AustLII Home | Databases | WorldLII | Search | Feedback

University of Melbourne Law School Research Series

Melbourne Law School
You are here:  AustLII >> Databases >> University of Melbourne Law School Research Series >> 2010 >> [2010] UMelbLRS 3

Database Search | Name Search | Recent Articles | Noteup | LawCite | Author Info | Download | Help

Wood, David --- "Punishment: Nonconsequentialism" [2010] UMelbLRS 3

Last Updated: 31 August 2010

Punishment: Nonconsequentialism

David Wood
Faculty of Law, University of Melbourne


A companion to “Punishment: Consequentialism”, and also “Punishment: The Future”, this paper examines various nonconsequentialist attempts to justify punishment, that is, attempts that appeal to claims concerning the innate worth or intrinsic character of punishment, quite apart from any consequential good or benefit punishment may be thought to produce. The paper starts with retributive theories, and turns then to the denunciation and expressive theories, before considering combined communicative- retributive theories.

* * *


A companion to “Punishment: Consequentialism”, this paper is concerned with nonconsequentialist theories of punishment, theories that seek to justify punishment through its supposed innate worth or intrinsic character, and not some consequential good or benefit it is claimed to produce, such as a higher general level of utility or well-being, or the rehabilitation of offenders. A further paper, “Punishment: The Future”, takes up the question of alternatives to punishment, and considers possible developments.

This paper starts with retributive theories, which hold (briefly) that punishment is justified as deserved in virtue of – as a fitting response to – the offender’s wrongdoing. The paper considers next the denunciation theory, according to which (again briefly) punishment is justified through its role in denouncing the criminal’s wrongdoing, forcefully voicing the community’s disapproval and rejection of his conduct. The paper then turns to two influential off-shoots of the denunciation theory. One is the expressive theory, which may be just denunciation under a different guise, though some versions go further in investigating the emotions that punishment is taken to express. The other off-shoot is the communicative theory, the most prominent versions of which do not simply replace the notion of expression with that of communication, but rather bring in retributive considerations, so seeking to merge the two traditions (and hence are referred to here as “communicative-retributive theories”).

However, all these theories face the same, serious challenge, if in different forms. The main task for communicative-retributive theories, as for the denunciation theory and also expressive theories, is to explain the link between the censure or condemnation element of punishment and the sanction or hard treatment element (Kleinig, Feinberg, Husak 92-4). Why not just single out the offender, reprimand and formally convict him, and maybe, at most, subject him to purely symbolic punishment? Why also impose some sanction or form of hard treatment, such as imprisonment or a fine? Involving, as it does, deliberately harming individuals, how can punitive hard treatment possibly be justified?

This problem similarly faces retributive theories of punishment, and indeed any theory with a retributive element. One can likewise ask why a fitting retributive response to wrongdoing need take the form of, or include any element of, hard treatment. The problem therefore confronts communicative-retributive theories in virtue of both elements. Jean Hampton has recognised that retributive responses to wrongdoing need not require punishment: “retribution includes all sorts of responses to human beings, only some of which are punitive” (Hampton 1685; and see “Retribution” below). However, for those who favour the abolition of punishment, the question is why a retributive response need ever be punitive.


Considerable work has been undertaken on retributivism since its resurgence in the 1970’s in penal theory, and social policy and popular sentiment more generally (Galligan 144-6; Duff, 1996:1-2). Numerous versions have been developed (Cottingham, Walker 1999). This section can consider only a few, although as indicated, the paper returns to retributive theory later in considering combined communicative-retributive theories (see “Communication, and Retribution Again”).

The simplest form of retributivism sees retribution as a basic, unanalysable, intuitively obvious, moral principle: “[t]he core of the idea of retribution is the moral notion that the wrongdoer ought to be punished” (Galligan 152); “[a]t the heart of retributivism is the contention that it is the wrongness of the criminal act that justifies the imposition of punishment on the offender” (Golash 49); “[g]uilt deserves punishment for the sake of justice” (Sadurski, 1985 (1): 223). The main drawback with this version of retributivism lies in its very simplicity – it provides no guidance as to how it is to be applied (Galligan 153, Wood D. 545-6).

A second view, traceable to Hegel (Hegel 119-131,Wood A. 109), Brudner 231-5), holds that punishment annuls, cancels out, rectifies or repairs the crime – punishment is the righting of wrong (Wood A. 110, Golash 50). This view goes beyond the basic principle view by offering an explanation of sorts, for why wrongdoing requires punishment, namely to annul, cancel out, rectify or repair the crime (Wood D. 547-9).

A third view, which goes back to Kant (Kant 472-8) in holding that justice “inheres in maintaining a certain relationship between members of the community” (Galligan 155), seeks to spell out the connection between the wrongful conduct that warrants a punitive response and the punitive response itself. In committing a crime the injurer upsets a pre-existing moral balance between the offender and other, law-abiding members of the community (Morris). To use other metaphors, the offender gains an unfair or illicit advantage, incurs a debt to society, or reaps a windfall. In undergoing fair, proportionate punishment, the balance is restored – alternatively, the offender is denied his advantage, repays his debt to society, returns the windfall. This view may be seen as developing the previous view in offering an answer, of sorts, to the question of how the punishment annuls the crime, namely through restoring the moral balance, and so on. Alternatively, it may be seen as bypassing talk of annulment and the like, and returning to the first view, answering directly the question of why crimes warrant punishment.

Moving beyond these three views, at least, as so briefly discussed above, Hampton’s writings on punishment are particularly interesting. While she developed different theories at different stages of her tragically brief career, attention is restricted here to her article “Correcting Harms versus Righting Wrongs: The Goal of Retribution” (Hampton). Hampton distinguishes there between two types of wrongs, namely those that cause harm or loss only (whether material or psychological), and those that cause, in addition, what she calls “moral injury” (Hampton 1692, cf. 1679, 1666). Wrongs that cause harm or loss alone only have to be repaired. However, wrongs that cause moral injury require a retributive response to right the wrong. The retributive response repairs the moral injury, just as civil law damages repair the wrongful harm or loss: “each compensates a different form of damage” (Hampton 1698).

To raise various points, first, although Hampton sees retributive justice as a matter of repairing moral injury, her theory could also be treated as an instance of the “annulment of wrongs” view of retributivism, in that it is through the repair of the victim’s moral injury that, she claims, the retributive response rights (Hampton: e.g. 1663-4) the wrong.

Secondly, a distinctive strength of Hampton’s account lies in its placing the victim at the moral centre, by seeing the primary or significant relation in punishment as that between victim and injurer, rather than injurer and fellow non-injuring citizens. This contrasts starkly with the standard criminal law account of criminal wrongs as primarily public wrongs, wrongs against the community, and only secondarily wrongs against individual victims. Indeed, this public nature is widely regarded as essential to justify the state’s role in punishing, and hence prior to that criminalising, conduct which causes such wrongs. Hampton’s theory opens the door to the conclusion that it is purely the extent or degree of wrongfulness of properly criminalized wrongs that justifies their criminalisation, whether or not such wrongfulness grounds their public nature.

Thirdly, although Hampton’s notion of “moral injury” may be highly appropriate in the case of the most heinous and despicable offences such as murder and rape, it could be questioned in the case of lesser offences, for instance, “a theft by a relatively poor person from a wealthy person” (Duff 1996: 37). However, far from being a ground for criticism, this appears to be a strength of Hampton’s account, precisely because it is highly pertinent to ask whether less seriously wrongful conduct should be criminalized, rather than subjected only to regulatory or civil liability. Indeed, the less serious the wrong, the more pertinent this question is. The issue of degrees of moral injury requires consideration, but Hampton does not appear to do this.

Fourthly, while not explicitly setting out to consider the issue, Hampton takes a significant step towards putting forward a criterion of criminalisation, by limiting such wrongs to those that cause moral injury. However, this is at most a necessary, not sufficient condition. While wrongs that cause moral injury call for a retributive response, such a response, as she significantly argues, need not be punitive (Hampton 1659-1660, 1685-1698). This raises the question of what further is required – is sufficient – for wrongful conduct to warrant criminalisation, and whether this lies in some threshold degree of moral injury.

Since she is not concerned to provide a criterion of criminalisation, Hampton does not consider what counts as sufficient moral injury for the responsible conduct to warrant criminal punishment, and hence prohibition by the criminal law. Neither, for the same reason, does she examine whether, to warrant criminalisation, the relevant conduct must cause not just some moral injury, but moral injury of a certain threshold of seriousness. Since, again, Hampton does not consider degrees of moral injury, or even ask how wrongful the conduct must be, how much moral injury it must cause, she does not provide even an abstract answer to the question of how wrongful the conduct must be, to warrant criminalisation. Neither does she consider how severe the punishment must be to repair the victim’s moral injury, and hence right the victim’s wrong.

Following on and fifthly, Hampton still requires some moral balance between crime and punishment – between the conduct warranting the retributive response and the retributive response itself – in order to determine whether a retributive response is proportionate to the wrongfulness of the conduct, and more specifically, whether this response need take the form of punishment, and if so, punishment of a certain degree of severity. Hampton requires, that is, a moral balance between offender and victim, rather than offender and the community.

However, such a balance appears vulnerable to the very criticism that Hampton levels against Herbert Morris (Hampton 1660-1; Morris). Morris, Hampton objects, mistakenly holds that what is morally objectionable about wrongs warranting punishment lies in the distributive injustices they cause, in the offender’s gaining an undeserved or unwarranted advantage over other, law-abiding members of the community, in short, being free riders. Instead, Hampton says, their wrongfulness lies in the very real and grave wrong the offender inflicts upon the victim. It is not that rapists and child abusers do something that we are all sorely tempted to do, but stoically refrain from doing – indeed, the idea is “indecent” (Hampton 1661). The wrongfulness of rape and other serious crimes lies not in such a comparative matter, but in their sheer wrongfulness, or as Hampton puts it, in the moral injury they cause.

Indeed, the objection to Morris seems to be (not that Hampton puts it quite like this) that he targets the wrong moral balance, namely the “general” moral balance between the injurer and other, law-abiding, members of the community (including the victim), instead of the “local” moral balance between injurer and victim. This raises the question of whether, her criticism of Morris notwithstanding, Hampton likewise offers an account of retributive justice that sees it as a species of, or reducible to, distributive justice (Sadurski, 1985 (2); Wood D. 549-551), albeit a “localised” distributive justice between offender and victim, rather than a “general” distributive justice between offender and community.

Finally in this discussion of retributivism (however see also “Communication and Retribution Again’, below), it has been asked whether retributive punishment is a requirement of justice (in particular, to victims and offenders), or more generally a matter of social policy, the reasonableness of which “must be considered in conjunction with other social policies that may reduce or increase the occasions for such punishment” (Golash 94). Although Hampton appears to take a justice rather than social policy view of retributivism (consider her very use of the term “retributive justice” (Hampton e.g. 1660), she acknowledges that the obligation to provide a retributive response is not absolute, and that there are “times when a society is right to put other concerns before retribution” (Hampton 1700). For example, rather than responding to slanderous speech by means of punitive damages in tort law, a society may decide that freedom of speech is so important as to “morally preclude[-] [it] from inflicting retribution” (Hampton 1700) in this way.


The nineteenth century judge and jurisprudential scholar, James Fitzjames Stephen provided the classic statement of the denunciation theory in writing that punishment aims to satisfy “in a regular public and legal manner”, that “feeling of hatred and the desire of vengeance” that crime “excites in healthily constituted minds” (Stephen 152). Or as Lord Denning put it: “[t]he ultimate justification of any punishment is not that it is a deterrent but that it is the emphatic denunciation by the community of a crime” (Cottingham 245). Punishment is justified insofar as it represents, symbolizes, or gives dramatic statement to the community’s deepest moral beliefs, and voices the society’s abhorrence, resentment, and disgust at the crime.

The obvious question the theory raises is, quite simply: why denounce? What is the point or purpose of doing so? What makes denunciation a principle of social action, which in some form is the state’s right, indeed duty, to implement – and, moreover, through the criminal law and punishment (Tasioulas 287 and n. 11; Hart 66; Williams 100)?

A parallel question arises not just with expressive and communicative theories, but more fundamentally with retributive theories, and any theory with a retributive element. Why need the appropriate retributive response take the form of hard treatment, and furthermore hard treatment of the type and amount that the particular retributive theory in question demands? Why is that the fitting amount of hard treatment, proportionate to the seriousness of the crime or the extent of the offender’s culpability? Hampton may well argue (“Introduction”) that a retributive response need not be punitive, require hard treatment, but the question here (as raised at the end of “Introduction”) is why an appropriate retributive response need ever be punitive, require hard treatment.

To explain the point of denouncing, the denunciation theorist – or “denouncer” (Walker 1980: 28) for short – has to give content to his aim. But can he identify any intrinsic benefit in denouncing criminal wrongdoing, let alone one with the moral weight to support punishment, and as a state institution at that? What makes punitive denunciation a principle of social and indeed state action? Alternatively, is the justification for denouncing criminal activity purely consequential, lying in affirming central, deeply-held social values, thus strengthening them and in so doing social cohesion or solidarity as well (see ‘Punishment: Consequentialism”: “Strengthening Social Values”) – which in turn can be expected to reduce criminal motivation, and hence crime itself? The denouncer thereby becomes a pluralist, in being a harm-reductivist or “reducer” as well. Less flatteringly, he becomes “simply a reducer who believes in a particular technique” (Walker 1980: 28), and so can no longer lay claim to put forward a separate theory of punishment.

A vital issue, then, is the content of the relevant values (a prior question is what are the relevant values), whether they warrant strengthening, and moreover, justify the use of the power and authority of the state, and its system of criminal law and punishment, to do so. (See further “Punishment: Consequentialism”: Strengthening Social Values”.) There are appalling values. Consider the suggestion that some classes or groups of people are morally inferior in virtue of their race, religion, sex or sexual preference, or because they are deemed primitive or uncultivated, or for whatever reason fit only to be demeaned and vilified. It is scarcely justifiable to nurture social cohesion and solidarity (however they are to be assessed) through such means, and in particular by inciting contempt and hatred for “outsider” groups, such as Jews, gays and gypsies.

One line of attack, then, is that denunciation is not an end in its own right, but only a means to strengthening social values, and thereby to the further end of crime-reduction. Furthermore, a would-be denouncer sensitive to liberal concerns may have to concede that, whatever the merits of strengthening social values and cohesion, this is not a legitimate state function, at least not for its system of criminal law and punishment.

A radically different response holds that, rather than collapsing into harm-reductivism, the denunciation theory presupposes, and is at best a mere adjunct to, retributivism – again, conduct only warrants denunciation in virtue of its being sufficiently wrongful to deserve a punitive retributive response (Walker 1980: 29).


Denouncers made do with the ordinary feelings and emotions of ordinary people, and did little to examine them or their underling values. Expression theorists, in contrast, turned their attention to the emotions and values to be expressed through punishment. An important fear to be allayed was that, if appropriate “retributive” emotions could not be identified, one is left only with a lust for revenge (but see Zaibert). As Anthony Duff points out, Stephen scarcely aided his cause by being rather cavalier about how he described the emotions that denunciation was supposed to voice, referring indiscriminately to “the feeling of hatred – call it revenge, resentment, or what you will” (Stephen 152, Duff 1996: 29). In examining such emotions, expression theorists drew upon, and in turn contributed to, a general awakening of interest in the emotions, a topic largely ignored in the western philosophical tradition (with some notable exceptions, such as Hume). Expression theorists – even if they did not necessarily see themselves as such – identified such relevant moral, virtuous or “retributive” emotions (Mackie 3; Murphy, in Murphy and Hampton 1-10) as the offender’s feeling of guilt (Moore 213-5), the victim’s and community’s resentment at the crime (Feinberg), “moral anger” (Golash 60-71) or even “retributive hatred” (Murphy, in Murphy and Hampton ch. 3; Hampton, in Murphy and Hampton ch. 4).

However, as communication theorists pointed out (Duff, 1996: 32-3), in concentrating on the moral emotions that punishment was supposed to express, expression theorists left unexamined the notion of expression itself, and hence remained unaware of its difficulties. Expression can be totally one-way. It requires an “expresser”, but not a recipient, let alone someone capable of responding. One can express whatever one likes into the night sky. In contrast, while a person may be only the bare recipient of an expressive act, he is a participant in a communicative process, which is essentially interactive, requires two-way traffic, addressing the recipient as a rational being and (in principle) an autonomous agent. Essential to communication, but not necessarily to expression, is some cognitive content, an element of understanding or comprehension.


Turning to communication, then, a prominent trend in recent philosophy of punishment is the development of theories which see the point or purpose of punishment as communicating to the wrongdoer, and the community more generally, the “fact, nature and wrongfulness” (Duff 1999: 51) of his crime. As mentioned (“Introduction”), the leading communication theories are in fact combined communicative-retributive theories, which set out to improve upon both existing retributive, and also denunciation and expressive theories. Indeed, such theories often include not just communicative and retributive elements, but other elements, for instance, of rehabilitation, harm-reduction or utility, and even non-punitive – indeed, anti-punitive – considerations, such as mercy.

However, the topic of communication is certainly not confined to retributive theories, or theories with some retributive element. All theories of punishment have some communicative component (Duff 1996: 50, and 1999: 48). For instance, deterrence theories and theories with some deterrence element, must demonstrate how deterrence justifies, or helps justify, actual punishment, as opposed to perfectly harmless measures – “smoke and mirrors” tricks of an illusionist, and the like – which succeed in deceiving people into believing that sentences are actually carried out, thereby producing the same general, if not specific, deterrent effect, as if they had been (Lacey 29). (Specific deterrence could be included where offenders themselves are deceived into thinking that real punishment has been inflicted on them, such as, fancifully, through the use of drugs, or brain surgery which implant false memories in them.) Less harm is done overall, through punishment not actually being inflicted, and so such systematic deceit can be justified on utilitarian grounds at least – unless the fact of deceiving so many people is accorded sufficient disutility to outweigh the utility of the deterrent benefit.

The main issue for a communicative-retributive theory is whether it looks to the condemnation or censuring role of punishment to justify the element of hard treatment – whether the theory sees hard treatment as justified through its role in carrying out what it sees as the communicative purpose of punishment, or whether the theory looks to some quite independent justification, such as deterrence.

On the one hand, “internalists”, such as Duff and John Tasioulas, hold that the very same value which justifies the condemnation or censure, must also justify its communication through hard treatment. As Duff puts it, “the communicative purpose runs all the way down, even to the justification (at I should emphasise, the level of ideal theory) of particular kinds of hard treatment punishment” (Duff, 1999: 51, emphasis added).

On the other hand, “externalists” deny that it is necessary, or even desirable, to look to the communicative or censuring role of punishment to justify hard treatment. Rather, they hold, its justification must lie in some independent value, such as harm-reduction. Von Hirsch, for instance, holds that hard treatment is to be justified, if at all, by another suggested function of punishment, namely the need for a “prudential supplement” (von Hirsch 1999: 70) to “keep predatory conduct within tolerable limits” (von Hirsch 1999: 69; von Hirsch and Ashworth 92). (In this regard, von Hirsch’s theory is consequentialist, illustrating how arbitrary it can be to try to rigidly demarcate consequentialist and nonconsequentialist theories. What presumably makes von Hirsch’s theory nonconsequentialist overall – if one can speak in such terms – is that, as he puts it, the deterrence or “prudential supplement” element in his theory is subject to his “censuring framework” (von Hirsch and Ashworth 22-24, 25; von Hirsch 1993: 14), as will be seen shortly.)

As von Hirsch explains (to take him first), people are not angels (von Hirsch 1993: 13; von Hirsch and Ashworth 23), and cannot be expected to be motivated solely by moral concerns and considerations, with no regard for their own interests. But neither do they respond only to sanctions and the threat of sanctions, and so require “tiger” (von Hirsch 1993: 11, 12, 14), “brute” (von Hirsch 1993: 13; von Hirsch and Ashworth 23) or “beast” control (von Hirsch 1999: 70; von Hirsch and Ashworth 18, 26, 137). Taking people as they are, neither angels nor tigers, but somewhere in between, one cannot appeal only to people’s capacity for rational moral motivation as one can with angels, but must appeal also to their capacity for prudential motivation. A “supplementary prudential incentive” (von Hirsch 1993: 13, cf. 14; von Hirsch and Ashworth 23-4, 34) – or alternatively “disincentive” (von Hirsch 1993, 13, 43, cf. 42) – is required, in addition to the element of censure or condemnation in the form of a sanction or hard treatment of some kind.

Von Hirsch’s angels cannot be perfect angels, of course, otherwise they would never lapse, not even occasionally, into criminal conduct. The point is rather that, on the infrequent occasions they do so lapse, at most formal condemnation or censure, with perhaps some merely symbolic punishment are sufficient to make them desist. Perfect angels are presumably motivated sufficiently by moral concerns so as not to require any prudential supplement to deter or dissuade them from engaging in criminal conduct in the first place. However, ordinary mortals, neither angels (whether real or tarnished) nor tigers, but somewhere in between, may at times require such a supplement in the form of a sanction. With sufficient strengthening of our capacity for moral motivation, the sanction or hard treatment could steadily diminish to nothing – all members of the community being elevated to the status of at least tarnished angels, formal condemnation and purely symbolic punishment come to be sufficient. (Presumably, the final stage in this ‘moral progress’ – in the operation of such a “decremental strategy” (von Hirsch 1993: 40, 45-6; cf. Braithwaite and Pettit 140-3), is that people become perfect, and not merely tarnished angels, and simply no longer engage in any form of criminal conduct.)

Von Hirsch is emphatic that the need for a prudential motivation supplement is the sole purpose or justification of the criminal sanction. No additional justification is provided by any possible role of expressing or communicating the censure, or its moral message, more effectively or forcefully, even if this is the case. Von Hirsch says (and likewise Duff, as will be seen shortly) that censure can be communicated equally well “in a purely (or primarily) symbolic mode” (von Hirsch 1993: 14), through “purely symbolic means” (Ashworth and von Hirsch 23).

What connects the elements of censure and sanction, and provides coherence to his theory, von Hirsch holds, is that the sanction, together with its prudential supplementary role, operates within a “censuring framework” (von Hirsch and Ashworth 22-24, 25; von Hirsch 1993: 14). This requires, at the very minimum, that the severity of the sanction is limited by the amount of the deserved censure, and hence the extent of the offender’s wrongdoing or culpability, thus excluding especially harsh sentences, beyond the limits of proportionality, whether for deterrence, incapacitation, rehabilitation, or any other purpose. It is simply too bad, von Hirsch thinks, if this means that the desired rate of crime reduction is not achieved. Society must then look to other crime-reductive means and methods, outside the criminal law and criminal justice system, or just put up with the higher crime-rate. Convicted persons are not to be used as a means to this (or any other) end.

Von Hirsch further warns that the hard treatment element in punishment may, if too hard, “drown out” the very moral message of censure or condemnation it is supposed to communicate (von Hirsch and Ashworth 77; von Hirsch 1999: 75, cf. 71), so that the recipient is aware only of this treatment, and not its meaning, the message it is supposed to convey (the very reason for being subjected it) – or alternatively, aware of the hard treatment only at the expense of its message of censure. Von Hirsch is clear that his theory excludes this possibility.

The pertinent question is whether “drown out” cuts in at a lower level than proportionality. If not, it plays no practical role, proportionality having done its work for it, in excluding any punishment that would “drown out” its moral message. (Another, at least theoretical, possibility is that “drown out” cuts in at different levels according to the type and nature of the crime, or some other consideration.) As von Hirsch points out, “[t]he supplementary prudential disincentive is just that, supplementary” (von Hirsch 1993: 43). It is not intended to replace the moral message, and neither to undermine the effectiveness of individuals’ capacity for rational moral motivation and agency. Presumably, individuals should still strive to strengthen this capacity, and be more like angels, less like tigers.

To raise just two criticisms, first, Tasioulas objects that censure is “drastically downgraded” (Tasioulas 290) within von Hirsch’s theory, because it does not look to the reasons for censuring to justify, not only the censuring itself, but how it is communicated to the wrongdoer, namely, through hard treatment. Rather, the hard treatment element of punishment is justified on crime-preventive grounds. Tasioulas sees the hard treatment element of punishment not just as one way among others to communicate censure, but as internally related to the element of censure (Tasioulas 286).

However, von Hirsch could well respond that it is his theory that gives censure a pre-eminent place, indeed, a higher status than do Duff and Tasioulas, precisely because he does not tie censure to hard treatment as they do. As noted, von Hirsch’s theory envisages the possibility of censure without sanction or hard treatment. Certainly, for von Hirsch’s “decremental strategy” to reach its natural endpoint of the abolition of punishment, would require people’s capacity for moral motivation to become much stronger – to the point where prudential motivation, and hence hard treatment, is no longer required, formal conviction and at most purely symbolic punishment being sufficient. “Censure without sanction” can only be envisaged if censure is the crucial element of punishment, as itself effectively producing compliance through its own intrinsic moral force.

Far from downgrading censure, if anything, it seems, von Hirsch downgrades the sanction or hard treatment element of punishment, relegating it to serving only prudential supplement purposes, so playing no intrinsic or essential role in communicating censure. Indeed, von Hirsch gives censure the highest possible moral standing by denying it any essential relation with hard treatment – ‘liberating’ it from any such connection, and hence ensuring for it genuinely independent status. The very fact that he envisages the possibility of censure without sanction, as the natural endpoint of his “decremental strategy”, scarcely makes it plausible that he would “downgrade” censure (not that the allegation is that he does so intentionally).

Secondly, Tasioulas criticises von Hirsch’s theory for a supposed lack of coherence, a fault he locates in the same source as the alleged downgrading of censure, namely von Hirsch’s appealing to independent, crime-preventive considerations, to justify the hard treatment element of punishment. More specifically, the allegation is that the leeway required for the operation of the prudential supplement means that von Hirsch’s theory is not coherent or sufficiently coherent – his “censuring framework” does not deliver on this score. Tasioulas dismisses the requirement to operate within this framework as an “ad hoc stipulation” (Tasioulas 291), the shortfall – if one can speak in such terms – not being made up elsewhere.

In response to von Hirsch’s “invok[ing] the preventative function of punishment in order to justify the communication of censure through hard treatment”, Tasioulas alleges that “immediately the danger of incoherence looms because what is introduced as a supplement to censure threatens to subvert the fundamentally communicative character of his theory” (Tasioulas 286). The immediate question is, simply, what counts as sufficient coherence. Coherence is hardly a tangible, quantifiable or measurable commodity. Von Hirsch holds that his censuring framework provides all the coherence that is required, that it is sufficient that the hard treatment element of punishment and its deterrence or crime-prevention role operates within, and hence is subject to, this framework (von Hirsch and Ashworth 22-4). Even if the censuring framework offers only “weak coherence” or “loose integration”, as it could be put, this is sufficient.

Furthermore, requiring a tighter relationship between the censure and sanction elements of punishment could be overly restrictive, stultifying even, negating the very flexibility that von Hirsch’s censuring framework sets out to provide (and that any pluralist theory needs, in order to properly accommodate the various elements and considerations it acknowledges). Moreover, such a tighter relationship threatens to frustrate the progress of the decremental strategy (a point taken up shortly in relation to Duff). Coherence (or integration) is not the only requirement of a theory of punishment. Von Hirsch could appeal to the need for flexibility to respond to changing social circumstances, in particular, the strengthening of potential criminal wrongdoers’ capacity for moral motivation, thus reducing reliance upon their capacity for prudential motivation. A related possible change (although this presents formidable problems of empirical assessment) is that deterrents could come to be perceived more strongly, so that, say, a five year prison sentence has the same deterrence effect as a ten year sentence had previously.

Turning to Duff’s version of the communicative-retributive theory, he points out that his account of punishment sets out to be “more ambitiously communicative” (Duff 2001: 88) than von Hirsch’s. For Duff, as mentioned, justified punishment functions as a response to criminal wrongdoing that seeks to communicate to the offender the wrongfulness of his conduct. However, this is only the beginning of a highly elaborate account of what punishment should seek to achieve. As Duff explains, his theory is both backward- and forward-looking. The message communicated by punishment must be a message focused on, and justified by the offender’s past offence. However, it is only forward-looking considerations, he thinks, that justify the use of hard treatment to communicate censure.

According to Duff, to draw on his own summary (Duff 2000: 412-414), a judicial sentence “should have several closely related aims” (Duff 2000: 412). First, “it should aim to communicate to the offender, to bring home to him more forcefully than his conviction might have done, the censure that his crime deserves” (Duff 2000: 412). Secondly, the punishment “should aim to persuade him to accept the censure as justified – to face up to and to recognise what he did as wrong – which will also be to repent that wrong” (Duff 2000: 413; his emphasis). Thirdly, the sentence “should aim to persuade him to see the need to reform his future conduct, so as to avoid such wrongdoing – and help him to begin to do so” (Duff 2000: 413). Fourthly, “it should aim to provide some reparation to those whom he has wronged, and thus to reconcile him with them” (Duff 2000: 413). “What matters here”, Duff continues, “is not material reparation...but moral reparation for the moral wrong that was done” (Duff 2000: 413; his emphasis). This requires at the very least an apology, both to the victim, and to the community “whose public values he has flouted” (Duff 2000: 413). (See further, the discussion of restorative justice in “Punishment: The Future”: “Condemnation without Hard Treatment’.) Fifthly – and this appears, rather than a separate stage, to mark the conclusion to the fourth stage – the punishment “should aim to reconcile the offender with those whom he has wronged, through this process of censure and symbolic apology” (Duff 2000: 413). Sixthly and finally, though Duff does not mention this in this particular exposition, it should seek to restore him to full membership of his moral community.

This process – of recognition, repentance, reform, reparation, reconciliation, and restoration – could be referred to, for obvious reasons, as the “R-Process”. Indeed, in summarising his account, Duff says that punishment performs its communicative role in three ways, which he appropriately calls “the three R’s” (Duff 2001: 107-112), namely in enabling or facilitating, first, the offender’s recognition of, and repentance for, his wrongdoing, secondly, the offender’s moral self-reform, and thirdly, the offender’s restoration to full membership of the moral community.

There is nothing fixed in concrete about this description. Stages could be added (for instance, Duff elsewhere includes remorse) or divided into two or more if not so easily deleted or amalgamated. Also, while some stages, such as recognition and repentance seem relatively discrete, others, such as remorse, are rather ongoing states of affairs with no clearly identifiable beginning or end.

Turning from exposition to criticism, a crucial question is whether Duff places an unnecessary impediment in the progress of a “decremental strategy”, of steady reduction in the severity of sentences, by tying sanctions as closely as he does to censure, in seeking to justify hard treatment “internally” to the condemnation or censuring function of punishment. Punitive hard treatment could become progressively less harsh, and possibly punishment as a state institution could eventually be abolished, but for its “integration” – its “internal” (Duff 2001: 30, 89) and “intrinsically appropriate” (Duff 2001: 87) relation – with the censure or condemnation function of punishment. Duff appears caught by existing conventions (Duff 2001: 276; Feinberg 400; Tasioulas 289) about what retributive justice requires, what counts as proportionate, retributive punishment, that is, punishment that treats the offender neither more nor less severely than his conduct or culpability warrants. Such conventions could function to keep punishment levels up, even though social changes mean that existing punishment levels are no longer required for deterrence or crime-reductive purposes. These social changes include particularly the strengthening of moral motivation, as mentioned, so that less reliance needs to be placed on prudential motivation, as well as the strengthening of prudential motivation itself.

Von Hirsch’s theory simply does not give rise to this problem, and thus appears preferable in this regard. It holds, as seen, that censure can be communicated without hard treatment, “in a purely (or primarily) symbolic mode” (von Hirsch 1993: 14), the function of hard treatment being quite separate, namely to provide an independent source of motivation for compliance.

On the other hand, and to Duff’s credit, where retributive conventions – conventions about the amount of hard treatment required for retributive justice purposes – become more demanding, require more severe punishments (as seems to have occurred in various countries, including the United States, over the last few decades), his theory could slow down the progress of such an “incremental strategy”, or any retribution-based trend of increasing punishment-severity.

A more basic criticism is precisely that which Tasioulas and Duff himself level against von Hirsch, namely that contrary to Duff’s claim to propose a “unitary” (Duff 2001: 89), indeed “teleological” theory (Duff 1996: 45-51), his theory lacks coherence. This is because the theory has two parts – one backward-looking and retributive and the other forward-looking and consequentialist – which, despite Duff’s claims, cannot be united or integrated. Indeed, it is not just that Duff’s theory (like von Hirsch’s) is consequentialist (in virtue of its forward-looking aspect of element), it is a far less attractive as a consequentialist theory than von Hirsch’s, in focusing on a comparatively narrow set of consequences, concerning only offenders (and penitent offenders at that), and only in relation to their moral reform. Von Hirsch’s theory, in contrast, is concerned with the consequences of criminal conduct to members of society generally, as potential victims of crime. Indeed, it could be suggested that von Hirsch’s prudential supplement could be ‘married’ with, what is really, according to this criticism, Duff’s offender moral reform supplement. Indeed, one could also bring in concern for victims, and their need for vindication (Hampton 1686), and so suggest a three-way marriage – if this is not to stretch the metaphor too far – of all major parties, offender, victim and the community.


Punishment involves deliberately harming individuals, and therefore stands in very strong need of justification. A companion to “Punishment: Consequentialism”, this paper examined nonconsequentialist theories of punishment – retributive theories, the denunciation and expressive theories, and finally communicative-retributive theories. A further paper (“Punishment: the Future”) looks ahead to possible developments in thinking and theorising about punishment, as well as to alternatives to punishment.


For very helpful comments and criticisms, I wish to thank Jes Bjarup, Zack Hoskins, John Lowndes, Paul Roberts, John Tasioulas, Andrew von Hirsch and in particular, Greg Roebuck.

Wiley-Blackwell Publishing gives David Wood proper credit as the original author of this article. Wiley-Blackwell also acknowledges that the article draws upon pages 545, 547 and 549-53 of an earlier article by the same author: Wood, D., “Retributive and Corrective Justice, Criminal and Private Law”, (2005) 48 Scandinavian Studies in Law (Perspective on Jurisprudence: Essays in Honor of Jes Bjarup) 541. Reproduced by kind permission of Scandinavian Studies in Law.

Works Cited

Braithwaite, John and Pettit, Philip. Not Just Deserts: A Republican Theory of Criminal Justice. Oxford: Oxford University Press, 1990.

Brudner, Alan. The Unity of the Common Law: Studies in Hegelian Jurisprudence. Berkeley: University of California Press, 1995.

Cottingham, John. “The Varieties of Retribution”. 29 Philosophical Quarterly (1979): 238-246.

Duff, R.A. “Penal Communications: Recent Work in the Philosophy of Punishment”. Crime and Justice: A Review of Recent Research (ed. Michael Tonry; University of Chicago Press) 20 (1996). 1-97.

------------. “Punishment, Communication, and Community”. In Mattravers, Matt, ed. Punishment and Political Theory. Oxford: Hart, 1999. 48-68.

-------------. “In Defence of One Type of Retributivism: A Reply to Bagaric and Amarasekara”. 24 Melbourne University Law Review (2000): 411-426.

------------. Punishment, Communication and Community. New York: Oxford University Press, 2001.

------------. Answering for Crime: Responsibility and Liability in the Criminal Law. Oxford: Hart, 2007.

Feinberg, Joel. “The Expressive Theory of Punishment”. 49 Monist. (1965): 397-423.

Galligan, D.J. “The Return to Retribution in Penal Theory”. Crime, Proof and Punishment: Essays in Memory of Rupert Cross. Ed. C.F.H. Tapper. London: Butterworths, 1981: 144-171.

Golash, Deirdre. The Case Against Punishment. New York: New York University Press, 2005.

Hampton, Jean. “Correcting Harms versus Righting Wrongs: The Goal of Retribution”. UCLA Law Review 39 (1992): 1659-1702

Hart, H.L.A. Law, Liberty and Morality. Oxford: Oxford University Press, 1963.

Hegel, G.W.F. Elements of the Philosophy of Right (trans. N.B. Nisbet, ed. Allen W. Wood) Cambridge: Cambridge University Press, 1991 (originally published 1821).

Husak, Douglas. Overcriminalisation: The Limits of the Criminal Law. Oxford: Oxford University Press, 2008.

Kant, Immanuel. Metaphysics of Morals. In Kant, Immanuel, Practical Philosophy (trans. Mary Gregor; general introduction, Allen Wood).Cambridge University Press, Cambridge, 1991. 353-603 (originally published 1797).

Kleinig, John. “The Hardness of Hard Treatment”. In Ashworth, Andrew and Wasik, Martin, eds., Fundamentals of Sentencing Theory: Essays in Honour of Andrew von Hirsch. Oxford: Clarendon, 1998. 273-298.

Lacey, Nicola. State Punishment: Political Principles and Community Values. London: Routledge, 1988.

Mackie, J.L. “Morality and the Retributive Emotions”. Criminal Justice Ethics 1 (1982): 3-10.

Moore, Michael. Placing Blame: A General Theory of Criminal Law. Oxford: Clarendon, 1997.

Morris, Herbert. “Persons and Punishment”. Monist 52 (1968): 475-501.

Murphy, Jeffrey G., and Hampton, Jean. Forgiveness and Mercy. Cambridge: Cambridge University Press, 1988.

Sadurski, Wojceich. (1) Giving Desert Its Due. Dordrecht: Reidel, 1985.

-----------------------. (2) “Distributive Justice and the Theory of Punishment”. Oxford Journal of Legal Studies 5 (1985): 47-59.

Stephen, J.F (White, C.J., ed.). Liberty, Equality, Fraternity. Cambridge: Cambridge University Press, 1967 (originally published 1973).

Tasioulas, John. “Punishment and Repentance”. Philosophy 81 (2006): 280-322.

von Hirsch, Andrew. Censure and Sanctions. Oxford: Clarendon Press, 1993.

-------------------------. “Punishment, Penance, and the State: A Reply to Duff’. In Mattravers, Matt, ed. Punishment and Political Theory. Oxford: Hart, 1999. 69-82.

von Hirsch, Andrew and Ashworth, Andrew. Proportionate Sentencing. Oxford: Oxford University Press, 2005.

Walker, Nigel. Punishment, Danger and Stigma. Oxford: Blackwell, 1980.

-----------------. “Even More Varieties of Retributivism”. Philosophy 74 (1999): 595-605.

Williams, Bernard. “Moral Responsibility and Political Freedom”. 56 Cambridge Law Journal (1997): 96-102.

Wood, Allen W. Hegel’s Ethical Thought. Cambridge: Cambridge University Press, 1990.

Wood, David. “Retributive and Corrective Justice, Criminal and Private Law”. 48 Scandinavian Studies in Law (Perspectives on Jurisprudence: Essays in Honor of Jes Bjarup, ed. Peter Wahlgren) (2005): 541-582.

Zaibert, Leo. “Punishment and Revenge”. 25 Law and Philosophy (2006): 81-118.

AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback