1 Introduction

The ultimate aim of criminal law is freedom. This is the defining idea of what we may call freedom theories of criminal law. They identify infringement of the freedom of others as the essential feature of crime (as opposed to harm, for instance). And they take prevention and remedying of such infringements as the purpose of the criminal justice system.Footnote 1

This way of framing crime and punishment in terms of freedom can easily be applied to typical cases. Imagine, for instance, that an assault causes the victim to be paralyzed from her neck down. No doubt her freedom will have been drastically reduced. She can no longer feed herself or move around as she pleases. Similarly, a person who has been robbed will experience a decrease in her freedom, at the very least because she will have less money to spend as she wishes, and likely also because fear and anxiety will restrict her in her daily life.

These are examples of how crimes cause a reduction in the freedom of others. Some proponents of freedom theories of criminal law make a stronger claim, however. A crime does not merely cause a person to have less freedom; it constitutes, in and of itself, a breach of the freedom of others.

If this latter claim is true, a crime must have a certain meaning that makes it an infringement of freedom irrespective of its consequences. Irrespective of whether the victim or others actually experience decreased freedom as a result of the crime, the crime means that the victim’s freedom has been breached. Say, for instance, that a person has been defrauded without noticing it, or photographed nude without their knowledge. These victims do not actually experience a loss of freedom, as stipulated. Yet, the theory claims, their freedom has been infringed upon.Footnote 2

The same distinction can be made with regard to punishment: It may cause an increase in freedom, for instance by preventing crimes, which threaten our freedom. And according to some freedom theories, punishment may also, in and of itself, remedy the infringement of freedom entailed by the crime. If this latter claim is true, punishment must have a meaning that makes it an expression of respect for freedom, whether or not the amount of freedom in society increases due to crime prevention etc.

For both crime and punishment, then, we can distinguish between an instrumental function and an immanent function. Upon the first, the two are means of (respectively) decreasing and increasing freedom. Upon the second, they instantiate (respectively) the denial and affirmation of freedom.

While the instrumental function of crime and punishment is easily grasped, as we saw in the examples above, the immanent function is far from intuitive and requires explanation. It is this immanent connection between crime/punishment and freedom that is the topic of this paper. My aim is to show that such a connection exists. I will do so by explicating the “natural meaning” of crime and punishment. This way of addressing the topic is inspired by Jean Hampton’s use of H. P. Grice’s concept of natural meaning. Hampton uses the concept to describe how crimes necessarily express a diminishment of the victim’s moral worth. Without contradicting her view, I will argue that there is a further natural meaning of crime as denial of mutual freedom. I do so by applying H. L. A. Hart’s notion of a natural right to freedom. This notion of natural right is not natural in the sense of being “eternal” or “intrinsic to human nature” or the like. It is natural in the conditional sense that if one accepts the notion of a right, then the notion of freedom follows. The concept of freedom is entailed by the “nature” of a right, if you will. Likewise, Grice’s concept of natural meaning does not imply that meaning arises out of nature or is limited to natural facts or the like. Rather, the concept is used to capture a logical relationship that does not arise from meaning that is due to convention, or what Grice calls “non-natural meaning”. It is this logical relationship, I will argue, that exists between crime and denial of freedom, and between punishment and affirmation of freedom.Footnote 3

Identifying an instrumental and an immanent connection between crime/punishment and freedom allows us to distinguish between two main types of freedom theory of criminal law: a consequentialist and a deontological type. The theory of John Braithwaite and Phillip Pettit can stand as an example of the first type. “The target of the criminal justice system should be to maximize dominion”, they write.Footnote 4 Punishment and other practices of the criminal justice system serve instrumental functions according to their theory. Punishment promotes the goal of maximizing freedom.

The theories of Kant and Hegel are primary examples of deontological freedom theories. They view punishment, and criminal law in general, as instantiating the value of freedom. They do not deny the instrumental function, of course. But they require of a just criminal justice system that the value of freedom be realized in and through punishment and the other practices of the system.

The existence of an immanent link between crime/punishment and freedom is therefore a prerequisite for the deontological type of freedom theory. If we can establish the existence of this link, the framing of crime and punishment in terms of freedom will become all the more plausible and appealing. That said, much more would have to be explored in order to defend the freedom theory as a comprehensive theory of criminal law, for instance regarding the theory’s ability to account for well-established legal principles. I do not here have space to conduct such a discussion. My aim is only to argue for one necessary condition of the deontological freedom theory, the inherent connection between freedom and crime/punishment.Footnote 5

However, there is a further reason why the argument of this paper might serve to strengthen the position of freedom theories vis-à-vis other theories of criminal law. The reason is that the freedom perspective is inextricably tied to the concept of rights. There is therefore a sense in which the framing of criminal law in terms of freedom is ineluctable. We cannot, if we are to accept the notion of a right, avoid accepting that a breach of rights through crime means denying the rightsholder’s freedom. This does not imply that we have no choice but to adopt the freedom theory of criminal law. But it does imply that the freedom theory is especially appropriate for criminal law in a democratic state under the rule of law. The freedom theory has a connection to the concept of the rule of law that competing theories, such as fair play theories, utilitarian theories or moral education theories, do not. Put differently: while “denial of freedom” is a natural meaning of crime within a system of rights, “freeloading”, “negative social utility” and “sign of moral ignorance” are conventional meanings of crime. As such, these ways of understanding crime may be more or less common, depending on the conventional meanings of the community, and may seem more or less appropriate to different types of crimes under different circumstances.Footnote 6 But they are in no way necessitated by the concept of a system of rights, i.e. the rule of law, like the freedom framing is.

Before I start to explain how I see the immanent connection between crime, punishment and the value of freedom, I hasten to add that I do not claim that this connection always exists in practice. On the contrary, the social and political realities of most countries are such that it would be false to say that crimes always entail a denial of mutual freedom, and even more so that punishment is always an instantiation of respect for mutual freedom. Excessively harsh sentences and horrendous prison conditions, where offenders are brutalized and denied basic rights, show how the criminal justice systems of most countries fail to affirm the freedom of all, including the offender.

This does not mean that the theory is wrong, however. Rather, we should say that it is the theory that allows us to see what is wrong in the current practice. The theory conveys an ideal to which practice can be compared. Deontological freedom theories, specifically, identify what is wrong not merely in the results of our current practice, but in the very practice itself. Accordingly, our criminal justice systems fail not simply because they do not deliver on the promise to increase our freedom–they fail not simply because recidivism is high and brutalized prisoners tend to make us less safe, and hence, tend to decrease, not to increase, our freedom. Our criminal justice systems fail also because prisons in which prisoners are brutalized undermine the value that the systems ought to instantiate.Footnote 7

2 The Expressive Functions of Crime and Punishment

Deontological freedom theories entail a claim that crime and punishment serve expressive functions. Crime does not merely cause material harm; it also expresses a denial of the victim’s right to freedom. In Kant’s terminology, a crime is not (primarily) a material wrong; it is a formal wrong, meaning that the form of the action is such that it is incompatible with equal rights for all. Likewise, punishment does not (primarily) remedy a material wrong, as compensation does. It remedies a formal wrong, by re-framing the criminal action in such a way that its form becomes compatible with equal rights for all. I will explain this theory below. For now, the main point is that it entails a view of both crime and punishment as expressive actions.

Indeed, it is this expressive function that distinguishes crime from other misfortunes and harms, and punishment from other burdens that may befall someone. Put negatively, if an event does not have meaning–if it does not express anything–it is not a crime, and hence, cannot be symbolically negated. An earthquake, for instance, does not communicate. It is therefore neither right nor wrong. Human action, on the other hand, can be meaningful, and hence potentially right or wrong.

To this someone might object that many cultures have viewed earthquakes and the like as punishments from God. We even find a curious example of the opposite, a human punishing nature for its crimes, in the story of the Persian king Xerxes who whipped the sea for its failure to obey him. These are not counterexamples, however. They merely show that people have attributed meaning to nature and the Cosmos, thereby including them within the realm of justice and injustice.

The first to coin the term “the expressive function of punishment” was Joel Feinberg, in the article by the same name.Footnote 8 He outlines a few ways that punishment communicates to the public at large: by authoritative disavowal,Footnote 9 by absolution of others,Footnote 10 by symbolic non-acquiescence,Footnote 11 and by vindication of law.Footnote 12 After having showed that punishment serves important expressive functions in (at least) these four ways, Feinberg raises the following challenge: Could not the same expressive functions be achieved by other means?For instance, could we instead of punishing a rapist hold a parade in honor of the victim, with banners and speeches denouncing the crime and expressing to the victim her equal worth? Could we do without any ritual at all, beyond a simple statement conveying the wrongness of the crime? Feinberg’s answer: “Perhaps, but when [the state] speaks by punishing, its message is loud and sure of getting across.”Footnote 13 His point is that as things are, punishment is the most effective way of expressing condemnation.

[C]ertain forms of hard treatment have become the conventional symbols of public reprobation. This is neither more nor less paradoxical than to say that certain words have become conventional vehicles in our language for the expression of certain attitudes, or that champagne is the alcoholic beverage traditionally used in celebration of great events, or that black is the color of mourning.Footnote 14

We might compare punishment to the custom of giving gifts: If a person does you a favor, you can express your gratitude with a simple “thank you”. But the message will get across louder and clearer if you also give a bouquet of flowers or a bottle of wine. Such are current conventions for expressing gratitude. The conventions could be different, however. The act of giving a bouquet of flowers could in some distant culture mean a threat. Giving wine could mean the equivalent of throwing down one’s gauntlet, i.e. a challenge to a duel. Hence, there is no inherent connection between the medium and the message.

Feinberg’s answer to his own challenge, then, is that punishment is necessary for the achievement of the mentioned expressive functions, not because punishment is inherently suited for this aim, but as a matter of convention. The goal of punishment is external to the practice of punishment. And because we might achieve the goal by other means, we must justify that we ought to use punishment instead of other media of communication. Feinberg concludes: “The problem of justifying punishment, when it takes this form, may really be that of justifying our particular symbols of infamy”.Footnote 15

How might one go about justifying punishment in this way? Since punishment involves the intentional infliction of pain, we would have to show that punishment is so much more effective in achieving the desired goals, compared to non-punitive sanctions, that the “price” of causing pain is worth it. But not only is it a matter of which sanction is more effective; this type of consequentialist justification presupposes that the value attained is greater than the cost of the sanction. To determine whether that is true, we must first determine which value punishment attains. The expressivist theory does not by itself answer that question. It merely says that punishment is justified because it is the most efficient way of expressing the public’s disapproval or condemnationFootnote 16 (or more specifically on Feinberg’s theory, because it best achieves the four expressive functions). That does not suffice as a justification, for the simple reason that the public’s condemnation may itself be immoral, e.g. in a society where the majority is racist and condemns disobedience of segregation laws.

Condemnation, disapproval or symbolic non-acquiescence etc. can only be justified if they come as responses to the flouting of normatively acceptable values. Further, not all types of value can justify every form of condemnation (e.g. the flouting of esthetic value does not warrant punishment of an architect who has designed an ugly building). The bottom line is that expressivism cannot stand alone as a justification of punishment.Footnote 17 It requires a theory of which value it is that punishment expresses, and why expressing that value justifies punishment. I turn now to Jean Hampton’s version of expressivism in search of such a theory.

3 Hampton’s Theory of the Meaning of Crime and Punishment

Hampton applies a Gricean theory of meaning to show how a criminal act may convey meaning. H. P. Grice distinguished between natural meaning and non-natural meaning. His example of the latter is, “Those three rings on the bell (of the bus) mean that the ‘bus is full’” Footnote 18 The meaning of the three rings of the bell is non-natural in the sense that it is purely a matter of convention. The convention could be different: three rings could mean there is room on the bus; one ring could mean it is full. Similarly, Feinberg’s conventional meaning of punishment, corresponding to Gricean non-natural meaning, allows for the possibility that the convention could be different and punishment would have a different meaning than it does. For example, under one convention, punishment means “absolution of others”. Under another convention, punishment could be taken to confer guilt on the offender’s family or clan, with whom the offender was identified. Likewise, punishment may express the conventional meaning that the offender has unfairly shed the burdens of law-abidance that every citizen must carry, thereby profiting by freeloading off others.Footnote 19 However, punishment may also mean almost the opposite, that the offender has damaged himself rather than gained profits from his immoral crimes, cf. Plato’s claim that it is better to suffer injustice than to commit it.Footnote 20

Hampton acknowledges that crime and punishment may convey meaning in the non-natural, conventional, sense, but asserts in addition that they are meaningful in Grice’s natural sense. Natural meaning can be understood as something akin to “implies”, “shows” or “equates to”. We say, for instance, “Those dark clouds mean it is going to rain”, and, “Today’s drop in share prices means I will lose money”. Or Grice’s example: “Those spots mean measles”. We are clearly not talking about the linguistic meaning of the words uttered, but rather about an actual or logical connection between the two phenomena referred to on each side of the word “mean(s)”.

For natural meaning, as opposed to for non-natural meaning, “x means p” entails p.Footnote 21 Therefore, you cannot say without contradiction, “Those spots meant measles, but he hadn’t got measles”. Or, “Today’s drop in share prices means I will lose money, but I won’t lose money”. You can, on the other hand, say, “Those three rings on the bell (of the bus) mean that the bus is full, but it isn’t, the bus driver is mistaken”.

Human actions can be meaningful in the natural sense. Hampton’s example is of an art expert who vandalizes a painting by slitting its canvas, an act that means that the critic considers the painting worthless.Footnote 22 Why can we infer this meaning? Because, Hampton says, to consider an artistic object valuable means “to preclude many kinds of treatment with respect to this object”. She discusses as an example the Book of Kells. If somebody spray-painted over its pages we would be furious. “[I]ts value generates certain entitlements. For as long as the Book of Kells has that value, it has these entitlements, which include being preserved, treated with care, and so forth.”Footnote 23

In the same way, Hampton claims, a criminal act means that the offender demeans the victim. “By victimizing me, the wrongdoer has declared himself elevated with respect to me, acting as a superior who is permitted to use me for his purposes.”Footnote 24 The aim of punishment is to deny this false claim of degradation conveyed by the crime. Punishment “negates the evidence of superiority”.Footnote 25 “The retributive punisher uses the infliction of suffering to symbolize the subjugation of the subjugator, the domination of the one who dominated the victim.”Footnote 26

However, if the purpose of punishment is to “bring low” the offender in order to re-establish the parties’ equal worth, it seems this function rests on the assumption that the offender actually succeeds in elevating himself and lowering the victim. Interpreting Hampton in this way, Heather Gert, Linda Radzik and Michael Hand conclude that she proposes an “offensive equation of power with value”.Footnote 27 David Dolinko makes a similar point when he asks why somebody should take a crime against her as evidence of her inferiority. “If you find your home burglarized, you may experience anger, or a sense of defilement, or fear that it will happen again, or all of these–but will you feel that the burglar has demonstrated that his moral value is greater than yours? Surely not!”.Footnote 28

Hampton’s answer is that crimes do not actually cause the victim to be inferior to the offender. Crime is merely an appearance of degradation. Therefore, Hampton says, “the retributive motive for inflicting suffering is to annul or counter the appearance of the wrongdoer’s superiority and thus affirm the victim’s real value”.Footnote 29 For Hampton, as for Kant, the victim’s real value is infinite, or as Kant says, “raised above all price”.Footnote 30 A victim cannot lose value, and an offender cannot have more value. Human value is priceless and inalienable.Footnote 31

This answer, however, seems to leave Hampton’s position vulnerable to the kind of critique that Dolinko makes: “But why should we care about nullifying precisely those claims?”Footnote 32 All kinds of false moral claims are made every day. Should it be the state’s responsibility to negate all such claims? Take the example of Randy Newman who had a huge hit singing: “Short people got no reason to live.” Clearly, we cannot justify punishing him in order to show the claim to be merely an appearance of the value of short people. (Perhaps the bridge of the song would have been enough to acquit him: “Short people are just the same as you and I.”).

To sum up, Hampton’s version of expressivism has answered the question of which value it is that punishment expresses (the value of equal moral worth). However, it is not clear why expressing that value would justify punishment.

4 Freedom and Right

The theory I propose takes as its starting point H. L. A. Hart’s account of rights and a similar insight provided by Hampton’s example of the art critic who slits the canvas of a painting.

The art critic’s act means that for her the painting is worthless. The reason the act has this natural meaning is because it is entailed by the meaning of artistic value. An object having artistic value means certain treatment of it is precluded. As Hampton put it, the object has entitlements as long as it is viewed as valuable. Vandalizing the object means disrespect of those entitlements, and hence disrespect of its value.

Hart’s theory of rights, as he lays it out in “Are There Any Natural Rights?”, similarly asserts a necessary connection between valuing the autonomy of an individual and according her moral rights. A person’s freedom is entailed by the concept of having rights, and vice versa. “There is no place for a moral right unless the moral value of individual freedom is recognized.”Footnote 33 Hart does not argue that this natural right to freedom is part of an eternal natural law; he is making the conditional assertion that “if there are any moral rights then there must be this one natural right”.Footnote 34

How, then, is the natural right to freedom entailed by the concept of moral rights? Hart distinguishes between special rights and general rights, which mirror each other. Special rights arise out of transactions or special relationships and accord the right to something that would otherwise be within another person’s sphere of autonomy. “[T]he claimant has some special justification for interference with another’s freedom which other persons do not have (‘I have a right to be paid what you promised for my services’).”Footnote 35 General rights apply to all and accord the right not be interfered with unless somebody has a special right to do so. “[G]eneral rights […] are asserted defensively, when some unjustified interference is anticipated or threatened, in order to point out that the interference is unjustified. ‘I have the right to say what I think.’ ‘I have a right to worship as I please.’”.Footnote 36

Hart’s point is that both special rights and general rights can only be understood against a background where all have a natural right to be free. It does not make sense to claim a special right to interfere unless persons by default have the general right not to be interfered with–that is, unless we are by default free. To claim, for instance, “I have a right to your services according to our contract” only makes sense if but for the contract you were rightfully free not to render the services. Further, the very possibility of according me special rights presupposes your autonomy. “For we are in fact saying in the case of promises and consents or authorizations that this claim to interfere with another’s freedom is justified because he has, in exercise of his equal right to be free, freely chosen to create this claim.”Footnote 37 Both types of right thus presuppose the autonomy of the individual: (1) for transferring rights to another, and (2) as the negative against which a positive right to interfere is conceivable.

Robert Nozick objects to this argument, and attempts to show its flawed logic by turning it on its head:

If there were cogency to Hart’s claim that only against a background of required nonforcing can we understand the point of special rights, then there would seem to be equal cogency to the claim that only against a background of permitted forcing can we understand the point of general rights.Footnote 38

Nozick’s point is the following: If I say that we all have a general right to take a walk or to eat dinner or to speak our minds, it only makes sense, applying Hart’s logic, if but for this general right, others could legitimately force us not to do so. In conclusion: Only if we do not have a natural right to be free can we understand our general rights to do such things as taking a walk or eating dinner.

Nozick’s argument does not work, however, for he has missed the point that general rights are negative. We do not have general rights to do anything in particular; we have general rights not to be interfered with when others do not have special rights to do so. Admittedly, Hart could have been clearer about this. His examples of general rights are “I have a right to say what I think” and “I have a right to worship as I please”. However, he does say that these sentences are asserted defensively, when an unjustified interference is anticipated. The general right is not really “the right to speak your mind”, but “the right not to be hindered in speaking your mind, save for justifiable reasons”. Likewise, the general right is really “not to be hindered in worshiping”.

Turning this on its head is implausible and in conflict with the way we normally conceive of rights. We do not normally think we are prohibited to do everything except that which we have a right to do. That would require an infinite amount of positive general rights, for instance, the general right to climb a tree on a Wednesday, which we would otherwise be prohibited in doing because we are by default unfree. And how could we even conceive of special rights as overriding general rights without confirming Hart’s point? If I have a general right to climb trees on Wednesdays unless others have a special right to force me not to, then I am in fact free. Nothing has then changed from Hart’s theory except that the name of general rights is misleadingly assigned to “the right to climb trees on Wednesdays” instead of “the right not to be hindered in climbing trees on Wednesdays”.

Instead of an infinite amount of positive general rights, we have one negative general right to freedom, and many positive special rights. This is essentially the same claim that Kant makes when he says that we have one innate right to freedom from which all positive rights stem.Footnote 39 According to Kant’s Universal Principle of Right, we have a general right to unrestricted freedom to the extent compatible with everyone else’s equal freedom. Special rights make exceptions to this general right to non-interference, but, importantly, exceptions that confirm the rule, i.e. our innate right to freedom (by presupposing our autonomy in creating special rights). If I let you use my car, your interference in my property is compatible with mutual freedom, because I have granted you a special right. This is the principle of volenti non fit injuria. Without consent, however, your use of my car would be wrong. It would not be a case of a special right overriding the general right to non-interference, because special rights presuppose respect for the autonomy of the person granting the right. In other words, one cannot have a special right without respecting freedom.

The upshot of this is that if you do interfere with someone’s general right without having a special right to override the general right, you are not respecting that person’s freedom. Hence, disrespecting rights means disrespecting freedom.

Not all breaches of right amount to crimes, however. Imagine, for instance, that you had good reason to believe that I had given you permission to use my car, but that I had not in fact done so. In that case, you would have infringed upon my property rights – you have used my car without permission – but you would not (necessarily) have committed a crime. In order to distinguish the meaning of this type of infringement of rights from the meaning of crimes, Hegel’s categories of wrongs may be helpful. Hegel distinguishes between civil wrong and crime in the following way: Civil wrong “negates only the particular will, but pays respect to the general right”.Footnote 40 In crime, “neither right in general nor my personal right is respected”.Footnote 41

Let us apply this to my example: If you use my car while thinking that you had permission, which you in fact did not have, you infringe upon my particular property right to my car, but your action does not amount to a denial of my right to my car as such. This is a civil wrong. Your action is compatible with your recognition that it is indeed my car. In other words, you do not express through your action that property rights do not apply to me. In fact, by acknowledging my right to give you permission, you implicitly respect the system of property rights whereby owners can grant special rights of use.

If, however, you use my car knowing that you do not have permission, you commit a crime against me. You intentionally deny me my property rights. Your action then implies that property rights do not apply to me; that I do not have the right to have property rights, i.e. that I lack “the very capacity for possessing rights”.Footnote 42 That, in turn, means that you deny, through your action, the norm of mutually respected property rights. You accord yourself more rights than you accord me, hence your action means a denial of equality before the law.

We thus finally arrive at a natural meaning of crime: Crime means a denial of mutual freedom. It is a denial of the norm that gives everybody, including the victim, an equal right to have rights. For this reason, crime is a public matter. We are all, in a sense, victims when a crime is committed, because crime is always also a breach of the norm of equal rights for all, which protects everybody’s freedom.Footnote 43

5 Punishment as an Affirmation of Mutual Freedom

I turn now to the natural meaning of punishment. Deontological freedom theories claim that punishment is not merely a conventional instrument for condemning crime; punishment inherently expresses respect for mutual freedom. The following is a Kantian argument for why it is so.

For Kant, right is mutual freedom. The Universal Principle of Right states: “Any action is right if it can coexist with everyone’s freedom in accordance with a universal law”.Footnote 44 This goes for punishment as well. Punishment is right if and only if it is compatible with everyone’s freedom, including the freedom of the person who is punished. This may sound strange. Convicted criminals rarely consent to their own punishment. Punishment is usually forced upon them, against their (free) will.

To this, Kant answers: “No one suffers punishment because he has willed it but because he has willed a punishable action.”Footnote 45 He has willed an action that is punishable under a law that he is obligated to obey. For a law to legitimately require my obedience, it must be “omnilateral”, which Kant defines as “derived from the particular wills of each”.Footnote 46 A legitimate law, then, must be such that each could have rationally willed that it be valid for herself and for everybody else. Only laws that are consistent with mutual freedom can be rationally willed by all. “My external (rightful) freedom is […] to be defined as follows: it is the warrant to obey no other external laws than those to which I could have given my consent.”Footnote 47

This implies that I do not have the right to disobey laws to which I could have given my consent. I do not have the right to disrespect mutual freedom. And even stronger: I do not have the right not to be hindered in disrespecting mutual freedom. Such a “hindering of a hindrance to freedom”Footnote 48 is indeed consistent with mutual freedom. Without the right to protect right by force, mutual freedom would be an illusion. Anyone could steal and kill with impunity. Society would be lawless. Freedom and legitimate coercion are therefore two faces of the same coin. In Kant’s words: “Right and the authorization to use coercion therefore mean one and the same thing”.Footnote 49

The upshot is that if one accepts the premise that right is mutual freedom, then one must accept that coercion can rightfully be applied in order to assure mutual freedom. Formulated negatively: Nobody can claim a right not to be hindered in hindering right. A criminal cannot appeal to the principle of right in order to claim that she rightfully denies the principle of right. She gives up her legal protection against coercion by committing an act against which coercion is legal.

The state thus has a right to apply force in order to uphold mutual freedom, and this grounds the right to punish lawbreakers. However, we have not yet considered the function of punishment and its natural meaning. How is it that punishment instantiates respect for the freedom of all, including the criminal?

For punishment to instantiate mutual freedom, it must entail treating the criminal as a free and rational person. Specifically, it must entail understanding the offender’s action as an expression of her rationality; a freely chosen act, for which the offender is accountable (presupposing a normal, adult offender). Nobody would deny, of course, that people often act in ways that are less than rational. We all do stupid things, and we are to varying degrees compelled by need or greed or circumstance. These are empirical facts. The relevant question, however, is whether the agent could potentially have acted from reasons; in other words, whether she has the capacity for rationality, and not whether she actually or always uses that capacity to the full extent. When holding someone accountable we impute this capacity to the agent; we blame or praise someone on the assumption that she was not entirely compelled to act in the way she did.Footnote 50 Treating someone as a free and rational person thus means taking a certain normative attitude toward that person: To interpret her actions in light of her capacity to choose them rationally.

It follows that if a person’s action is taken as an expression of her rationality, the action must also be rational for all other rational persons. Hence, a rational action makes an omnilaterally valid claim; a claim to which all rational persons could have consented.

Let us apply this to my example from above: You have taken my car, knowing that you did not have permission. The only way that this action could be omnilaterally valid was if the right to property was invalid. Only if there did not exist a norm that grants owners of things the right to exclude others from use of the thing without permission, could it be rational for everybody to make use of things without permission. But you cannot annul the right to property for all. That would entail a unilateral imposition of your will on others. Your act can only be rational if it is exclusively self-imposing: You do not deny property rights generally, but only for yourself.

This is how we can make sense of Kant’s enigmatic claim that “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”.Footnote 51 The function of punishment is to turn the crime against the criminal–to “bring his misdeed back upon himself”Footnote 52–by taking his message of denial of law seriously as a rational maxim. It is rational, and hence consistent with mutual freedom, to refrain from a right, but not to force others to refrain from their rights. In the words of Arthur Ripstein: The criminal “makes a rule only for herself; the law responds by limiting its application to her alone. Her hindrance to freedom is thus hindered by sealing it off”.Footnote 53

Many have interpreted Kant as advocating a primitive form of retribution.Footnote 54 Allegedly, he intends the objective harm of the crime to be balanced with an equivalent harm in the form of punishment. This critique against Kant’s penal theory, claiming that it entails torturing the torturer and raping the rapist, interprets his retributive principle too literally. Such punishment would not be consistent with the humanity or dignity of the offender. The acts of torture and rape are necessarily degrading, and therefore not something to which the victim could ever rationally consent.Footnote 55

Allen W. Wood claims that Kant’s theory of punishment is inconsistent with Kantian ethics precisely because it cannot be rational to universalize such actions as rape. As Wood says, it cannot be the case that “someone (presumably, the state’s executive authority) is entitled to act toward the criminal in a manner that accords with what his maxim would imply if it were universalizable (which it necessarily is not)”.Footnote 56 This, however, misconstrues the way in which the criminal’s act is universalizable. Abstaining from one’s rights is universalizable; unilaterally changing the rights of others is not. It cannot, therefore, be the latter–the crime–that punishment rightfully universalizes, by raping the rapist, for instance. It must be the former–the withdrawal of the right not to have one’s freedom constrained–that is rightfully universalized. Put differently: a wrong cannot be universalized; only right can.

The act of “bringing his misdeed back upon himself” is thus a way of treating the criminal as rational by re-interpreting his act in such a way that it is consistent with mutual freedom. It is a way of respecting the rationality imputed to the offender, by hindering that the offender’s hindrance to freedom is left standing, as if it were omnilaterally valid.

In Hegel’s phrase, punishment exposes the “nullity” of the crime; it shows that the crime is nothing on this normative level. Crimes have material effects, but they have no effect upon right. A thousand homicides do not change the norm that prohibits homicide. Right, Hegel says, “is precisely what refuses to be set aside”.Footnote 57 The point is the same as Kant’s point. Right is omnilateral; one cannot change it unilaterally.

Where do we stand now, with regard to the question of whether punishment can be justified? At first glance, it seems we could raise the same objection against Kant and Hegel as against Hampton: If right cannot be changed by crime–if crime is a nullity–why bother to express that through punishment? In other words: Why is it so important to negate the message of the crime if we already know that the message is false?

The reason, I propose, is that mutual freedom is not simply a moral principle, in the same way that Hampton’s notion of equal moral worth is a moral principle. Mutual freedom is also a political principle. It is a principle for regulating the political community by according everyone rights to the extent compatible with everyone else’s equal rights. If rights could be infringed upon without being restored, then they would exist only in principle, and not in reality. The right to freedom would exist only on the normative level, where it cannot be negated (where crime is a nullity). As a political principle, however, mutual freedom requires realization in society. People are actually, politically free only if the system of law is upheld, that is, only if breaches of rights are sanctioned. In Ripstein’s formulation: “Punishment upholds the supremacy of the law in space and time”, that is, not simply as a moral principle, but in social reality, in the space and time of our actual lives.Footnote 58

There is a forward-looking dimension to this: When crime is sanctioned, it creates assurance that law will be upheld, deterring prospective criminals.Footnote 59 In this way, punishment serves an instrumental function also on this version of the freedom theory; it causes freedom in society to increase.

And there is a backward-looking dimension (or better, a present-looking dimension): When crime is sanctioned, mutual freedom is upheld here and now, and not as a consequence of future decreases in crime. Punishment instantiates the principle of mutual freedom by ensuring that the unilateral claim entailed by the crime is not universalized, as if it were omnilaterally valid. To return to my example: If you were not hindered in the continuing theft of my car, it would be as if my right to my car was invalid. You would de facto have the power to change my rights, even though it is normatively impossible. By denying the offender success in unilaterally changing other people’s rights, punishment means, in the natural sense, the affirmation of the norm of mutual freedom.

6 Conclusion

I have in this article argued for the existence of an immanent link between crime, punishment and freedom. In order to explicate this immanent link, I have distinguished between what H. P. Grice calls natural meaning and other non-natural meanings that crime and punishment have by convention. Crime means, in the natural sense, a denial of equal rights for all, and because of the logical connection between rights and freedom, crime therefore means a denial of mutual freedom. Punishment means that the offender is treated as a free and rational person, by understanding the maxim of her action as a rational withdrawal of rights for herself, rather than as an irrational claim to withdraw other people’s rights. Punishment thus affirms the freedom of the offender as well as the rights, and hence the freedom, of all.

This presupposes that punishment is conducted in a manner consistent with the dignity of the offender. Sadly, that is not the case in most countries, where, among other things, prison conditions are often degrading to inmates. What I have presented here is thus an ideal theory. I have argued that punishment can be sufficient to express an affirmation of mutual freedom under the right circumstances, not that it is always sufficient. Neither have I argued that punishment is always necessary for such an affirmation. My argument is consistent with the possibility that there could be other ways of sanctioning crime that also affirm mutual freedom–other, restorative, non-punitive ways in which the offender can, in certain cases, be held accountable for her actions.

If that is the case, Feinberg is right to note that the problem of justifying punishment may really be that of justifying our particular symbols of infamy. We must justify which sanction, under which circumstances, best achieves the purpose of the criminal justice system. However, so long as the purpose is to uphold mutual freedom, only those sanctions that instantiate the value of mutual freedom are justifiable responses to crime.

The natural meanings of crime and punishment that I have here described have not, of course, been universally recognized. Hart’s point is valid: If you do not have a concept of rights, you do not have a concept of a natural right to freedom, and neither do you then conceive of crimes as entailing a denial of this concept. This does not mean that we have here merely a conventional meaning of crime. If a man in, say, ancient Babylon raped his wife, the act would mean a denial of her freedom, even though he and others at the time would not conceive of it as such. The man would not be treating the woman in accordance with the right not to be raped, even if they both were not conscious of this denial of right, and hence, of freedom.

In modern societies, where we are accustomed to think in terms of rights and hence to value individual freedom, this natural meaning of crime as denial of equal freedom is much closer to an established meaning of crime, although, as we have seen, we often do accord other meanings to crimes as well. Once we acknowledge the concept of rights, we thereby accept the value of individual freedom, and we cannot then fail to accept that a denial of a person’s capacity for rights means a denial of mutual freedom. Unlike the other conventional meanings of crime, this meaning of crime is not optional once we accept the value of freedom. We could, of course, deny the value of freedom, but at the cost of giving up the notion of rights, and hence, the constitutional state itself. The freedom perspective on crime and punishment is thus not only intuitively appealing because we value freedom highly in our culture–it is also implied by something else we do and ought to value highly: The rule of law.