writes: the wrongdoer is no longer
deprived of peace and cast out, nor can the general public vent their
anger on him with the same lack of constraint.21 The decision
whether the criminal should be expelled from the community or
can remain is decided by the anger of the community as such that
is, by the ruler and not by the general anger (allgemeine Zorn) of the
masses. In the first period, the ruler decides for the expulsion.
Whether the ruler carries out the cruelties in person or simply allows
them, whereupon they would be then carried out by the crowd, is
actually not the core issue: in both cases the general anger dictates the
punishment not the ruler and his or her anger. The general anger is
reactive, though: it is a ressentiment that results from powerlessness.22
In this case, powerlessness results assumedly from the impossibility of
22. Cf. GdM ii 5. Genealogy of morality, ed. Ansell-Pearson, pp. 43ff.
155
reversing the crimes effects. Anger is something different from a

readiness to render compensation. Even if Nietzsche thoroughly
regards the pleasure provided by the infliction of suffering as a compensation,23 it is possible that the angered individuals remain unsatisfied with the criminals suffering. For the victim, the pleasure from the
suffering is only a surrogate.
The fact that there is war between the crowd and the criminal
points in this direction. Since this war does not occur between states
and since the state cannot wage war with individual persons, it therefore can only be a war between the population and the criminal. A war
can, however, only occur between two communities; for that reason
here it concerns itself with only a copy and with even the mimus of
a bellicose attitude:24 the mimus was only a farce for the Romans. In
fact, 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 context
of this second period, one may assert that the first period is typified by
a diffusion of unrest.
The first period was, therefore, more a period of private justice. In
the second period, punishment is no longer determined by the crowd,
but instead by the commonwealth and polity as such. Only then are we
dealing with a public punishment in the proper sense; in contrast to
that, in conjunction with the first period Nietzsche writes punishment 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 the
individual quite so seriously, because these do not seem to be as
dangerous and destabilizing for the survival of the whole as they did
earlier.26
For this reason Nietzsche is here in agreement with Hegel punishments can be milder. Nowadays one would essentially interpret the
danger and the risk of collapse as Hegel does in } 218 of his Elements of
the philosophy of right: the more stable the state, the less the risk that a
crime incites further crimes and, hence, makes a stricter general
deterrence necessary. Unlike with Hegel, however, with Nietzsche this
23.
24.
25.
26.
GdM ii 5ff. Genealogy of morality, ed. Ansell-Pearson, pp. 439.

GdM ii 9. Genealogy of morality, ed. Ansell-Pearson, pp. 501.
GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, p. 51.
156
is not what the concern is; rather, the concern is about the risk that a
criminal might agitate the entire society or might bring about civil
war. In which case is this a lesser risk? Perhaps it is less if the quantity
and 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 being
a direct threat by the majority of the population. This is even more
probable, though, if the state possesses, in a stable manner, the real
monopoly on violence, which makes it possible for the state successfully to resist the anger of its citizens. This may be the definition of
a states grow[ing] in power.
The state puts the criminal under its protection against the general
anger, in order that the crime be totally acquittable and then truly
acquitted. Unlike an enemy, with whom one remains unreconciled
even after his or her defeat and death, public punishment should
provide for either the complete expunction or the complete acquittability of the debt contracted by the crime. In this regard, three
aspects matter.
First, the state must see to it that the wrongdoer is isolated from his
deed.27 The consequence resulting from this is that, unlike an
enemy, the criminal is not to be destroyed. In this manner, the criminal
cannot be reduced to his or her crime, and can later become again a
member of the commonwealth. Yet one should not forget the fact that
Nietzsche would in no way endorse this development. For he writes in
Essay i 13 that there is no being behind the deed . . . the doer is
invented as an afterthought, the doing is everything. Were one to
cite, as Michel Foucault did, a phrase from the French philosopher and
politician Gabriel de Mably as being the criterion for modern penal law,
in contrast to the pre-modern penal law that says punishment . . .
should strike the soul rather than the body then Nietzsche more
distinctly stands on the side of pre-modern penal law.28
Secondly, an equivalent has to be found for a deed isolated in this
way, or, more exactly, multiple equivalents have to be found29 each
one probably determined by the gravity of the deed. The equivalent,
therefore, can only be an illusion; (public) punishment can only be
28. Michel Foucault, Discipline and punish: the birth of the prison, trans. Alan Sheridan, second
edition (New York: Random House, 1995), p. 16. The original quotation is taken from
Gabriel de Mably, De la legislation, in uvres comple`tes (12 vols., London, 1789), vol. 9,
p. 326, and reads: Que le chatiment . . . frappe lame plutot que le corps.
157
based on an illusion. Punishment is not an equivalent, but only the

waiving of an equivalent.
Thirdly, this should result in a compromise with the anger of
those immediately affected by the wrong-doing. Nietzsche does not
further specify the contents of this compromise. One may guess,
however, that the compromise consists in inflicting suffering on the
criminal and declaring him or her as being inferior to the victim.
Indeed, in most of the penal theories, punishment is defined as the
bel),30 not merely as some kind of answer to
infliction of an evil (U
the crime (a definition of punishment without a constitutive infliction
of suffering is nevertheless possible; rehabilitation, for example, really
goes beyond compensation, but is not thought of as being an evil).
Already in Essay ii 5, Nietzsche writes that, in the case in which the
actual power of punishment, of exacting punishment, is . . . transferred to the authorities, the victim attains the elevated feeling of
despising and maltreating someone as an inferior.31 One should
not overlook the point that the state takes no pleasure in inflicting
suffering on the criminal or in his or her humiliation. The compromise can, however, partly consist in the state guaranteeing relative or
increased security.
The principle of acquittability (Prinzip der Abzahlbarkeit) still does
not say anything about the degree of punishment. The criminal can
render compensation by a mild punishment, as well as by death after
manifold tortures for longer durations. The principle of acquittability
only requires that every crime can be expiated or that any debt can be
acquitted 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 the
first time in the premises of the third period of punishment:
As the power and self-confidence of the community grows, its penal
law becomes more lenient; if the former is weakened or endangered,
harsher forms of the latter will re-emerge. The creditor always
becomes more humane as his wealth increases; finally, the amount of
his wealth determines how much injury he can sustain without suffering
from it.32
30. Cf., for instance, in RL } 49e, Ak vi:331ff. Practical philosophy, ed. Gregor, pp. 4725.
Also, see above, Chapters 13.
158
Nietzsche seems at this point to introduce, astonishingly, both

another principle of penal law and another psychological basis of it.
The principle of penal law no longer appears to be a compensation in
proportion with the power relation, but deterrence instead. Likewise,
the psychological motivation of the ruler no longer appears to be
the 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 could
allow itself the noblest luxury available to it, that of letting its
malefactors go unpunished. What do I care about my parasites, it
could say, let them live and flourish: I am strong enough for all that!33
At this stage even deterrence is dispensed with: the parasites are

allowed to weaken the community. The corresponding psychological
premise is no longer the will to power, nor is it even the search for selfpreservation and security, but instead it is only an indifference to the
world or the dissipation of luxury.
The alternation in the psychological premises conforms to
Nietzsches conception of the biological process of deployment and
exhaustion of energies. In Nietzsches thinking, the active energies do
not remain constant, but instead attempt primarily to assert themselves: they strive for power. However, at one point they will exhaust
themselves and then they will only tend toward merely reactive selfpreservation or even toward disappearance toward the nihil of nihilism. The concluding judgment of Essay ii 11 can be applied to this last
stage of biological development:
A system of law conceived as sovereign and general, not as a means for
use in the fight between units of power but as means against fighting in
general . . . this would be a principle hostile to life, an attempt to
assassinate the future of man, a sign of fatigue and a secret path to
nothingness.34
Not only in the contemporary debate over penal law does impunity
hardly count as a serious option.35 Also, those authors who are
not supporters of maximizing general deterrence being an aim of
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. 33651.
159
punishment share the view that total impunity would be a weighty

incentive for crime; so that at the very least, owing to this reason of
minimal general deterrence, no commonwealth can dispense with the
existence of punishment without challenging the existence of the
community itself. The true power of a state can never be so great that
it can dispense with every kind of punishment. In situations in which a
state enjoys a growing power but does not wish completely to abdicate
its security, the degree of punishment may decrease; at most, impunity
remains the ever unreachable horizon of this mitigation: impunity
points to a utopia. Nietzsche writes, in fact, not that it is not impossible to imagine society so increased in power, but instead that it is not
impossible to imagine society so conscious of its power (emphasis
mine). This power consciousness relates to no realistic perception;
rather, it should be understood as being a symptom of a biological or
psychological condition a condition of exhaustion.
No acquittal of debt takes place in the period of impunity; in this
respect, impunity is outside of the sphere of justice defined by the
principle of universal acquittability and comparability. The lack of
comparison and equivalence fulfills the definition of luxury.
A luxury is something to the price of which one pays no heed. In
the period in which the criminal was fought against like an enemy,
however, there likewise existed no justice because there was at that
time a state of war and the criminal was completely delivered to the
anger of the people: there was no system of acquittability in general.
According to Nietzsche, the deeper and the lower the position of
the creditor in the social scale, the more the creditor feels a lust to
inflict suffering as payback, because through punishment of the
debtor, the creditor takes part in the rights of the masters.36 The victims
of the criminal as well as the other simple citizens sharing the general
anger enjoy this right of the masters. The rulers who hold themselves
to be especially powerful and decide in favor of absolute impunity that
is, universalized pardon also enjoy this prerogative of the most
powerful man; however, he or she does this only on account of an
illusion of power. Both the simple, angry citizen and the indifferent, allpardoning ruler, on the one hand, are powerless, but, on the other
hand, they succumb to the illusion of power.37 In contrast, true power
consists in forc[ing] those who are less powerful [in this case, the
160
less powerful are both the criminal and his or her victim] . . . to
reach a settlement amongst themselves.38 In fact, only in the
second and third periods can a compromise be reached between
the interest of the criminal (impunity) and the interest of the victim
(a bellicose, informal cruelty).
6.2. Nietzsches appraisal of justice

The conclusion nearly a lesson! that Nietzsche delivers at the end
of his explanation of punishment seems at first glance to raise a direct
objection to my version. He writes:
Justice, which began by saying Everything can be paid off, everything
must be paid off, ends by turning a blind eye and letting off those
unable to pay, it ends, like every good thing on earth, by [canceling
itself]. The [self-cancellation] of justice: we know what a nice name it
gives itself mercy.39
With these lines of text, Nietzsche seems to see impunity as being the
logical conclusion of justice or acquittability, and therefore not as
being a deviation from strict justice. For Nietzsche speaks expressly
not of sublimation, but of the [self-cancellation] of justice (emphasis
mine). One would rather expect that mercy would be designated as
being an extrinsic cancellation of justice, being beyond justice, and
thus as it is according to Nietzsche himself, being beyond the law
( Jenseits des Rechts).40
Nietzsches charge against justice is that it seeks acquittal of debt
while nevertheless allowing those unable to pay to escape punishment.
There are at least two ways to understand this. Either, one sees the
turning of a blind eye to those who are insolvent pertaining merely to
the fourth period, that is, to absolute impunity. But then the charge of
self-cancellation is incomprehensible. Or, one understands it in such a
way that the turning of a blind eye indeed completely unfolds in the
fourth period, though it was already therefore, from the outset
present in the second and third periods. In this respect, even public
39. GdM ii 10. Genealogy of morality, ed. Ansell-Pearson, pp. 512. Translators note: in the
translation, Aufhebung is rendered as sublimation, but in the Nietzschean sense, as in
the Hegelian context, rendering it as suppression, cancellation or sublation
might prevent any confusion.
161
punishment as acquittal of debt means the turning of a blind eye; in

short, it means not truly acquitting the debt. In this context, one
ought to recall two points:
1 The criminal is a promise breaker, not only a malefactor, which is
why he or she loses all rights and safeguards and, furthermore, all
mercy as well.41
2 Nietzsche stresses that you live in a community, you enjoy the
benefits of a community (oh, what benefits! sometimes we underestimate them today).42
One can gather from both of these points, first, that the criminals
debt is qualitatively especially large and, secondly, that the creditors
assets are likewise qualitatively especially large. Now, Nietzsche never
himself adopts the principle of justice that everything can be compensated for [acquitted].43 On the contrary, nobler things cannot be
acquitted; and the institution of the state and its establishment
wholly rank among these nobler things, as is demonstrated, for
example, by Nietzsches interjection oh, what benefits! We are,
therefore, allowed to proceed under the assumption that for
Nietzsche the acquittability is not (at least not always) at hand. In
the second period, the acquittal, which is then public punishment,
thus means that the debt remains unpaid at least partly: the debtor
gets off relatively easily.
The following explanation of this paradox seems obvious: the
justice that leads to this renouncement of a real equivalent stems
from the ressentiment of the victim. Nietzsche constructs a contrasting
alternative theory to this theory held by Duhring. The latter regards
the connection of crime with ressentiment and with revenge in the
following manner:
In our moral deliberations, we have already regarded each action which
originally, 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 for
retribution or, precisely, revenge.44
41.
42.
43.
44.

Eugen Duhring, Cursus der Philosophie als streng wissenschaftlicher Weltanschauung und
Lebensgestaltung (Leipzig: O. R. Reisland, 18945), p. 224.
162
Now, Duhring understands justice to mean retribution (Vergeltung).

Hence justice rests on the need for revenge. Why, however, ought
this need to remain a ressentiment or a counter-sensation instead of
leading to the action of revenge? Duhring explains the rejection of the
vendetta by stating that it yields a private war that is instigated and
continued over and over again.45 In order to avoid a civil war, one
establishes a system of conflict resolution in virtue of which one seeks
appeasement of revenge by means of atonement and compensation.46 In return for forgoing revenge, the victim receives, in addition
to public peace, the guarantee that evils will be imposed upon the
criminal and that he or she will feel guilt and remorse:
But the high rates with which one allows ones own bodily injury and the
killing of dependants to be bought off after the fact should not lead to
passing over the fact that the readiness to make an earnest material
sacrifice could also vouch for the dignity of the changed will and, hence,
for true remorse and for a peaceful disposition. The need for revenge
does not disappear, however, only by subduing and harming the injurer,
but it is also balanced out if the malefactor himself assumes his
castigation honestly by humbling himself by the true confession of
guilt and by the accomplishment of an apology and a punishment.47
Nietzsche refutes this explanation in two steps. First, he shows that

justice and public punishment have different origins. Secondly,
according to Nietzsche, the punishment does not give rise to either
any feeling of guilt or any bad conscience in the criminal, but rather
the opposite.
Immediately, at this point, a new paradox arises. Should the punishment in accord with justice originate not from emotional reactions
but instead from active emotions,48 then how can the renunciation
of a real equivalent be explained on the basis of these active emotions? The answer lies in the aforementioned distinction between
two actors, that is, between the population and those who wield power.
In general, Nietzsche always responds to inquiries into the nature
or into the aim of an institution with a counter-inquiry into the
power relation out of which the institution arises. The attempt of
the person in power to localize the matter and head off further or
45.
46.
47.
48.
Duhring, Cursus der Philosophie, p. 225.

Duhring, Cursus der Philosophie, p. 225.
Duhring, Cursus der Philosophie, pp. 2256.
163
more wide-spread participation and unrest is implemented by the

punishment.49 This striving for civil peace, which Duhring also saw as
the ground for public punishment as compromise and renunciation
of retribution, is not a final aim in Nietzsches conception unlike in
Duhrings conception. In Nietzsches conception, the rule of law
and civil peace were not established for the sake of the safety of the
individual citizens, but instead as a means for ensuring power for
those in power. Nietzsche emphasizes this distinction:
A system of law conceived as sovereign and general, not as a means for
use in the fight between units of power but as a means against fighting in
general . . . [requiring] that every will should regard every other will as
its equal, this would be a principle hostile to life . . . a sign of fatigue and a
secret path to nothingness.50
Therefore, the states efforts at establishing a civil peace among

other things, by means of public punishment should be regarded as
a weapon. The prevalence of power can be internal as well as external.
The internal prevalence of power consists in the situation in which
individual human beings for instance, the victims no longer inflict
the punishment, but instead the state alone; thereby, the state
enforces what has generally been later termed the monopoly on
violence. The external prevalence of power consists in the fact that
civil peace represents a considerable advantage in struggles against
other states. Nietzsches praise of war as a principle that is favorable to
life finds repeated expression in his works.51
The result of the evolution from private revenge to public punishment first appears negative, though, for the population in general and
for 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, but
an ancient, powerful, human-all-too-human proposition.52 At this
point, the people lose the ability to make someone suffer (Leidenmachen) and receive as consolation only the ability to see somebody
suffer (Leiden-sehn) in which the suffering constantly becomes milder
and more discrete the more powerful the state becomes. The population
51. For instance, in Human, all too human (i } 224); Thus spoke Zarathustra (iv 312); and
Twilight of the idols (} 38).
164
experiences a clear frustration because the wrong-doer is [instead]

carefully shielded by the community from this anger, especially
from that of the immediate injured party, and given protection.53
Previously, private punishment meant tak[ing] part in the rights of
the masters,54 which disappears at this point. The fiction of equivalency 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 of
the remorse of the criminal that provides consolation to ressentiment.
This remorse, which Duhring holds both to be something real and
to be the aim of punishment, proves to be a mere means to the
prevalence of the (states) power for Nietzsche.
Nietzsche stresses that change of perspective that plays out between
the philosophy of right and the philosophy of penal law, on the one
hand, and his own conception, on the other hand. According to
Nietzsche, punishment has no purpose (Zweck); instead, the concept
of punishment represents a whole synthesis of meanings.55 Punishment refers neither to the purpose of the population (that is losing its
ability to make someone suffer) nor to the purpose of power (that is
limiting itself), nor does it effect any remorse in the criminal, as we
shall later see. Not long before Nietzsches On the genealogy of morality
(1887) appeared, two of the most significant penal law theorists
placed the purpose of punishment at the center of penal law: Rudolf
von Jhering and Franz von Liszt.56 All of the great outlines of theory in
the tradition, from Platos Gorgias to Hegels Elements of the philosophy of
right, through Beccarias On crimes and punishments and Benthams
Theory of legislation, attempted to justify the institution of punishment
with an end. In this respect, Nietzsche represents a radical break in the
theory of penal law; only a narrow bridge remains between Nietzsche
and the tradition: both approve of the institution of punishment.
Nietzsches conception, however, rejects every dimension of justice
and any justification of punishment as being mere illusions.
Admittedly one may tend toward the illusion that Nietzsche praises
justice when he writes: If it actually happens that the just man
53.
54.
55.
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).
165
remains just even towards someone who has wronged him . . . well,
that is a piece of perfection, the highest form of mastery to be had on
earth.57 In so doing, however, three basic aspects should not be
overlooked. First, Nietzsche is referring in no case to justice, but
instead to the just man. Secondly, the sentence is formulated hypothetically (If it actually happens . . .), and its condition is not fulfilled
since it is something which we would be wise not to expect and
should certainly find difficult to believe.58 Furthermore, Nietzsche
never takes an idealistic position such as that: Nietzsche categorically
disapproves of any idea whatsoever. Thirdly, the just character of a
human being who stands the closest to justice does not stem from his
or her appreciation for justice. The human who stands close to justice
is 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 is
primarily 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 or
that is a value simply does not exist in Nietzsches 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 and
only more skilled in crime (the punishment strengthens the power to
resist; we must certainly seek the actual effect of punishment primarily
in the sharpening of intelligence),60 or their wills are broken and they
become dulled (if it does happen that a mans vigour is broken,
resulting in his wretched prostration and self-abasement).61 A third
possibility is as good as excluded: the real pang of conscience, precisely
amongst criminals and convicts, is something extremely rare.62
In order to understand this, it must briefly be mentioned where
and in whom the feeling of guilt eventually arises. Nietzsche pointedly
57.
58.
59.
60.
61.
62.
GdM ii
GdM ii
GdM ii
GdM ii
GdM ii
GdM ii
11. Genealogy of morality, ed. Ansell-Pearson, pp. 523.

11. Genealogy of morality, ed. Ansell-Pearson, p. 53.
14 and 15. Genealogy of morality, ed. Ansell-Pearson, pp. 59, 60.
166
summarizes the explanation when he speaks, in Essay ii 22, of the man

or woman who has been incarcerated in the state to be tamed, and
has [invented] bad conscience so that he can hurt himself, after the
more natural outlet of this wish to hurt has been blocked.63 We have
seen how the state has monopolized the infliction of suffering by
public punishment. Now the state is exercising a monopoly of power
against 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 being
by the will to power and feels the infliction of suffering as a joy. And
the convict is just as well locked up as the normal citizens, if not even
more 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 the
judicial executive procedures inhibits the criminal himself from
experiencing his act, his mode of conduct, as reprehensible as such:
because he sees the same kind of action practised in the service of
justice and given approval, practised with a good conscience.64
At first glance, this explanation seems puzzling because it appears to

confuse the question of guilt and punishment with the question of
discrimination. The convict is deliberately discriminated from the rest
of the citizens because he or she has committed a crime. Normal citizens
are handled differently. This normative explanation is irrelevant in
Nietzsches conception, because normative elements do not provide
any explanation; instead, they themselves must be descriptively
explained.
Yet Nietzsches 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 heavily
oppressed than the normal citizens; unlike them, the criminal must
struggle against the power of the state so that he or she has no
necessity to require the internalization of what Nietzsche terms the
Leiden-lassen, literally mak[ing] someone suffer. Secondly, unlike
normal 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.
167
to go without a real equivalent (the consoling equivalent is only a

compromise) the criminal experiences something that can be held
to be an equivalent by him or her. The criminal is handled in the same
way he or she acted; with the punishment, debts are cancelled. The
suffering of the victim is not seen by the convict, since the state has
separated the convict from the victim as well as from the revenge of
the 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 to
the ideal of rehabilitation, prisons serve more to harden the antisocial
inclinations of criminals than to convert them to a law-abiding way of
life,65 admittedly seem to corroborate Nietzsches assessment of
imprisonment. Yet even if the convict is only extremely seldom
reformed, then, thanks to the punishment, at least the will of a
number of the criminals will be broken. At least some convicts will
therefore not be recidivists, that is, they will be rendered harmless or
rehabilitated thanks to the punishment after their term has been
served which is the goal of specific deterrence.
6.4. Does Nietzsche provide a theory of punishment?

What does Nietzsche himself then suggest? The acquittance (Abzahlung) rests on an illusion. I also do not believe that Nietzsches answer
would be expulsion from society. Neither the right of the masters
nor pardon can be an option. Although Nietzsche does not declare
himself to be against cruelty and against the joy that comes from
inflicting suffering, anger as a counter-sensation, as ressentiment, is also
not recommendable in Nietzsches perspective. One alternative
remains: mercilessness and cruelty against the hated, disarmed
enemy who has been defeated,66 but indeed without anger: rather
with joy, without reactive feeling, without ressentiment against the
person who has broken his or her word. The criminal simply
attempted to prevail by his or her breach of promise. Nietzsche has
no moral objections against that, because only the power relation can
be considered as being a norm. The criminal, whose crime is joyfully
sanctioned 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.
168
the internalization of ressentiment, that is, the internalization of the

anger of the powerless human being toward the criminal, is harmful
to the whole process of the prevailing of power, meaning symptomatic
for the triumph of powerlessness.
In fact, Nietzsche speaks of someone who has broken his or her
word not as a base, abject human being, but instead like a soldier in
a war, a metaphor that most of the time carries positive connotations
in Nietzsches works and refers to something noble. The criminal
rejects the system of equivalence and as a result considers him- or
herself as being incomparable. If the criminal fails, one may indeed
feel joy over the defeat, though one may not judge him or her morally.
Nietzsche had already mentioned this option more extensively in
the Second Essay:
Not so long ago it was unthinkable to hold a royal wedding or full-scale
festival for the people without executions, tortures, or perhaps an
auto-da-fe . . . To see somebody suffer is nice, to make somebody suffer
even nicer that is a hard proposition, but an ancient, powerful,
human-all-to-human proposition . . . No cruelty, no feast: that is what
the oldest and longest period in human history teaches us and
punishment, too, has such very strong festive aspects!67
It only begs the question whether one may even speak of a public
punishment under these conditions. Provided that the criminal is
punished not by private citizens for example, in the form of a
lynching, which, incidentally, displays more anger than joy but
instead by the state, one could, however, finally speak of a public
punishment without at the same time approving of Nietzsches view.
What does Nietzsches amoral view show us about public punishment as a feast of cruelty? Nietzsches argument seems bipartite and
two-sided. On the one hand, he has recourse to the classical justification of punishability: the criminal did not comply with the reciprocity
that is fundamental to the commonwealth, and he or she attacked
the commonwealth; thus, the criminal is no longer able to be a
normal member of the commonwealth. On the other hand, in
Nietzsches conception, the further classical premise is lacking that
the 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
169
deterrence and the various forms of retributivism allow except in

the 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: Nietzsches idea
of the most desirable punishment would be a mixed form of justice
and cruelty. Such a mixed theory would admittedly be implausible
as such.
In reality, the matter looks slightly different. Nietzsches primary
premise 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 of
the requirements of social life that it constitutes.68 It is not the
severity of the punishment, according to Nietzsche, that initially demonstrates the necessary and powerful aid to mnemonics;69 instead,
from the outset, it is the exclusion from society or the inacquittability
of the crime. Nietzsche cites as one of the means of mnemonics
only the death penalties with torture: stoning, breaking on the wheel,
impaling, drawing and quartering or trampling by horses, boiling
in oil or wine, flaying, cutting flesh from the body, the smearing of
honey 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 mnemonics71 that teaches strict
obedience.
Nietzsches conception of punishment as being a mnemonic is
concerned merely with the prevailing of power of the commonwealth
and leaves justice and the rule of law to the respective power relations
in whatever forms they may finally find themselves. A genealogy of
punishment is what Nietzsches explanation is; by the same token, it is
also a resolute recommendation for punishment. However, there is
surely no inspiration for a justice oriented debate about punishment
to be found in this recommendation. Apart from that, in our age it
would potentially not be an efficient mnemonic aid to implement
Nietzsches suggestion; instead, it would rather provoke hopefully
a great deal of resistance as well as major social unrest. If our legal
systems, with all their structural and functional problems, are more
stable than Nietzsches conception then his single core argument has
68.
69.
70.
71.

Cf. GdM ii 3. Genealogy of morality, ed. Ansell-Pearson, p. 42.
170
miscarried. Thereby our age would differentiate itself from

Nietzsches conception just as those two situations that Voltaire somewhat sarcastically described in respect to retaliative punishment were
differentiated 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 sometimes
have in a situation of extreme famine: you eat them no more once
you have bread.72
72. Chapter 14 of Voltaire, Commentary on the book On crimes and punishments, by a
provincial lawyer, in Voltaire, Political writings, ed. and trans. David Williams
(Cambridge: Cambridge University Press, 1994), pp. 24479 (p. 264).
7
WHAT IS THE PURPOSE
OF PUNISHING CRIMES
AGAINST HUMANITY?
7.1. The characteristics of crimes

against humanity
The concept of crimes against humanity fulfills two roles. On the one
hand, it is intended to remedy the loopholes in the international
legal system. On the other hand, it constitutes a new kind of crime,
that is, a kind of crime that entails characteristics absent from any
other sort of crime.
The first loophole that the concept of crimes against humanity was
intended to close in the positive international legal system was the one
that arose as a result of the impossibility of prosecuting not only a
crime committed against the combatants and against the civilian
population of the enemies, but also those committed against their
own civilian populations. In this way, the concept of crimes against
humanity extends the concept of war crimes to include new categories
of victims. The second loophole that was supposed to be closed was
the impossibility of applying this extension of the humanitarian international law to crimes already committed. The Nuremberg and Tokyo
tribunals, which were charged with prosecuting war crimes and crimes
against humanity committed by Germans and Japanese, respectively,
during the Second World War, prosecuted crimes committed before
their own institution, that is, before the concept of crimes against
humanity arose prior to their own founding, thus injuring the basic
legal precept of nulla poena sine lege in favor of the enforcement of a
minimal natural law legal framework in international matters. Interestingly enough, the International Criminal Court (hereafter ICC) does
not contain this retroactivity in its Statute, though the national penal
171
172
legislations contain it.1 A third loophole is only now being gradually

closed, whereby the ICC was given a jurisdiction encumbered neither
by a statute of limitations nor by the place where the crime has been
committed, at least in those states where the statute calling it into
existence was ratified. This jurisdiction is subsidiary, however, since
the ICC prosecutes exclusively either those cases that the individual
national judiciary authorities did not try, or those in which a clearly
inappropriate decision was made.
The definition of crimes against humanity seems totally able to
close these loopholes, at the very least for the following three reasons.
First, an international criminal tribunal such as the ICC, which is by
no means a supreme court but a court that has a narrower jurisdiction,
and which decides cases without any possibility of appeal, that is, as
the last resort, cannot be responsible for hearing those cases related to
the highest crimes that have already been heard and not grossly
unfairly handled by any domestic court. Not only is the definition
and the treatment of the gravest crimes (in most countries the gravest
form of homicide) very different from one country to another (take,
for example, American first-degree murder, German Mord and French
meurtre avec premeditation, which are quite different concepts and are
handled in very different ways); even if they were consistent across
national borders, the numbers of cases to be handled, even if only in
appeal, would widely exceed the capacity of a single court. Secondly,
because crimes against humanity are considered according to their
definition as systematic attacks, which implies that they are serial
crimes, they are the most significant crimes from a quantitative point
of view. Thirdly, more often than not such systematic attacks occur
either in a failed state or in a criminal state, that is, for the former, in a
state that cannot hinder such attacks, or for the latter, in a state that
commits such crimes itself. As a result of that, there is a special need
for the subsidiary intervention of an international institution. In
short, the concept of crimes against humanity is useful as a criterion
to allow only the most severe and the most unpunished crimes to be
tried before such international tribunals. Were it only for these very
reasons, a new category of crime would be required and it has been
established in the criminal justice systems of each nation as the mere
1. Retroactivity is still the main reproach addressed to the Nuremberg Tribunal; see Larry
May, Crimes against humanity: a normative account (Cambridge: Cambridge University
Press, 2006), p. 211.
punishing crimes against humanity?
173
consequence of its existence in positive international law in the form

of 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 harmonization
of criminal prosecution made possible by the establishment of a
global standard and of an appropriate priority rule for criminal prosecutions. Although these pragmatic reasons seem sufficient to justify
the existence of crimes against humanity a specific sort of crime in
international law I suspect, however, that some of the characteristics
of this new kind of crime, as well as the justification given for its
punishment, 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 systematic
attacks2 committed pursuant to or in furtherance of a state or
organizational policy to commit such attack.3 An individual criminal
destroying an entire village for personal reasons does not commit with
that act any crime against humanity. A crime against humanity is not
directly a crime against the institutions, as is high treason, abuse of
power, bribery, etc. Yet crimes against humanity are not merely violations of the rule of law, but instead an attack against the political
institutions themselves; among other reasons, because they clearly
intend to exclude entire groups from obtaining political representation and from the process of political deliberation, as well as possibly
to deprive them of access to due process before impartial courts. I do
not agree with Antoine Garapons qualification of this as the violation
of the droit davoir des droits (the right to have some rights),4
because any victim of an ordinary murder definitively loses this right
to have rights, although he or she is not necessarily a victim of a crime
against humanity. Crimes against humanity are crimes against basic
political institutions.
2. Crimes against humanity abandon the usual sanction structure
found in penal law between the individual crimes, as well as between
the individual punishments. Unlike the domestic criminal law of each
2. Art. 7.1 of the Rome Statute of the International Criminal Court (hereafter, Rome
Statute): www.icc-cpi.int/NR/rdonlyres/EA9AEFF7-5752-4F84-BE94-0A655EB30E16/
0/Rome_Statute_English.pdf.
3. Rome Statute, Art. 7.
4. Antoine Garapon, Des crimes quon ne peut ni punir ni pardonner (Paris: Editions Odile
Jacob, 2002), p. 134.
174
country, such varied crimes as homicide, kidnapping, torture and

rape are subsumed under Article 7 of the Rome Statute. According
to Article 26.3, committing, soliciting or facilitating the crime, or even
providing the means for its commission, are all considered to be
committing the crime against humanity. The applicable penalties
as defined by Article 77 are the same for all of these deeds, whereas
the amount of punishment greatly varies in the domestic legal contexts, according to which of these deeds was committed. Neither of
the two options for assigning the amount of punishment (either
imprisonment for a certain number of years or life imprisonment) is
in any way specifically related to one of the categories of crimes that
are treated separately in the basic penal codes of the individual
nations. We can see one of the consequences in the case of a person
A committing a crime incited by a person B. The nature of the
incitement by B, however, does not amount to creating a state of
duress for A. Ordinary penal law shall not always punish incitement
to a crime; in such cases it punishes only incitement to the gravest
crimes, and then it punishes them mostly less than the commission of
the crime itself. In the Rome Statute there is no mention of a mitigation of punishment for the incitement to a crime compared with
commission of a crime. On the contrary, the Rome Statute does not
contain anything that would prohibit a reverse sanction structure
(such as punishing someone less for a crime depending on the level
of participation of that person, which goes against the usual sanction
structure in most countries). In fact, a political leader who has commanded large-scale massacres without having ever killed anybody
himself may be subject to a severer punishment than a soldier
who has carried out his orders and is also declared by the court not
to have acted under duress.
3. Crimes against humanity are not subject to the statute of limitations, whereas in most countries even the highest crimes are subject to
the statute of limitations. Even in countries such as Germany, in which
murder is not subjected to any statute of limitation, most of the crimes
(rape, kidnapping, etc.) included in Article 7 of the Rome Statute of
the ICC are subject to the statute of limitations when they are not
committed in the context of a crime against humanity. First and
foremost, at the time Germany decided to exclude murder from the
statute of limitations, it did so precisely in order to ensure that
the prosecution of crimes against humanity that were not yet part
of the penal code could still be made possible; were this not the
175
intention at the time, murder would most likely have been subject to
the statute of limitations. In Austria, there is no statute of limitations
for murder, but there is a dramatic reduction of the amount of
punishment should the case be brought twenty years after the execution of the crime.5
4. The exclusion of crimes against humanity from the statute of
limitations is often considered as being morally justifiable by the
monstrousness that all crimes against humanity pose. This monstrousness is usually considered as implying two consequences. First, it
implies that those who commit crimes against humanity are supposed
to be monsters who deserve extraordinary punishment. Secondly, it
implies that the monstrousness that crimes against humanity represent should remain unforgettable, because guilt is allegedly incommensurable. Thus, it is held that punishment should have an
expressive function, that is, that it should represent an expression of
the inextinguishable, most deep-seated abhorrence of all humankind.
In this view, no punishment can fully fit the crime, when the crime is a
crime against humanity, yet an extraordinary punishment can at least
be regarded as an expression of the inextinguishable guilt of the
criminals. The monstrousness of some crimes against humanity, as
well as the impossibility of finding any commensurable punishment to
fit the gravity of the crime, is often employed as the justification for
the existence of the legal category of crimes against humanity, and
it implies not only the exclusion of the statute of limitations (as
mentioned in the first point), but also the elements mentioned in
the first and second points, though interpreting them in a peculiar
way. Indeed, the absence of the usual precisely codified sanction
structure (the second point) is explained in this context as the manifestation of the fact that the crime is so monstrous that it lies beyond
the reach of the sanctions in ordinary penal law and of any quantitative scale of guilt and amount of punishment. But what about point
one, that is, about the political significance of crimes against humanity
as it is treated by the argument that the monstrousness of crimes
warrants sanctions above those against conventional crimes? Those
who explain the specificity of crimes against humanity as resulting
from their monstrousness do not emphasize the endangerment of the
public institutions brought about by the commission of the crime;
5. Austrian Penal Code (Strafgesetzbuch), Sec. 6, Art. 57(1). See May, Crimes against
humanity, p. 216.
176
rather, they emphasize that committing the crime is a moral evil, that
is, the alleged inability of the criminal ever to belong again to the
political community under the rule of law.
7.2. An alternative conception of crimes against humanity

In most conceptions of crimes against humanity, the preceding four
characteristics create a complex that I would like to challenge.
I believe the first characteristic to be true, as well as the second, but
I would justify the second in a completely different way, and I will
entirely reject the third and fourth ones. Therefore, I will begin my
inquiry into all four characteristics with the first one.
Crimes against humanity are crimes against the basic political institutions of states governed by the rule of law. I would like to introduce
a distinction that is essential for determining the appropriate amount
of punishment for this crime: I shall distinguish between (A) the time
at which it is being committed and (B) the time after it has been
committed.
(A) We find in the tradition of natural law two rights that are
intended to provide a remedy to a situation in which a crime against
the very institutions of a legitimate political community is being
committed, that is, tyrannicide and the right of resistance. By definition, this situation is one in which the rule of law is not enforced, and
thus no action before a court is possible. Therefore, tyrannicide and
right of resistance cannot be codified by positive law. The German
Basic Law (Grundgesetz) declares: All Germans shall have the right to
resist any person seeking to abolish this constitutional order, if no other
remedy is available.6 Yet, for the reason just mentioned, the law cannot
organize the exercise of this right to resistance; and it belongs not to
the penal law, but to the constitutional law. Tyrannicide and right
of resistance are no punishment of the tyrant, but rather provide for
the tyrannicide a legal excuse for performing an action that would
be, under normal circumstances, a punishable offense. I would like to
discuss, however, the punishment of crimes against humanity.
What would then comprise the goal of punishing a person who is at
any given time in power and is committing a crime against humanity
while holding office? It seems to me that the purpose of punishment
can be either most importantly incapacitation, that is, in this case
6. Art. 20(4) of the German Basic Law.
177
the removal of the criminal who commits a crime against humanity

from power, or general prevention, that is, in this case, deterrence of
the members of foreign governments from committing a crime
against humanity. Yet, the prosecution under the charge of crime
against humanity, as it is defined not only in international law but
also in internal law with its already mentioned characteristics, is either
a superfluous or a useless tool. It is superfluous when it is possible to
indict and to depose a member of the government or an entire
government under charges derived from ordinary penal law, and the
parliament is not so corrupt as either to tolerate or to support the
crimes committed by the government. It is useless when either no such
procedures of destitution or impeachment are foreseen by the constitution, or the criminal members of the government exercise such a
power or threaten such violence that nobody will dare to prosecute.
Prosecution may be impossible not only for reasons of internal dysfunction, 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 them
even against prosecution of crimes against humanity. Since any
humanitarian intervention employing military force is prohibited,
except in the presence of aggression and with the consent of the
United Nations Security Council, international law seems clearly not
to head in the direction of a prosecution of crimes against humanity
while the criminals are still in office. The concept of crimes against
humanity, as it is understood by todays human rights law, does not
provide the tools necessary for the clearing of the legal obstacles
on the path toward intervention against a government that violates
massively the fundamental rights of its citizens.
In fact, if the arguments made in favor of an unconventional
punishment for crimes against humanity were to emphasize a deterrent effect at all, then it is the general deterrent effect that is supposed
to result from sentencing the criminal political leader after he or
she has left office. Effective prosecution and conviction of political
leaders allegedly deter other political leaders from committing crimes
against humanity. Yet, first, no evidence has ever been provided that
political leaders who see in crimes against humanity the only means
available either to promote political goals the righteousness of which
they may be truly convinced of or simply to remain in office can be
motivated to change their mind under the influence of the punishment of their foreign colleagues, although the motive for promoting
178
political goals they are deeply committed to or of remaining in office

is certainly very strong.7 Secondly, were the aim of punishment deterrence, punishment should then be imposed directly following the
commission of the crime, that is, while the perpetrator still occupies
office, and not after he or she has eventually been deposed.
(B) Let us now investigate the situation after the criminal has
been deposed. Once the criminal has been divested of power, the
preventive aim of punishment may be either: general deterrence
about which I have expressed my aforementioned doubts or specific
deterrence, which implies hindering the criminal in regaining power.
Traditionally, three options are available in the case of a leader who
has committed crimes against the institutions of his own state and has
been deposed (for example, in cases of conspiracy, high treason,
abuse of power, etc.), with each option corresponding to a characterization of the former leaders traits as a leader, as well as of the
situation in which he or she is presently. Let us assume that the former
leader poses an immediate and acute threat for the institutions,
because of having at his or her disposal as many supporters, as many
means, as many weapons, etc., as he or she sees fit in order to have a
chance to regain power. The first possible characterization implies
that the former leader would lose this ability to regain power again if
he were no longer to be present within the states territory. The
classical solution for the aforementioned: banishment. This solution
does not have anything to do with the suggestion made by Aaron
Fichtelberg to banish the criminal from all of human society.8 In fact,
banishment from all of human society ultimately amounts to
condemning a person to an all-too-likely death, like that of a dangerous, wild animal, as Fichte demonstrated. The second characterization
is that, even in exile, the criminal would remain as much of a danger
for the state as if he or she were actually present in the state. The
classical solution for the aforementioned: death. Finally, the third
characterization of the former leader is that, after being deposed,
the former leader no longer poses a threat to the states institutions.
The former leader is thus condemned in order to sever the ties
between the former leader and the people, as well as to demonstrate
that the succeeding government does not share the same intentions as
7. See Aaron Fichtelberg, Crimes beyond justice? Retributivism and war crimes, Criminal
Justice Ethics, 24, no. 1 (2005), 3146 (p. 34); on p. 35 he rightly says that the threat of
punishment might even be counterproductive.
8. Fichtelberg, Crimes beyond justice?
179
the preceding leader so that the crimes the earlier government committed can be strictly outlawed. But, in general, the condemnation is short
and a pardon is rapidly granted.9 If necessary, a punishment may be
added to some period of ineligibility for office. The old adage states that
most of the time in politics punishment is either death (the second
characterization) or a short prison term (the third characterization).
It is interesting to note that in positive international law none of
these three options is available to a tribunal such as those of the
former Yugoslavia, of Rwanda or of the ICC. Capital punishment has
been 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 for
crimes against humanity unlike the exception made for the statute
of limitations; the gravest crimes against humanity carry no greater
sanctions than the worst instances of first-degree murder. After the
Nuremberg and the Tokyo trials, capital sentences were no longer
passed by an international criminal court. Were capital sentences
allowed to be passed, most Western democracies would not participate
in those trials, because most of them are expressly prohibited from
extraditing to a court that can pass capital sentences even in cases
of crimes against humanity. Banishment is not an option because
the Universal Declaration of Human Rights (Article 9), the European
Convention on Human Rights (Article 7(1)) and the Final Act of
the Conference on Security and Cooperation in Europe (Helsinki
Accords of 1975, Principle 10) guarantee an individuals access to
his or her own country at all times. Admittedly, were the criminal
leader to be judged by an international court, rather than by a domestic jurisdiction of the country of which he or she is a national, the
sanction would then probably be carried out outside the country,
which may be considered as de facto banishment. Finally, the sanction
for the leaders of crimes against humanity cannot be a short-term
sentence, but must be either a life sentence or a long period of time in
prison. Two astounding points come to light when all of the aforementioned is considered as a whole:
1 The punishment for crimes against humanity radically diverges
from 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; see
Michael P. Scharf, Justice vs. peace, in Sarah B. Sewall and Carl Kaysen (eds.), The
United States and the International Criminal Court: national security and international law
(Lanham, Md.: Rowman and Littlefield, 2000), pp. 21336 (p. 189).
180
political institutions, even though both crimes are closely related by

their very nature.
2 The severity of sanctions handed down for crimes against humanity
does not exceed the most severe sentences already existing for
other crimes, whereas those other crimes, in most countries, are
subject to the statute of limitations.
I will come back to a discussion of the aforementioned observations at
a later point. At this point, I will consider punishment of executants
merely acting upon orders given to them by the politicians. I will again
distinguish between the situation during the commission of the crime
and the situation after the crime has been committed.
It seems obvious that during the commission of the crime no
significant number of executants, if any, can be prosecuted, if there
should happen to be a systematic state or organizational policy in
the framework in which they find themselves carrying out their criminal actions (Rome Statute Article 7.1). After the crime has been
committed, the most important specific deterrent measure that must
be taken is the dissolution of all criminal political organizations that
had planned the crimes against humanity, guided them and, finally,
carried them out. Yet, what might be the purpose of punishing the
numerous individual executants? As I have already mentioned, one of
the characteristics of crimes against humanity as is also the case with
war crimes is that, on the one hand, command cannot be employed
as a justification of the crime, and on the other hand, the commission
of crime is not considered as being more serious than mere solicitation of an offense, in short, implying that leaders and the executants
are all responsible for the crime against humanity.
Yet, this is in the realm of pure theory, since only a very small
proportion of the executants are prosecuted and convicted, and they
are usually sentenced to less severe punishment than the leaders,
moreover to sentences not any severer and often more lenient than
for ordinary crimes (one thinks of a case heard before the International Criminal Tribunal for the Former Yugoslavia in 1996, in
which Drazen Erdemovic, a multiple murderer, was sentenced to only
five years in prison owing to mitigating circumstances).10 Even voluntary 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 at
www.un.org/icty/erdemovic/trialc/judgement/erd-tsj980305e.htm.
181
Gestapo, has not been systematically punished, only the gravest cases
being prosecuted. The sanctions enforced on a few executants can be
really efficient neither from the point of view of specific deterrence
(a few isolated individuals alone are not able to rebuild the criminal
organizations to which they belonged), nor from the point of view
of general deterrence (were the probability to be sentenced extremely
low, the deterrence effect would not be sufficient to justify such
measures). A much more efficient solution, from the point of view of
general deterrence, would be a massive punishment of all executants,
were it possible at all.
If such a massive punishment would meet with resistance from
the criminal who was indicted and sentenced, that would threaten
the existence and stability of the state under the rule of law, and if the
kind of very limited moderate punishment that I have described
would allow it to restore the state institutions to the rule of law, all
theories of punishments, whether deterrent or retributive, would
prefer the latter option. However, there is a quintessential difference
between deterrent and retributive theories with respect to this solution. A retributive theory of punishment considers it as an exception
made to the law, which can be made only in order to protect the
existence of the law in a situation in which the enforcement of the law
is not possible. In the Doctrine of right, Kant evokes the situation in
which the number of accomplices involved in a homicide is so great
that the state, in order to have no such criminals in it, could soon find
itself 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, because
there is no external justice at all in it.11 According to Kant, who,
at this point, undoubtedly belongs to the retributivist school, the
malefactor in such a situation should be sentenced to a mitigated
sentence, yet not through the letter of the law, but instead through
a decree of pardon issued by the monarch. For a theory of deterrence,
these more lenient punishments provide no exception to the law of
criminal justice, but are instead part of it, because penal law must, as
its sole purpose, ensure the enforcement of the rule of law as quickly
as possible. In this regard, once the danger has passed that the
malefactor who committed crimes against humanity might regain
power under normal rule of law, the probability of recidivism will
disappear. The probability of recidivism is much lower than with other
182
crimes, and because of that the punishment should likewise be much

less severe than what is usually prescribed by the normal penal code.
I will come back later to the retributivist position. The conclusion
that I will first draw from these points may be perceived prima facie
as repugnant. Because the characteristics that typify crimes against
humanity are formed by their political dimension and by the lack
of the usual sanction structure, they ought to lead to more lenient
punishments than do ordinary individual crimes. General and
specific deterrence simply do not require the same means, in cases
involving crimes against humanity, as cases involving ordinary crimes
perpetrated on a more individual scale.
This conclusion might well become even more repugnant were we
to carry out the following thought experiment. At this point, before
going any further, I would like to stress the point that I have absolutely
no sympathy for the person, or the ideas, mentioned below, whose
policies especially his policy advocating wholesale extermination of
several ethnic, political and religious groups rightly deserve the
deepest moral condemnation and the most vehement rational rejection. But I would like to make reference to the reasoning used by the
French defense attorney Robert Badinter, who finally obtained the
abolition of capital punishment after he became the French Minister
of Justice, after having struggled against it for his clients before
numerous criminal courts. Badinter explained that the more antipathy a criminal case provokes, the more clearly pleas may be made on
the grounds of principle.12 Let us begin our thought experiment by
imagining that the New York Times discovered, over forty years after the
close of the Second World War, that Adolf Hitler did not die in his
bunker in Berlin in April 1945, but instead escaped and ever since had
been living inconspicuously somewhere in the Irish countryside.
Except for retributivist arguments, I see no criminal justice theoretical
argument that could be presented to demonstrate that he should still
be punished after forty years; at that point, he no longer poses a threat
to any political institutions. In order to be absolutely sure that he
would not pose any further threat, one could consider depriving him,
at the very most, of his right to free speech. Drawing this repugnant
conclusion, that is, pleading for the imposition of the statute of limitations, even in cases involving the most despicable criminals, contradicts
one of the aforementioned characteristics of crimes against humanity,
12. See Robert Badinter, LAbolition (Paris: Fayard, 2000).
183
but it does not necessarily mean making a plea for the lifting of the
finding of guilt, as demonstrated by the Commission for Truth and
Reconciliation, a body that was intended to establish the facts but not
to make any condemnations.13 However, I suspect that a certain
amount of time in prison would have been necessary to rehabilitate
some very serious criminals of the apartheid era because they hold the
most firmly rooted convictions. I will confine myself only to forwarding the thesis that in cases of crimes against humanity the conventional prescriptive period should never be exceeded, but instead even
reduced.
7.3. The boundaries of our moral intuitions

As I mentioned above, a rejection of this repugnant conclusion can
only rely on a retributivist theory of punishment. Such a theory typically refers to the fourth characteristic of crimes against humanity that
I mentioned in Section 7.1. Retributivism rejects my repugnant conclusion, because the punishment can never be equal in gravity to the
crime; this is, for instance, Hannah Arendts thesis. Clearly, the monstrousness of all crimes against humanity is apparent, but the monstrousness of the crime cannot define a wholly new category of crime.
A cannibal who has eaten his consenting victims is, in my view, more
monstrous than someone who kills for stealing a million dollars. Yet,
the latter will be sentenced to a more serious punishment. The life
sentence will be (for instance, in Germany) the same as for someone
who tortured and killed a baby before the eyes of its mother, which
most of us will find more monstrous. And there are many actions that
are monstrous and unpunishable. Someone may cynically declare,
while eating in a luxury restaurant, that what makes him particularly
enjoy the meal is the thought that so many useless people are dying
of starvation somewhere in Africa. Or someone may declare to her
dying life partner at the hospital that she never loved him, had
betrayed him, and then leave without even saying good-bye. Both are
monstrous actions, though fully legal and protected rights. Whether a
homicide, a rape, a kidnapping, a manslaughter, etc. is monstrous is
neither a criterion for ascribing it to the category of crimes against
humanity, nor an element of the crime itself. The purpose of penal
law, as is the purpose of the legal system in toto, is not to retribute moral
13. See May, Crimes against humanity, p. 239.
184
evils and perversions, but to protect society, that is, the rights of the
citizens. This implies that penal law amounts to a means to restore the
rule of law between all persons including the criminals as quickly
and as efficiently as possible.
If we were now to formulate the retributivist objection to my repugnant conclusion in the guise of a classical criticism, we should say
that the retributivist would postulate that crimes against humanity
are deserving of severer punishment than similar crimes not barring
widespread and systematic attacks14 committed pursuant to or in
furtherance of a State or organizational policy to commit such attack.15
However, I see several serious difficulties with such a view being utilized
as a justification for disregarding the statute of limitations.
1 Were we to compare a homicide committed within the framework
of crimes against humanity with an ordinary homicide perpetrated on the individual level, we would be hard pressed to postulate
the former as being more deserving of severer punishment. Yet,
should a campaign of discriminatory, forced sterilization be punished more severely than, say, an ordinary first-degree murder? The
answer is not at all obvious.
2 Should a crime against humanity consisting of a massive extermination of human beings utilizing industrial methods be punished
more severely than a conventional first-degree murder? Were the
latter already punished by a life sentence, I see only a death sentence, or more likely a long and public torture, as a severer punishment. Yet, would that not constitute in itself a crime against
humanity? The letter and the spirit of international law punishing
crimes against humanity are clearly not shaping its course toward
such extreme punishments. (For this reason I disagree with Larry
May, who sees proportionality as requiring restraint in punishment
in relation to emotional factors: I see it, on the contrary, as
heading in the same direction.)16
3 Retributivism is not compatible with the lack of an outline of the
traditional sanction structure in Articles 7 and 25 of the Rome Statute.
One may object to my repugnant conclusion stating that penal law
should 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.
185
the harbor of the statute of limitations would severely offend popular

sentiment at least as much as not punishing crimes against humanity
more so than conventional crimes. One can understand this argument
in one of the following two ways. Either this argument posits that no
moral judgment can be morally true, were it to contradict popular
sentiment. Or the argument posits that such a repugnant conclusion
would lead the people to overthrow the judiciary power and possibly
even the rule of law.
Popular sentiment cannot be allowed, however, to be a criterion
for the moral judgment of punishment, among others things
because it is inconsistent in many regards. Popular sentiment mixes
retributive, general and specific deterrent elements. It can protest
against punishments that seem too lenient, but it can allow the victim
of a car accident to receive much greater compensation than the
victim of rape or of serious physical injuries perpetrated by an insolvent criminal. Furthermore, common sentiment is more often prone to
immoral judgments. For centuries it authorized criminal trials and
sentences against animals or against witches; it authorized torture
either as a punishment or as a method for interrogating a suspect;
it authorized the punishment of an entire family for a crime committed by one of its members. All this and more was fully acceptable to
the common sentiment. Nowadays, death by starvation of millions
of people is still not perceived in common feeling as a violation of
basic human rights, although such texts of international law as
the Universal Declaration of Human Rights (1948) rightly recognize
subsistence as a fundamental human right. Furthermore, in many
countries, a majority of the citizens is still in favor of capital
punishment, as well as of increasing restrictions on the right of
asylum.
Were one to consider the risk of the people possibly overthrowing
the institutions of the judiciary, as well as the political institutions,
if crimes against humanity were to be subject to the protections of
the statute of limitations, there would be no evidence in favor of
this view. South African institutions of state have suffered no harm
to 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 of
the apartheid regime by the new government there would have been a real threat of
military upheaval in South Africa: see chapter 6 of Amy Gutmann and Dennis
Thompson, Why deliberative democracy? (Princeton: Princeton University Press, 2004).
186
committed crimes against humanity have never endangered the

democratic rule of law.
For these reasons I see no valid moral objection against applying
the statute of limitations and against mitigating punishment. Though
the existence of the concept of crimes against humanity is a pragmatic criterion for judging the gravest crimes, and though it is the
necessary first step on the path to the harmonization of criminal
prosecutions by setting a global standard and an appropriate priority
rule, these pragmatic reasons seem to me not to justify the existence of
crimes against humanity as a fundamentally new type of crime that
requires new types of punishment and goals of punishments, as well as
justifying 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 especially
monstrous. This paradox reminds us that it is not the role of penal
law to take revenge on criminals for our past powerlessness. Rather,
the role of penal law is to protect against future crimes and to
secure the rights of the victim as well as of those of the perpetrator
within the rule of law. The victims of crimes against humanity might
well be more fortunate if governments who self-sufficiently congratulate themselves for having created the ICC which I certainly find a
most useful institution of global justice actually developed the right
of asylum rather than reduced it. The central issue should be not
whether or not Milosevic should be sentenced to life in prison or to
only four years behind bars; instead, the central issue should be how
best to wipe torture off the face of the earth.18
18. In this regard, Andrew Altman and Christopher H. Wellman, in A defense of
international criminal law, Ethics, 115, no. 1 (2004), 3567, confuse the role of the
punishment of criminals against humanity with the role that ought to be played by
military humanitarian intervention.
CONCLUSION
In Part i, we examined the Kantian, retributivist thesis, according to

which the criminal is punished, because he or she merits it. Two interpretations of the Kantian concept of right are possible (see Chapter 1).
On the one hand, according to the liberal interpretation, which does
not refer to the internal disposition of human beings, right is defined
merely as being the coexistence of the freedom of action of all
human beings according to the principle of equality of right. On
the other hand, according to the moral interpretation, which substantially refers to the internal dispositions of human beings, right
should implement the content of the categorical imperative as far as
possible through the application of coercion. In accordance with both
of these interpretations, two dimensions can be distinguished in the
Kantian retributivist thesis: one genuine dimension of legal ethics,
which is a dimension independent of the internal dispositions of
individuals (see Chapter 2), and a dimension of personal ethics (see
Chapter 3). The dimension of legal ethics of the Kantian retributivist
thesis contradicts the requirement for the coexistence of all freedoms,
of which the freedom of the criminal is a part too, and of which it
ought again to become a part at least in the most direct way. The
dimension of personal ethics contradicts the postulate of the highest
good, which rather requires forgiveness toward the criminal as long as
this forgiveness is compatible with the concern for the safety of the
other citizens. Hence, Kants concept of right as well as his concept of
virtue require rehabilitation instead of retribution, which ought to
occur 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 a
rehabilitation occurring after a period of specific deterrence was
187
188
conclusion
supported on the basis of the liberal interpretation of Kants concept

of right by Fichte (see Chapter 4) and Hegel (see Chapter 5): the
criminal is punished, in order that he or she might once again become part
of the commonwealth. Fichte begins with the question of the treatment
of the criminal and by a thought experiment initially draws the
logical conclusion of crime. The logically consistent consequence
of it would not be retaliation, but instead the exclusion of the
criminal from the commonwealth and the revocation of his or her
legal status. The public punishment derived from right as such
emerges as an alternative to this fate that is better for the criminal
so long as the alternative punishment amounts neither to the death
penalty nor to torture but rather to a sentence of incarceration. Such
an expiation, therefore, is not an evil, but rather a chance, which
Fichte constructs in such a way that the criminal can be reformed
and then finally reintegrated into the commonwealth. Hegel, in turn,
chooses as his guideline not the treatment of the criminal, but the
reestablishment of right as the negation of the negation in which
crime is the negation and the punishment is the negation of the
negation. In so doing, the negation of the negation is in no case, as
it 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 the
evil 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 in
conveying to the convicted criminal the significance of the punishment.
In Part iii, it finally turned out that retributivism could often demand
punishment without any respect for humanity in the person of the criminal.
Nietzsche offers a plausible account of the genesis of retaliative
punishment (see Chapter 6), in which the motive of the institution
of 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 conscience
and remorse in the criminal. When the state establishes a monopoly
on violence and guarantees public security, the people are deprived
of the exercise of this cruelty and punishments become comparatively
mild, such as those we have known since the end of the eighteenth
century. In order to keep the promise I made in the introduction
to draw conclusions resulting from the rehabilitative alternative to
retributivism, I finally argue (see Chapter 7) for a treatment of even
conclusion
189
the most serious and most inhumane criminals in other words

those who have perpetrated crimes against humanity that is no
more severe than how other criminals are treated. Even these
cases ought to be subject to the statute of limitations: the degree
of punishment ought not to be higher which is to say, proportional
to the crime but ought, instead, to tend to become less; the
idea ought to be abandoned that the monstrousness of such crimes
justifies the punishment and indeed, the most severe possible
punishment.
Further conclusions can be drawn from my critique of retributivism
as well as from my alternative proposal. At this point, I would like
briefly to outline how the penal system should be reformed according
to my alternative proposal.
1. There ought to be no fixed duration within the judgment for
incarceration. Instead of designating a certain duration, the degree of
punishment ought to be determined by a goal. The criminal ought to
expiate the punishment for as long as he or she still poses more of a
danger to the commonwealth than the other citizens do. In order for
that to occur, the criterion employed to suspend punishment may
not be the absolute safety of society, because every normal citizen
who consistently remains true to the law could, at any time, commit a
crime. Even now, punishments are reduced by sentences of parole,
and in the case of good behavior prisoners are released before their
sentences are up. Others are allowed to work outside the prison
during the day. Such leeway, already customary nowadays, ought
to be expanded to the general abolition of every designation of
duration.
2. The principle of proportionality between crime and punishment
has to be rejected. Admittedly, it is to be expected that, on average,
thieves can be rehabilitated in a shorter period of time than multiple
murderers who have tortured their victims. One would, therefore,
probably observe a clear, statistical correlation between the gravity of
the deed and the length of the process of rehabilitation. But this link
should not be a matter of principle, and there would certainly also be
exceptions.
3. A reform of penal law ought not to involve punishments becoming generally severer or generally milder. In effect, some punishments
would become milder, others severer, all depending on the behavior
of the individual convict.
190
conclusion
4. Psychological assessments of the convicts should occur regularly

and they should be thorough. A process such as the following is
conceivable. At the beginning of serving a sentence the criterion for
a discharge would be especially strong. The criterion would weaken as
time passes and would finally reverse so that gradually the burden of
proof might stand on the side of the commonwealth to prove that the
convict must, in fact, remain in prison.
5. Retributivism likes to contend that only retaliation is just for the
victim and every other punishment would be unjust, for the criminal
as well as for the victim. In the current penal system, this argument
rather appears to be an empty consolation for the fate of the victim
that turns out to be scandalous. Anyone who suffers grave bodily
injury is better off having it occur in the course of an automobile
accident for which he or she is not liable, rather than by a serial
killer, because the damages awarded to victims of criminal acts are
especially low. One should make it clear to the public what a criminal
trial 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 as
a whole regarding public security. Therefore, a twofold process is
conducted, even if both concerns are to be dealt with in the same
trial. Instead of exclusively and systematically calling for harsher
punishments as a way to win popular support, politicians should
primarily think of a better and more respectful compensation for
the victims.
6. An attempt ought always to be punished in the same degree as
the crime itself. This is not the case nowadays. Instead, attempts at
only some of the graver crimes are punishable, while carrying a lesser
degree of punishment. Feinberg convincingly traced back this situation to the conjunction that retributivism wants to create between the
gravity of a deed and the wickedness of the criminal:
I find no intuitive plausibility at all in first basing criminal liability
on moral blameworthiness, but then basing moral blameworthiness in
turn upon the actual harm or absence of harm caused. The ancient
view, that liability should be based not upon blameworthiness at all but
instead directly upon the amount of harm caused, seems to me more
honest, though no more plausible. It seems almost as if the retentionist
is so fixated on actual harm that he keeps searching for the question
to which it is the right answer. Not the question: What ought to be
the basis of criminal liability? Moral blameworthiness is a plausible
conclusion
191
answer to that question. Not the question: What is the basis of moral
blameworthiness? The traditional multiplicity of culpability conditions,
motives, mitigations, aggravations, and so on, answers that question. How
about 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 the

inner, 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, and
not of morality.
Beyond these concrete suggestions for the reform of penal law, as
the result of my inquiry I must address four points of more general
interest.
First, penal law cannot be about the (purely moral) wickedness of
human beings, but instead only about the observance of the laws or
about the infringement of right. Even if we were all devils,2 as Kant
notes in Toward perpetual peace, it would still not mean that we all ought
to be punished with deterrence in mind, before having committed any
misdeed. Only those who cannot be hindered from committing
crimes by moral or external reasons should be punished.
Secondly, it is not retributivism, but the justification of punishment
as a rehabilitation that is compatible with that path via which Kant
wants to reach global perpetual peace, the condition by which the
state of law would reign over the entire world. The modern tradition
of natural right that Kant critically draws upon recognizes three just
intentions for waging war: legitimate self-defense, recovery of ones
own goods stolen by the enemy and punishment of an unjust enemy.3
This punishment of the vanquished enemy is intended to mean
deterrence by the theorists of just war for instance, Hugo Grotius,
Samuel von Pufendorf, Emer de Vattel, etc. The aim is a peace that is
just toward both parties. In this respect, punishment is allowed only in
so far as to bring about security for the victor. This theory pursues the
goal of ensuring that no future injustice is carried out that would
present a just ground for a further war. If this deterrent intent turns
out 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: Oxford
University 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. 3956.
192
conclusion
number of wars. Since the victor has already obtained supremacy,

the punishment is often mild. Herein lies the strategy of the theory
of just war for the preservation of peace.
Let us now imagine what consequences a retributivist conception
of punishment might have for this theory. If the vanquished enemy
should fail to understand that the victors case is just, the retributive
punishment by the victor will be perceived by the vanquished as
being unjust. Unlike in national law, in international law there are
no higher-ranking judges, so that every sovereign is his or her own
judge, and although there cannot be two parties whose suits are
ultimately equally just, there can be two bona fide parties waging war
whose reasons for war appear completely legitimate and prima facie
just. Unlike a punishment that is meant to deter and to protect against
the recurrence of the grounds that led to war, retributivism requires
the vanquished not only to assume liability for all damages caused in
the war (compensation, reparations, etc.), but also to serve a severe
punishment. If the vanquished enemies waged the war in good faith,
this retributive punishment will then be perceived by them to be
unjust and thus may consitute a possible causa iusta for a future war.
For this reason, such a retributive punishment stumbles into a contradiction with Kants primary objective, that is, with perpetual peace.
Kants first preliminary article deprives every retaliation of its prerequisite by requiring that possible discoveries about the past should
not be taken into consideration:
Causes for a future war, existent even if as yet unrecognized by the
contracting parties themselves, are all annihilated by a peace treaty, no
matter how acute and skilled the sleuthing by which they may be picked
out of documents in archives.4
Even though Kant had the material claims of the warring parties
in mind in this passage, this remark certainly also applies in the realm
of penal law. Even in the Doctrine of right, Kant recommends the same
strategy for (civil) peace on the national level, as the following repeatedly cited passage shows:
Accordingly, every murderer anyone who commits murder, orders
it, or is an accomplice in it must suffer death; this is what justice, as
the idea of judicial authority, wills in accordance with universal laws that
are grounded a priori. If, however, the number of accomplices (correi)
4. ZeF Ak viii:344f. Practical philosophy, ed. Gregor, p. 317.
conclusion
193
to such a deed is so great that the state, in order to have no such

criminals in it, could soon find itself without subjects; and if the state
still does not want to dissolve, that is, to pass over into the state of
nature, which is far worse because there is no external justice at all in
it . . . then the sovereign must also have it in his power, in this case of
necessity (casus necessitatis), to assume the role of judge (to represent
him) and pronounce a judgment that decrees for the criminals a
sentence other than capital punishment, such as deportation, which
still preserves the population. This cannot be done in accordance with
public law but it can be done by an executive decree that is, by an act of
the right of majesty which, as clemency, can always be exercised only in
individual cases.5
Here abiding by the law is set aside in favor of preserving civil peace for
the very reason of preserving it, because civil peace is the sole prerequisite for a state of law in which laws can be issued and enforced.
This problem, to which retributivism is exposed within the Kantian
philosophy of right, is actually related, thirdly, to a fundamental
problem: retributivism is not compatible with an important principle
of the Kantian philosophy of right, that is, with the permissive law
(see Sections 2.4 and 4.4). The permissive law allows an exception
to be made to laws implementation, so long as this exception only
concerns what has occurred up to then and not that which either is
occurring in the present or will occur in the future. Such an exception
should provide for the peaceful establishment of a state of law where
the application of the law to the existing situation would lead to
regression to the state of nature or into civil war. However, the implementation of the law in the present and in the future is not allowed to
be postponed ad calendas graecas,6 but instead should occur in the
most direct way possible.7 Now, Kant defines right as being the sum
of the conditions under which the choice of one can be united with
the choice of another in accordance with a universal law of freedom,8
by which the power of choice (Willkur) of all human beings therefore, also of the criminal is meant. Thus, if there is in fact a way
enabling one to reintegrate the criminal into society in the most direct
6. ZeF Ak viii:344f. Practical philosophy, ed. Gregor, p. 321.
7. Cf. RL Ak vi:247. Practical philosophy, ed. Gregor, pp. 4012. Also, cf. Reinhard Brandt,
Das Erlaubnisgesetz, oder: Vernunft und Geschichte in Kants Rechtslehre, in Brandt
(ed.), Rechtsphilosophie der Aufklarung (Berlin: De Gruyter, 1982), pp. 23385.
194
conclusion
way, then this way is commanded. If this way actually exists at all, then
it is, per definitionem, rehabilitative punishment.
The categorical imperative of right consists not only in an ideal
legal system, but also in the steps that are necessary for either the
establishment or the reestablishment of this legal system. Writings
such as Toward perpetual peace and Idea for a universal history with a
cosmopolitan purpose belong to the latter dimension of the categorical
imperative of right. Even though } 49e devoted to penal law is an
editorial section of the Doctrine of right, which has to do primarily with
the ideal legal system, and therefore with the first dimension, penal
law can be categorized into the second dimension.
In this respect, Gustav Radbruchs dichotomy between right as a
realm of justice in compensation (commutative justice) and an extralegal realm of distributive justice and social utility is to be resolutely
rejected.9 Penal law, which incontrovertibly constitutes an essential
component of right, pertains neither to commutative justice nor to
corrective justice. The former presupposes the agreement of all
parties over the conditions and the items of the trade. Be it only for
this reason, punishment cannot rest on commutative justice. Not even
retributivism can be described as a trade. Eye for eye, tooth for
tooth10 or the retribution of what one merits does not mean
trading eyes or teeth, or trading the deserts of a human being for
the infliction of evils by another human being. Now, the return to the
status quo ante pertains to corrective justice. However, how can the status
quo ante of a murder victim be reestablished even to a small extent?
In this respect, Cesare Beccarias rhetorical question is irrefutable:
Can the wailings of a wretch, perhaps, undo what has been done
and turn back the clock?11 The punishment, rather, has to do with
distributive justice, because it assures every human being of integration
or of reintegration into the commonwealth regardless of his or her
achievements or actions up to then for the simple reason that every
human being as a being capable of reason has an innate, unforfeitable
right 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. Richard
Bellamy, trans. Richard Davies (Cambridge: Cambridge University Press, 1995), p. 31.
conclusion
195
Fourthly, the categorical legal imperative of integration or reintegration into the commonwealth likewise derives a deeper grounding
in the Kantian foundation of morality, namely in the empirical circumstances of the exercise of autonomy in a limited being capable of
reason who can also act against his or her duty, irrationally. On the
one hand, a human being should bear the responsibility for his or her
actions because he or she has freedom of action. On the other hand,
the consequences of actions carried out in the past should not lead to
the consequence that a person can no longer act freely in the future.
These two dimensions do not limit one another. The consequences of
the actions hitherto done are limited by the commandment that
requires treating every human being as a being capable of reason with
an inalienable freedom of action. Now, this inalienable freedom of
action is not limited by the consequences of past actions. Rather, it is
limited by the same inalienable freedom of action of the other human
beings. It is only for this reason that the Kantian concept of right is an
object of the categorical imperative, which we may also term the
categorical imperative of right, to borrow Otfried Hoffes
expression.12
We therefore find a twofold hierarchization of moral goods that
can collide with one another. First, the commonwealth would receive
precedence over individual freedom in the case of a collision, which
is why it is allowed provisionally either to limit or to suspend the
freedom of the criminal. Secondly, in the case of a collision, individual
freedom receives precedence over the consequences, which the
actions of individuals should otherwise imply; this is why crime should
be forgiven by society with consideration for the first priority, that is,
with consideration for the security of the commonwealth in the most
direct manner possible.13
At the center of the Kantian idea of autonomy, under the guidance
of which we always ought to act and to which, therefore, the legal
system 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 the
Kantian 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. 15972; and Section 2.3 in Jean-Christophe Merle,
Justice et progre`s: contribution a` une doctrine du droit economique et social (Paris: Presses
Universitaires de France, 1997).
196
conclusion
development of the human being as a being capable of reason, and

not the painfully exact accounting of the misdeeds of humanity in the
epoch of the as Kant terms it baby walker.14 As Albert Camus
observed in the debate over the death penalty, one could equally
say of retributivists that: we all know todays executioners are humanists.15 It is just that we do not hold the same concept of humanity
as they do.
14. The epoch of the Gangelwagen, a device used by children learning to walk. Was
ist 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 1957
Nobel Prize winner, trans. Richard Howard (Michigan City, Ind.: Fridtjof-Karla
Publications, 1959), p. 50.
BIBLIOGRAPHY
Primary texts
Beccaria, Cesare. On crimes and punishments, in Beccaria On crimes and punishments
and other writings, ed. Richard Bellamy, trans. Richard Davies (Cambridge:
Cambridge University Press, 1995), pp. 1113.
Bentham, Jeremy. An introduction to the principles of morals and legislation, reprint
of 1823 edition (Oxford: Oxford University Press, 1907).
Camus, Albert. Reflections on the guillotine: an essay on capital punishment by the
1957 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 Weltanschauung
und Lebensgestaltung (Leipzig: O. R. Reisland, 18945).
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:
Naturrecht und Staatswissenschaft (1818/19), in Hegel, and Vorlesungen uber
Rechtsphilosophie, ed. Reinhard Lauth (4 vols., Stuttgart and Bad-Cannstatt:
Frommann-Holzboog, 19734), vol. 1, pp. 217351.
Philosophie des Rechts (1824/25), in Hegel, Vorlesungen uber Rechtsphilosophie,
ed. Reinhard Lauth (4 vols., Stuttgart and Bad-Cannstatt: FrommannHolzboog, 19734), vol. 4, pp. 67752.
Principes de la philosophie du droit, trans. Jean-Francois Kervegan (Paris: Presses
Universitaires de France, 1998).
Hobbes, Thomas. Leviathan, ed. A. R. Walter (Cambridge: Cambridge University
Press, 1904).
Jhering, Rudolf von. Law as a means to an end, ed. Joseph H. Drake, trans. Isaac
Husik (Union, NJ: Lawbrook Exchange, 1999).
197
198
bibliography
Kant, Immanuel. Kants gesammelte Schriften (Berlin: Koniglich Preuische

Akademie 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 preuischen
peinlichen Rechts (Halle: Hemmerde und Schwetschke, 1796).
Krause, Karl Christian Friedrich. Abhandlung uber die Idee und die Eintheilung der
Philosophie und der Mathematik und den innigen Zusammenhang beider, in Paul
Hohlfield and August Wunsche (eds.), Philosophische Abhandlungen: aus dem
handschriftlichen Nachlasse (Leipzig: Schulze, 1889), pp. 540.
Grundlage des Naturrechts oder philosophischer Grundri des Ideals des Rechts,
(Jena: Gabler, 1803).
Mably, Gabriel de. De la legislation, in uvres comple`tes (12 vols., London, 1789).
Mill, John Stuart. April 1868 speech on capital punishment, in Mill,
Utilitarianism, ed. George Sher (Indianapolis: Hackett, 2001), pp. 6570.
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. 13181616.
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 and
J. Kemp, seventh edition (4 vols., London: Kegan Paul, Trench, Trubner
and Co., 1909?).
Schroeder, Friedrich-Christian (ed.). Die Carolina: die Peinliche Gerichtsordnung
Kaiser Karls V. von 1532 (Darmstadt: Wissenschaftliche Buchgesellschaft,
1986).
Seneca. De ira, in Moral essays, trans. John W. Basore (3 vols., London: Heinemann/
Cambridge, Mass.: Harvard University Press, 1928), vol. 1, pp. 106355.
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 provincial
lawyer, in Voltaire, Political writings, ed. and trans. David Williams
(Cambridge: Cambridge University Press, 1994), pp. 24479.
Secondary texts
Altman, Andrew and Wellman, Christopher Heath. A defense of international
criminal law, Ethics, 115, no. 1 (2004), 3567.
Ataner, Attila. Kant on capital punishment and suicide, Kant-Studien, 97, no. 4
(2006), 45282.
Badinter, Robert. LAbolition (Paris: Fayard, 2000).
Bedau, Hugo. Retribution and the theory of punishment, Journal of Philosophy,
75, no. 11 (1978), 60120.
Benn, Stanley. An approach to the problems of punishment, Philosophy, 33
(1958), 32141.
bibliography
199
Bianchi, Hermann. Abolition: assensus and sanctuary, in Alexander R. Duff

and David Garland (eds.), A reader on punishment (Oxford: Oxford
University Press, 1994), pp. 33651.
Bosanquet, Bernard. The philosophical theory of the state (London and New York:
Macmillan, 1899).
Braithwaite, John. Crime, shame and reintegration (Cambridge: Cambridge
University Press, 1989).
Braithwaite, John and Pettit, Philip. Not just deserts: a republican theory of criminal
justice (Oxford: Clarendon Press, 1990).
Brandt, Reinhard. Das Erlaubnisgesetz, oder: Vernunft und Geschichte in
Kants Rechtslehre, in R. Brandt (ed.), Rechtsphilosophie der Aufklarung:
Symposium Wolfenbuttel 1981 (Berlin: De Gruyter, 1982), pp. 23385.
Breazeale, Daniel. Philosophy and truth: selections from Nietzsches Notebooks of the
early 1870s (Atlantic Highlands, NJ: Humanities Press, 1979).
Brugger, Winfried. Darf der Staat ausnahmsweise foltern? Der Staat, 35 (1996),
6797.
Vom unbedingten Verbot der Folter zum bedingten Recht auf Folter?
Juristenzeitung, 35, no. 4 (18 February 2000), 16573.
Byrd, Sharon. Kants theory of punishment: deterrence in its threat, retribution
in its execution, Law and Philosophy, 8, no. 2 (1989), 151200.
Clark, Michael. A non-retributive approach to punishment, Ratio, 17, no. 1
(2004), 1227.
Cohen, Hermann. Schriften zur Philosophie und Zeitgeschichte, ed. Albert Gorland
and Ernst Cassirer (Berlin: Akademie Verlag, 1928).
Deith, John. On the right to be punished: some doubts, Ethics, 94 (1984),
191211.
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), 3146.
Flechtheim, Ossip Kurt. Die Funktion der Strafe in der Rechtstheorie Hegels,
in Flechtheim, Von Hegel zu Kelsen: rechtstheoretische Aufsatze (Berlin: Duncker
und Humblot, 1963), pp. 920.
Zur Kritik der Hegelschen Straftrechtstheorie, Archiv fur Rechts- und
Sozialphilosophie, 54 (1968), 53948.
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 quon 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).
200
bibliography
Hampton, Jean and Murphy, Jeffrie. Forgiveness and mercy (Cambridge:

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), 291314.
Kant on wrongdoing, desert, and punishment, Law and Philosophy, 18 (1999),
40741.
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. 21333.
Konigliche Volker: zu Kants kosmopolitischer Rechts- und Friedenstheorie (Frankfurt
a.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: Kants penal theory reinterpreted,
Utilitas, 9 (1997), 321.
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 (}} 1316), in Jean-Christophe Merle (ed.), Johann Gottlieb Fichte:
Grundlage des Naturrechts (Berlin: Akademie Verlag, 2001), pp. 12537.
Kersting, Wolfgang. Sittengesetz und Rechtsgesetz: die Begrundung des Rechts
bei Kant und den fruhen Kantianern, in Reinhard Brandt (ed.),
Rechtsphilosophie der Aufklarung: Symposium Wolfenbuttel 1981 (Berlin: De
Gruyter, 1982), pp. 14777.
Wohlgeordnete Freiheit: Immanuel Kants Rechts- und Staatsphilosophie, second
edition (Berlin: De Gruyter, 1993).
Kleinig, John. Punishment and moral seriousness, Israel Law Review, 25, no. 3
(1991), 40121.
Klug, Ulrich. Abschied von Kant und Hegel, in Jurgen Baumann (ed.), Programm
fur ein neues Strafgesetzbuch: der Alternativ-Entwurf der Strafrechtslehrer (Frankfurt
a.M.: Fischer, 1968), pp. 3641.
Koller, Peter. Probleme der utilitaristischen Strafrechtfertigung, Zeitschrift fur
die Gesamte Strafrechtswissenschaft, 91 (1979), 4595.
bibliography
201
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 KlausMichael Kodalle (ed.), Karl Christian Friedrich Krause (17811832):
Studien zu seiner Philosophie und zum Krausismo (Hamburg: Meiner, 1985),
pp. 8092.
Lazzari, Alessandro. Eine Fessel, die nicht schmerzt und nicht sehr hindert:
Strafrecht, in Jean-Christophe Merle (ed.), Johann Gottlieb Fichte: Grundlage
des Naturrechts (Berlin: Akademie Verlag, 2001), pp. 17386.
Lesch, Heiko Hartmut. Der Verbrechensbegriff: Grundlinien einer funktionalen
Revision (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
McTaggart, John E. Studies in Hegelian cosmology (Cambridge: Cambridge
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 Aldo
Masullo and Marco Ivaldo (eds.), Filosofia trascendentale e destinazione etica
(Milan: Guerini, 1995), pp. 30325.
Justice et progre`s: contribution a` une doctrine du droit economique et social (Paris:
Presses Universitaires de France, 1997).
(ed.). Eigentumsrecht, in Merle (ed.), Johann Gottlieb Fichte: Grundlage des
Naturrechts (Berlin: Akademie Verlag, 2001), pp. 15972.
Mohr, Georg. Unrecht und Strafe, in Ludwig Siep (ed.), G. W. F. Hegel:
Grundlinien der Philosophie des Rechts (Berlin: De Gruyter, 1997), pp. 95124.
Murphy, Jeffrie G. Does Kant have a theory of punishment? Columbia Law
Review, 87 (1987), 50932.
Neumann, Ulfried and Schroth, Ulrich. Neuere Theorien von Kriminalitat und
Strafe (Darmstadt: Wissenschaftliche Buchgesellschaft, 1980).
Ottmann, Henning. Philosophie und Politik bei Nietzsche (Berlin: De Gruyter,
1987).
Pelczynski, Zbigniew A. (ed.). Hegels political philosophy (Cambridge: Cambridge
Piontkowski, Andrei A. Hegels Lehre uber Staat und Recht und seine Strafrechtstheorie,
trans. Anna Neuland (Berlin: De Gruyter, 1960).
Pogge, Thomas W. Is Kants Rechtslehre comprehensive? Southern Journal of
Philosophy, 36, supplement (1997), 16187.
Is Kants Rechtslehre a comprehensive liberalism? in Mark Timmons (ed.),
Kants Metaphysics of morals (Oxford: Oxford University Press, 2002),
pp. 13358.
Primorac, Igor. Punishment as the criminal right, Hegel-Studien, 15 (1980),
18798.
202
bibliography
Primoratz, Igor. Banquos Geist: Hegels Theorie der Strafe, Hegel-Studien, 29,
supplement (1986).
Pugsley, Robert A. A retributivist argument against capital punishment, Hofstra
Law Review, 9 (1981), 150123.
Quinton, Anthony M. On punishment, in Harry Burrows Acton (ed.),
The philosophy of punishment (London: Macmillan, 1969), pp. 5564.
Radbruch, Gustav. Rechtsphilosophie, ed. Erik Wolf, seventh edition (Stuttgart:
K. F. Koehler, 1970).
Ricur, Paul. Le Juste (Paris: Editions Esprit, 1995).
Ripstein, Arthur. Equality, responsibility and the law (Cambridge: Cambridge
Rosen, Fred. Utilitarianism and the punishment of the innocent: the origins
of a false doctrine, Utilitas, 9, no. 1 (March 1997), 2337.
Roxin, Claus. Strafrechtliche Grundlagenprobleme (Berlin: De Gruyter, 1973).
Schuld und Verantwortlichkeit als strafrechtliche Systemkategorien, in
Roxin(ed.), Grundfragen der gesamten Strafrechtswissenschaft (Berlin: De
Gruyter, 1974), pp. 17197.
Scala, Klaus. Hegels Begriff der Strafe und die moderne Strafvollzugsproblematik,
Hegel-Jahrbuch (1987), 16470.
Scharf, Michael P. Justice vs. peace, in Sarah B. Sewall and Carl Kaysen (eds.),
The United States and the International Criminal Court: national security and
international law (Lanham, Md.: Rowman and Littlefield, 2000),
pp. 21336.
Scheid, Don E. Kants retributivism, Ethics, 93 (1983), 26282.
Schild, Wolfgang. Ende und Zukunft des Staatsrechts, Archiv fur Rechts- und
Sozialphilosophie, 70 (1984), 71112.
Schwarzschild, Steven S. Kantianism and the death penalty, Archiv fur
Rechts- und Sozialphilosophie, 70 (1985), 34377.
Seelmann, Kurt. Versuch einer Legitimation von Strafe durch das Argument
selbstwiderspruchlichen Verhaltens des Straftaters, Jahrbuch fur Recht und
Ethik, 1 (1993), 31526.
Wechselseitige Anerkennung und Unrecht: Strafe als Postulat der
Gerechtigkeit? Archiv fur Rechts- und Sozialphilosophie, 79, no. 2 (1993),
22836.
Shue, Henry. Torture, Philosophy and Public Affairs, 7, no. 2 (197788),
12443.
Sorell, Tom. Moral theory and capital punishment (Oxford: Blackwell, 1987).
Stillman, Peter G. Hegels idea of punishment, Journal of the History of
Philosophy, 14 (1976), 16982.
Tunick, Mark. Hegels 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.
34 (1990), 50114.
Willascheck, Marcus. Why the Doctrine of Right does not belong in the
Metaphysics of Morals, Annual Review of Law and Ethics, 5 (1997), 20527.
Wolf, Jean-Claude. Verhutung oder Vergeltung? Einfuhrung in ethische Straftheorien
(Freiburg i.Br.: Alber, 1992).
bibliography
203
Strafe als Widerherstellung eines Gleichgewichts, Jahrbuch fur Recht und

Ethik, 11 (2003), 199216.
Wood, Allen W. Hegels ethical thought (Cambridge: Cambridge University Press,
1990).
Kants Doctrine of right, Introduction to Otfried Hoffe (ed.), Immanuel
Kant: Metaphysische Anfangsgrunde der Rechtslehre (Berlin: Akademie Verlag,
1999), pp. 1939.
Zaczyk, Rainer. Das Strafrecht in der Rechtslehre J. G. Fichtes (Berlin: Duncker und
Humblot, 1981).
INDEX
abolitionism, 20; see also impunity

acquittal of a debt, see debt
adiaphoron, moral, 27
agreement, 21
Altman, Andrew, 186
Arendt, Hannah, 183
Aristotle, 21, 100
Ataner, Attila, 59
attempt, 92, 190
autonomy of the will, see freedom
of action
Badinter, Robert, 182
banishment, 178, 179; see also exclusion
from society
Beccaria, Cesare, 2, 18, 20, 70, 901,
164, 194
Benn, Stanley, 108
Bentham, Jeremy, 1819, 20, 164
bestiality, 42
Bianchi, Herman, 1, 158
Bosanquet, Bernard, 107
Braithwaite, John, 5, 7, 13
Brandt, Reinhard, 59, 103, 193
Brugger, Winfried, 2
burden of proof, 69
Byrd, Sharon B., 44, 459
Camus, Albert, 196
cancellation (Aufhebung), 115, 124, 126,
129, 1345, 1368, 160
categorical imperative, 2543, 589, 72,
90, 119, 130, 194, 195
child, 645, 66, 67
Christianity, 7
Cicero, Marcus Tullius, 100
Clark, Michael, 102
coercion, 301, 33, 38, 3941, 42, 46,
67, 74, 914, 125, 12845
Cohen, Hermann, 5960
Commission for Truth and

Reconciliation, 183
common sense, 113, 1224, 143, 149,
1845
compensation, 21, 545, 73, 92, 127,
1367, 151, 1523, 155, 1568,
161, 192
conscience, bad, see remorse
consciousness, common, see common
sense
consent, 113, 114
consequences, internal to the act, 1389,
1434
contract, 2931, 54, 152
contribution, 57
conversion, 7980
Cooper, David, 129
corporal punishment, see torture
crime against humanity, 149
cruelty, 1534, 160, 167, 168
death penalty, 1, 5, 20, 31, 478, 4950,
53, 545, 5960, 601, 68, 70, 94, 95,
106, 117, 1334, 157, 169, 1789,
182, 184, 185, 193, 196
debt, 21, 54, 74, 150, 1523, 156, 157,
15961, 167, 168
Deith, John, 80
desert, see merit
deterrence: general, 23, 8, 18, 19,
5771, 94, 978, 1012, 104, 107, 110,
119, 135, 138, 1401, 1778, 1801,
1812, 191; specific, 4
Duhring, Karl Eugen, 150, 1614
duty: of right, 323; of virtue,
Ebbinghaus, Julius, 26
enslavement, see slavery
Erdemovic, Drazen, 180
evil, see wickedness
204
index
example, deterrence through, 93, 95,
104, 138
exclusion from society, 94, 96, 99, 100,
1023, 105, 1534, 167;
see also banishment
expiation contract, 94, 99, 102
expressivism, 72, 175
Feinberg, Joel, 191
Feuerbach, Anselm, 2, 142
Fichte, Johann Gottlieb, 14, 42, 83,
85106, 11618, 134, 144, 153,
178, 188
Fichtelberg, Aaron, 178
Flechtheim, Ossip K., 108, 11113
Fletcher, George P., 167
force, 1324
forgiveness, 78, 81
Forster, Wolfgang, 105, 106
Foucault, Michel, 1, 156
freedom of action, 195
freedom of the will, 35, 624, 65
Garapon, Antoine, 173
goal of punishment, 723, 162, 164,
165, 189, 191
Grotius, Hugo, 19, 21, 191
Grundgesetz (German Basic Law), 176
Guillarme, Bertrand, 111
guilt, 150, 175, 183, 191
Gutmann, Amy, 185
Hampton, Jean, 72, 127
happiness, 55, 756, 778, 80
harm principle, 37
Hart, H. L. A., 70, 73, 82
Hassemer, Winfried, 4, 5, 10, 107
Hegel, Georg Wilhelm Friedrich, 10,
13, 14, 42, 83, 85, 105, 10745,
149, 155, 164, 188
highest good, 55, 756, 90
Hill, Thomas E., 44, 55, 72, 80, 149
Hitler, Adolf, 182, 184
Hobbes, Thomas, 2, 18, 19, 1512
Hoffe, Otfried, 4, 6, 25, 34, 44, 51, 56,
107, 108, 195
Hohfeld, Wesley Newcomb, 92
Holtman, Sarah, 44
Honderich, Ted, 108
Honneth, Axel, 126
Hosle, Vittorio, 42
Hufeland, Gottlieb, 26
humanity, crime against, see crime
against humanity
humanity as an end in itself, 323, 40,
46, 61, 64, 93, 119
humiliation, 157
205
impunity, 15860; see also abolitionism

imputability, 624
incapacitation, 3, 19, 71, 801, 94, 104, 176
incorrigible prisoners, 945
innocents, punishment of the, 19, 22, 48, 74
insight of the criminal, 114, 145
internalization, 147, 151, 168
International Criminal Court (ICC),
1715, 179
International Criminal Tribunal for
Former Yugoslavia (ICTFY), 180
intervention, military humanitarian, 177
ius strictum, 137
Jhering, Rudolf von, 164
Joseph II, Holy Roman Emperor, 133
justice: commutative, 21, 151, 15962,
1645, 194; corrective, 135;
distributive, 21, 174; Hobbesian, 152;
in general, 169; natural law, 169
Kant, Immanuel, 10, 1314, 213, 2443,
4471, 7283, 90, 92, 101, 11820,
136, 149, 181, 1878, 191, 1926
Kaufmann, Matthias, 88, 108
Kersting, Wolfgang, 25, 26, 28, 3941
Kervegan, Jean-Francois, 108
Klein, Ernst Ferdinand, 132, 134, 135, 142
Kleinig, John, 127
Klug, Ulrich, 107
Koller, Peter, 3
Kostlin, Christian Reinhard, 107
Krause, Karl Christian Friedrich, 1056
Kuhl, Kristian, 3, 10, 30
Landau, Peter, 3, 106
Lazzari, Alessandro, 88
Lesch, Heiko H., 110
liberalism, 25, 3443, 106
Liszt, Franz von, 164
Locke, John, 34
Ludwig, Bernd, 28, 29
lynching, 168
Mably, Gabriel Bonnot de, 156
McTaggart, John Ellis, 107
marriage, 42
maxim, 28, 33, 56, 75
May, Larry, 172, 175, 183, 184
mercy, 1601, 167
merit, 74, 91
Merle, Jean-Christophe, 43, 87, 90, 195
Mill, John Stuart, 20, 34, 37
Milosevic, Slobodan, 186
Mitnick, Kevin, 69
mixed theories, 5, 12, 4453, 889, 104,
10720, 169
206
index
mnemotechnic, 153, 169

Mohr, Georg, 1089, 11819
monstrous crimes, 175, 183
morality, 234, 2534, 38, 42, 44, 45, 1423
Moses, 19
Murphy, Jeffrie G., 72
natural law, 18, 169, 176, 191
necessity, see right of necessity
negation: of the crime, 10745; of the
negation, 112, 122
Neumann, Ulfried, 6, 9
Nietzsche, Friedrich, 14, 14770, 188
nulla poena sine lege, 171
nullity of crime, 115, 116, 11718,
129, 137
Nuremberg, tribunal of, 171, 179, 180
Ottmann, Henning, 151
pardon, 31, 68, 159, 167, 179, 181
parole, release on, 69
Paul, Saint, 80
peace, civil, 163, 193; see also war, civil
permissive law, 103, 193
personality, 614
Pettit, Philip, 13
Pfaelzer, Marianna, 69
Piontkowski, Andrei A., 107
Plato, 17, 38, 73, 116, 164
Pogge, Thomas W., 34, 62, 65, 66
positivism, legal, 38
Primorac (Primoratz), Igor, 108, 111,
115, 122
promise, 2931, 152, 161
property right, 41, 60, 65
proportionality principle, 5, 1920, 21, 43,
45, 49, 70, 74, 82, 113, 1202, 124,
127, 159, 168, 174, 183, 184, 189
Pufendorf, Samuel von, 2, 19, 21, 191
Pugsley, Robert A., 59
Quinton, Anthony M., 108
Radbruch, Gustav, 194
recidivism, 9, 10, 13, 57, 68, 69, 95, 147,
165, 181
recognition, 99100, 11618, 131
reconciliation, 113
reform, see rehabilitation
rehabilitation, 3, 5, 7, 9, 13, 578, 6171,
81, 85106, 108, 117, 1356, 145,
147, 167, 189, 194
remorse, 14, 147, 149, 162, 164, 166
republic, 25
resentment, see ressentiment
resistance, 176
responsibility, 67, 13
ressentiment, 154, 161
restoration of right, 112, 113, 1278, 194
retaliation, 6, 72, 90, 112, 115, 117, 1201,
1257, 141, 1434
retribution, 6, 8, 49, 507, 74, 92, 127,
141, 161
retributivism, 2, 34, 6, 7, 214, 43, 517,
72, 812, 110, 119, 141, 143, 147,
181, 183, 192
revenge, 56, 78, 812, 114, 1256, 135,
137, 1612, 163; see also vengeance
Ricur, Paul, 5
right, concept of, 857, 89, 90, 97, 99,
101, 116, 1445
right of necessity, 46, 52, 94, 193
Ripstein, Arthur, 80, 108, 139
Rosen, Fred, 3, 74
Roxin, Claus, 7, 9, 11, 12, 13, 107, 1478
savage, 67
Scala, Klaus, 108
Schadenfreude, 149
Scharf, Michael P., 179
Scheid, Don E., 44, 51
Schild, Wolfgang, 108
Schmalz, Theodor, 26
Schopenhauer, Arthur, 2, 107
Schroth, Ulrich, 6, 9
Schwarzschild, Steven S., 59
security, public, 95, 140, 1579, 190,
191, 195
Seelmann, Kurt, 108, 111, 11617, 141
self-defense, right of, 91, 137, 191
Seneca, 17, 73
Shue, Henry, 2
Sidgwick, Henry, 17, 20, 23
slavery, 56, 601, 66, 95
social contract, 85, 94, 102
Sorell, Tom, 60
sovereign, 91
state of nature, 52, 53, 57, 589, 101,
154, 181, 193
statute of limitation, 1745, 1846, 189
Stillman, Peter G., 108
suffering, see torture
suicide, 37, 59, 61
talion law, 17, 1920, 213, 512,
567, 87, 956, 101, 113, 1212,
127, 142
teleology, 3941, 65
Thompson, Dennis, 185
threat of punishment, deterrence
through, 2, 468, 70, 934, 98,
1045, 119
Tokyo, tribunal of, 171, 179
index
torture, 1, 5, 20, 48, 96, 106, 128,
1523, 154, 157, 163, 166, 1689,
184, 185, 186
Tunick, Mark, 108
tyrannicide, 176
universalization, 56
usefulness, 99, 1001
utilitarianism, 18, 44, 51, 68, 100
van den Haag, Ernest, 2
Vattel, Emer de, 191
vengeance, 767
victim, 9, 81, 92, 127, 130, 137, 155,
157, 15960, 162, 163, 171, 173,
186, 190, 194
207
virtue, 746, 77
Voltaire, 170
war: civil, 156, 159, 193; international,
1912; see also peace, civil
Wellman, Christopher H., 186
wickedness, inner, 43, 55, 734, 75,
76, 778, 79, 80, 101, 102, 122,
176, 184, 1901
Willascheck, Marcus, 34
Wolf, Jean-Claude, 55, 107
Wolff, Christian, 2, 39
Wood, Allen W., 36, 1089, 114
worthiness of punishment, 73, 812
Zaczyk, Rainer, 93

German Idealism and The Concept of Punishment Modern European Philosophy

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

German Idealism and The Concept of Punishment Modern European Philosophy

Загружено:

Авторское право:

Доступные форматы

This page intentionally left blank

GERMAN IDEALISM AND THE

MODERN EUROPEAN PHILOSOPHY

Some recent titles

Robert M. Wallace: Hegels Philosophy of Reality, Freedom, and God

GERMAN IDEALISM AND

CAMBRIDGE UNIVERSITY PRESS

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

Cambridge University Press has no responsibility for the persistence or accuracy

. . . we all know todays executioners are humanists.

part i desert as the sole justification

2 Kants legal justification of punishment

3 Kants moral justification of punishment

part ii punishment as a means of

5 Hegels negation of crime

part iii retributivist inhumanity

7 What is the purpose of punishing crimes

Oliveira, Herve Pourtois, Alexandre Travessoni Gomes, Luiz Moreira

Groundwork of the metaphysics of morals

(Religion innerhalb der Grenzen der bloben Vernunft)

Lecture on The philosophy of right

Note on translations of primary and secondary literature

A nearly total unanimity prevails with regard to the fundamental

german idealism and the concept of punishment

which was already systematically employed as an interrogation method,

By serving a prison term, a prisoner should eventually become able to

On the other hand, theories of general deterrence often draw the

german idealism and the concept of punishment

Retributivism criticizes theories of general deterrence as well as

observes: The goal of rehabilitation is necessarily contained in opting

In view of this antinomy of the punitive goals, a priority rule must be

german idealism and the concept of punishment

The form of retributivism that is given priority regards its moral

differentiation must be taken into account. There are always two

Thirdly, retributivism relies on the assumption that the consequences

german idealism and the concept of punishment

A short thought experiment may illuminate this point. Let us

german idealism and the concept of punishment

reason, the criticism of retributivism that will be made in this book

Even if an especially high recidivism rate were to be observed in regard

german idealism and the concept of punishment

The restriction, which Roxin admits partly contradicts retributivism,

The justification for this mixed theory is found by Roxin, as by so

According to the division of labor in the mixed theories, every theory

out of hand the criticism against rehabilitation theory that it allows

german idealism and the concept of punishment

complete categorical imperative to such an extent in the legal system

1.1. The retributivist turn

desert as the sole justification for punishment

pursuit of happiness (Hobbes), or happiness through the least possible

the two kantian concepts of right

mentions making an example of a criminal; as a means to specific

desert as the sole justification for punishment

more severe crimes rather than lesser crimes, as well as to deter

the two kantian concepts of right

desert as the sole justification for punishment

means (rigorosa) for preventing the recurrence of wrongs by others;

the two kantian concepts of right

desert as the sole justification for punishment

1.2. Two Kantian concepts of right

the two kantian concepts of right

coexistence of empirical freedoms according to the universal law of

desert as the sole justification for punishment