Introduction

That a criminal offender’s genuine remorse should be treated as a mitigating factor in sentencing is a well-settled legal principle in many if not virtually all jurisdictions.Footnote 1 For convenience, let us call this the ‘remorse principle’. Perhaps because this principle is so well-established, there is comparatively little sustained explanation of its rationale in either case law or jurisprudential commentary. In contrast, critics of the principle have in recent years put forward a number of criticisms of the idea that remorse should be a mitigating factor in sentencing. (See, for example, Duff 2001, pp. 120f; Ward 2006; Bagaric and Amarasekara 2001.) This has perhaps had the effect of putting the remorse principle on the intellectual back foot.

The present article seeks to defend the remorse principle as a proper part of sentencing law. It does so by articulating what I shall argue is the preferable rationale for the principle, namely that mitigation serves as a communicative recognition of the offender’s remorse. While exposition of that rationale will be aided by considering some criticisms of the remorse principle, the main concern here is not to answer critics point by point. Rather, the chief purpose is to work out why we should want to keep and defend the remorse principle in the first place.

It should be stressed that in looking for a rationale or deeper theoretical basis for the remorse principle, I am not assuming that the principle is merely an isolated ‘pre-theoretical intuition’ that must be abandoned as an arbitrary prejudice if no theory can be found to support it. Rather, I would suggest that it is among those deep-seated commitments which can serve to test general theories as much as they are tested by them (cf. Rawls 1999, p. 18, on reflective equilibrium). That is to say, a theory about sentencing may well have less force if it cannot satisfactorily account for the remorse principle.

The course of the discussion is as follows. I first present brief definitional sketches of both remorse and the remorse principle. I then critically consider what I take to be the four basic types of rationale for the remorse principle: the redundancy rationale, the inducement rationale, the reward rationale, and the recognition rationale. These four rationales are not mutually exclusive, though neither do they form a harmonious whole. The first three will be argued to have certain weaknesses and limitations, and to be not as persuasive as the fourth, which views mitigation as a form of recognition of the remorseful offender. This recognition rationale will be seen, moreover, to fit best with a broadly communicative approach to sentencing. Then, to help round out the account of the recognition rationale, I consider two arguments against the remorse principle, neither of which, I shall argue, is ultimately persuasive. Finally, I briefly address the question of whether the remorse principle entails that a lack of remorse should be an aggravating factor in sentencing.

Defining Remorse and the Remorse Principle

Before we consider the possible rationales for the remorse principle, it is important that we have a clear working definition of remorse in front of us. (‘Contrition’ is a virtual synonym for ‘remorse’, as understood here, but it will be simplest to use just the one term.) In this article, then, remorse is taken to be the mental and emotional suffering a person experiences when and as she recognises that she has done a moral wrong, especially one that involves wronging another person.Footnote 2 This remorseful experience will also typically involve a desire to make things right, to apologise, and to mend the harm, and a resolve not to repeat the wrongdoing. We would also normally expect the remorseful person to attempt to realise those desires through some sort of reparative act, such as apology or compensation. Different instances of remorse can properly differ in intensity, and even the same remorse can vary over time. Further, remorse can take many kinds of wronging of another as its object. Not all remorse is over wrongs that are criminal offences and not all criminal offences are proper occasions for remorse. Remorse can also, of course, sometimes (perhaps often) be distorted, deluded, or faked. In this article, we will be concerned with genuine and lucid remorse in response to a criminal offence that is also a moral wronging of another. This is not to minimise the practical difficulties in proving that an offender experiences genuine and lucid remorse, but simply to focus the discussion on the philosophical issues.

Though a remorseful person can at the same time or in relation to the same events experience shame and guilt feelings, remorse is not reducible to either. Shame involves a sense of having failed to realise some valued standard, especially in the eyes of certain others, while guilt feelings are taken to be the pained sense of having disobeyed an authority figure or breached a binding rule.

With regard to the remorse principle, we can define it in fuller terms as follows: where a sentencing judge is satisfied that an offender feels genuine remorse, then the judge, first, must always at least consider whether or not to mitigate the sentence due to the offender’s remorse, and, secondly, should normally so treat that remorse as a mitigating factor. That second limb means that there is a strong presumption that remorse be treated as a mitigating factor and that a failure to do so would require explanation (in a fully articulated set of reasons for sentence). This preserves ultimate judicial discretion, but within certain constraints. A weaker version of the remorse principle would be simply that remorse may be treated as a mitigating factor. This would preserve a large degree of discretion for the sentencing judge. In contrast, a stronger version of the principle would eliminate discretion altogether and require that remorse must always be a mitigating factor. It would appear, though, that the middle ground version of the principle (wherein remorse should, other than in exceptional cases, be a mitigating factor) best reflects settled law and practice.

Does Remorse Make Punishment Redundant?

The first—and perhaps most obvious—rationale to be considered is that the offender, by experiencing remorse, has herself done at least part of what the sentencer would have done. She has thus ‘pre-empted’ the sentencer, and so made redundant at least part of the sentence that would have been imposed. The extent to which part of the sentence has been made redundant should therefore be the extent to which the sentencer should mitigate the sentence. Such a rationale draws upon a general principle of parsimony, which requires that we try to achieve the desired end of a sentence by means of the ‘cheapest rate’ possible and avoid imposing any punishment beyond that which is necessary to achieve that end (see Bentham 1987, pp. 102–104).

Redundancy and Consequentialist Sentencing

This redundancy rationale will generally fit most readily with consequentialist approaches to sentencing. Thus, if one thinks the main aim of sentencing is to deter the offender from re-offending, or to rehabilitate or reform the offender, or simply to prevent or minimise his further offending by incapacitating him, then the remorseful offender seems to have done the work already, at least to some degree. (Not all consequentialist aims fit well with the redundancy rationale, however. The aim of general deterrence, for example, would not seem to be pre-empted by offender remorse.)

The assumption here, of course, is that a remorseful person is less likely to re-offend in the longer term, due to his behavioural and/or attitudinal change experienced as part of his remorse. Of course, genuine remorse does not guarantee that the person will never do anything wrong again. Mere re-offending is not proof of non-genuine remorse. Moreover, the assumption is not that the remorseful offender must be no more likely than non-offenders to commit a relevantly similar crime in future. Rather, the proposition in question here is simply that an offender who is genuinely remorseful about a particular offence will be less likely to commit a relevantly similar offence in the future than he would be if he were not remorseful.

While this is perhaps a reasonable common sense assumption, it appears that there is as yet insufficient positive evidence from psychology to confirm this assumption. As forensic psychotherapist Murray Cox notes, ‘[t]he disturbing fact is that there appears to be no firm evidence that had [the remorseless offender] shown remorse, there is less likelihood of him re-offending’ (Cox 1999, p. 17; see also Horne 1999, p. 31; Gilligan 1999, p. 33). This is not, of course, positive evidence that remorse does not reduce recidivism. Rather, the absence of evidence either in support of or against the proposition simply means that no firm scientific conclusion can yet be drawn. This is clearly an area where there is a great need for good empirical research. In the meantime, particularly in the absence of positive disproof of it, it can be argued that the common sense assumption that remorse normally reduces the risk of recidivism is still legitimate to work with, albeit somewhat tentatively, and that the onus lies on those who would disprove it to advance the relevant evidence.

Redundancy and Retributivist Sentencing

The redundancy rationale does not sit so well, however, with retributivist approaches. It seems simply a misdescription to say, for example, that the remorseful person ‘is already punishing himself’ (Walker 1980, p. 129). This is because the suffering of punishment, imposed from without, is not to be identified with the suffering of remorse, which rises up from within. Though they may share some features, they are ultimately different kinds of suffering, with distinct qualities and, indeed, distinct meanings, and cannot serve as simple substitutes for each other.

If one were to adopt a more communicative, censure-focused conception of retribution, there would still be a difference that forestalls redundancy. The self-rebuke or self-reproach inherent in remorse cannot serve as a simple substitute for the denunciation or censure by another, especially one who occupies a formal office such as that of judge. For all their obvious common elements, such self-reproach and such censure remain different kinds of speech act, and so the former cannot make the latter simply redundant.

No Identifiable Stage of Sentence is Made Redundant

Even within the limits of its applicability, the redundancy rationale appears to assume that there is some specific stage or segment of the sentence that is made redundant by the offender’s remorse, and that it is this excess that is dropped when the sentence is mitigated. An objection may be raised here that in practice, and even in principle, it is not plausible to say that specific and precise stages of a sentence are devoted to specific functions, as if the first year of a two-and-a-half year prison term is for specific deterrence, the second year for general deterrence, the next three months for incapacitation and the final three months for denunciation. Rather, all four aims are intended to be met by the whole two-and-a-half years’ prison time, just as all twenty of a building’s support columns are designed to bear the weight of the building together (cf. Duff 2001, p. 121).

Let us say that the sentencing judge decides that, of that two-and-a-half year sentence, fifteen months can be taken off since she is satisfied that the offender’s remorse makes specific deterrence and incapacitation redundant. Is the judge saying that fifteen months of the original sentence was for specific deterrence and incapacitation? How else, it could be asked, could she have arrived at a discount of fifteen months on the ground that specific deterrence and incapacitation are no longer needed? If this is how the judge reasoned, then in principle it would seem that she could take the analysis further and say that twelve months had been for specific deterrence and three months for incapacitation.

Ultimately, however, it seems implausible that a sentencing judge could explicitly and specifically identify discrete portions or even precise amounts of penalty as having been made redundant by the offender’s remorse, and precisely reduce the sentence accordingly. Most judges, I suspect, would reduce sentence on the grounds of redundancy only imprecisely and quite possibly more to symbolise redundancy than to measure it. This suggests that the redundancy rationale is really more metaphor than a literal account of judicial thinking (actual or desirable). There is nothing wrong with that, of course, but it does point to limitations in the rationale as a complete justification of the remorse principle.

A related point worth mentioning here is that it is not possible to say how much of a sentencing discount remorse should afford, generally speaking. Any general rule that the presence of genuine remorse will ordinarily entitle an offender to, say, a 25 per cent discount, or even a discount within a range of 10 to 50 per cent, is unlikely to be useful. Different cases of remorse will vary in their depth, intensity, effect, meaning, and so on. Any universal rule, or even rough rule of thumb, will almost inevitably ride rough-shod over the details of particular cases. Of course, parity between cases is important, but this does not require a general rule.

The Communicative Inadequacy of the Language of Redundancy

We can see, then, certain limits to the redundancy rationale for the remorse principle. But even where it does fit well enough within an existing sentencing aim, there is still something lacking. What is under-described in this rationale is the nature of that ‘lack of necessity’ and the manner in which ‘unnecessary’ punishment is acknowledged in reducing sentence severity. Casting the matter in communicative terms helps to bring out the issues here more clearly. In a conversation, ‘saying too much’ may sometimes be more than merely idle—it may create embarrassment and awkwardness, or even, in some cases, cause offence and hurt. When we say of someone’s comment ‘That was unnecessary’, we do not normally mean that what was said was irrelevant or a waste of words; there is usually a sense that the comment created an awkwardness in the conversation, and may even have been hurtful or insulting to the interlocutor. The words do not just go off into empty space; they change the nature of the communication and may, depending on the case, turn reproach into insult or abuse. The redundancy rationale leaves this sort of concern unarticulated because the language of redundancy is generally insensitive to the possible significance of the excess. The sentence is seen primarily as a mechanical tool for achieving a set goal; its communicative import is, if it figures at all, only a peripheral matter. This encourages the redundancy rationale to see the excess sentence simply as meaningless waste, as an inefficient mechanical effort that achieves little or nothing, while being more or less unaware of the substantive communicative significance to be found in that excess, such as insult or abuse. We shall return to these themes below.

Can Offering Mitigation Induce Remorse?

On this next rationale, the promise of mitigation of sentence is seen as a proffered benefit intended to induce the offender to come to feel remorse. An inducement differs from a reward in that it is prospective, in contrast to the retrospectivity of reward. Also, a reward is deserved or merited while an inducement is not, as it is simply offered as a means to get the offeree to do something. (Of course, it is possible that a benefit could serve as both an inducement and a reward.)

Deterrence theorists are perhaps the most likely to endorse the inducement rationale because it involves the same logic as deterrence, but simply inverts the concepts: instead of discouraging crime by imposing suffering, we encourage remorse by providing (or, at least, offering to provide) a benefit.

For an inducement to be effective it is usually necessary that the person induced believe that she is at least likely to receive the inducing benefit for doing the thing desired by the inducer. Once the person has done that desired something, however, it is strictly immaterial, from the point of view of the inducer, whether that inducing benefit is actually given, for the inducement has already succeeded. Thus it is, strictly speaking, not the mitigation of sentence, but rather the offer to mitigate sentence that might induce remorse. Nonetheless, there may be good reasons to give the benefit and mitigate sentence. For example, the person induced may cease doing the desired thing if her expected benefit is withheld or withdrawn. Also, failure to provide a promised benefit may weaken the effectiveness of future such inducements (whether of that same person or others) because confidence in those inducements will be undermined.

The logic of inducements is simple enough, but it is also fairly obvious that it does not provide a good basis for the remorse principle. Where the thing to be induced is remorse, and the inducement is mitigation of sentence, then a fundamental contradiction is manifest. Given the characterisation of remorse used in this article, it would appear that a person cannot come to have genuine remorse only because he is induced to have it by the prospect of a lessening of sentence. The object of such a person’s understanding is simply not that of a genuinely remorseful person.

It is different when the law proffers lighter sentences for offenders who plead guilty or help the police in order to encourage such pleas and help. These are, generally, good things which assist the community, and the benefit to the community is forthcoming regardless of why the offender pleaded guilty or co-operated. The benefit is not undermined because the offender only pleaded guilty or co-operated with the police in order to get the lighter sentence. (There is, of course, a problem where an innocent person who believes he cannot successfully contest the charges against him pleads guilty simply in order to get a lighter sentence, but we shall have to leave that aside here.)

With remorse, it is different. Though we welcome the offender’s remorse and indeed may seek to encourage and elicit it, it is nonetheless very much an experience that must come from within the individual herself, as her acknowledgment of what she did. This is not to say that remorse cannot properly be partly a response to the efforts of others to elicit it, for indeed it can. But such remorseful responsiveness is not, if it is lucid, induced by some extraneous thing such as a sentencing discount. It is a response to the nature of what one did. While the nature of what one did can well be ‘brought home’ to one by others’ reactions to one’s deed (such as their horror, anger, reproach, rebuff or punishment), ultimately remorse cannot be lucidly induced by anything other than contemplation of its own proper object: the wrong one has done another. In other words, one may properly elicit remorse by drawing the offender’s attention to his wrongdoing, but one cannot induce remorse with extraneous benefits and at the same time preserve it.

Should Remorse be Rewarded with Mitigation?

In the case of rewarding remorse with mitigation, the reward is seen as due to the offender simply in the light of the offender’s remorse. The mitigation is a deserved benefit for achieving something good. It is not due, for example, because the offender’s legitimate expectation of a benefit would be let down or there is some quasi-contractual obligation to provide what was promised. Moreover, once a person has done the desired thing, it is not immaterial whether or not the reward is given, unlike the case in relation to inducements. It will be a moral lack, an injustice, not to provide the reward. Because of the centrality of desert and merit to this approach, we can see that this rationale for the remorse principle is conceptually related to retributivism. Bad acts deserve to be met with suffering, but good acts deserve to be met with benefits.

The idea of desert can, of course, be cast in various ways here. One is the idea of the deserved reward being earned. Something’s being earned carries with it the notion that that which is done as the earning is in some way a ‘burden’ which is to be compensated by receiving that which has been earned. In this case, the mitigation of sentence serves as the earned reward. But what exactly is the earning? Remorse is by definition a kind of suffering, but is it the kind of suffering that calls for compensation? At first glance, this would seem an inappropriate way to respond to remorseful suffering. It is not, like many, perhaps most, other forms of suffering, a simple burden to be relieved or compensated if possible. It is, rather, something to be ‘gone through’, given that it is an integral part of the person’s acknowledgment of what he did. The lucidly remorseful person, at least, would be unlikely to see his remorseful suffering as a burden calling for ‘recompense’ in the way a worker calls upon her employer to pay her for her labour.

Perhaps, though, the desert here is more a matter of the sentencer feeling a moral pull toward welcoming the good of the offender’s remorse by way of a ‘celebratory’ or ‘congratulatory’ reward. Such a reward is more a celebration of success than a compensatory payment of earnings. Academic or sports prizes are good examples. The academic or sporting efforts were not (simply or solely) burdens to be compensated but rather are valuable achievements in themselves that we wish to acknowledge by way of a celebratory or congratulatory reward. In this way, the reward of mitigating sentence in response to offender remorse may perhaps be more of a congratulatory type of reward in response to a worthy or desirable achievement.

Some may, however, object to the whole idea that remorse deserves any sort of reward, compensatory or congratulatory, on the basis that feeling remorse is what a wrongdoer ought to feel, and it is not appropriate to reward people for simply doing or feeling what they ought to do or feel. (See, for example, Bagaric and Amarasekara 2001, p. 364; Tudor 2005 in reply.) It is true that we do not usually seek to reward those who fulfil negative obligations. The person who got through his day without assaulting anyone is not rewarded for such restraint, which is no more than his duty. But this does not seem true of positive obligations. We may think that the person who lifted the drowning infant out of the duck pond did no more than she ought to have done, but we nonetheless usually feel strongly inclined to reward her in some way—and not only as a gesture of thanks but also to celebrate the doing of a good deed. Similarly, we may wish to reward (perhaps only in some subtle way) the person who feels compassion at the plight of a homeless person, even though we think that that is the least one should feel. The point is that rewards (particularly congratulatory rewards) are not given solely for good works above and beyond one’s duty but can be given for doing or feeling what one ought, and remorse (like compassion) may be simply no more than that.

Nonetheless, there does seem to be something inadequate about the idea of ‘celebrating’ or ‘congratulating’ someone for feeling remorse. There is something not quite right in the ‘tone’ of celebration or congratulation here; these modes of response seem insensitive to the sombreness, the darkness even, of remorse, which after all is a response to wrongdoing, often enough very serious wrongdoing. Celebration and congratulation in this context would most often seem quite out of place. Remorse is not really an ‘achievement’, at least not the sort that invites celebration or congratulation in this way.

This inadequacy emerges even more clearly when we turn to consider the perspective of the remorseful person herself. The lucidly remorseful person may well have struggled her way to a clear perception of what she has done, but it is not a work that one completes with a sense of proud achievement or satisfaction at a job well done. The lucidly remorseful person may well look upon any such congratulatory prize in genuine puzzlement. ‘What is there to celebrate?’, we can imagine her responding. A reward that does not really ‘speak to’ the person rewarded is an odd sort of reward.

The reward rationale, then, also seems to under-describe or misdescribe the phenomenon it seeks to deal with. Someone originally inclined toward the congratulatory version of the reward rationale might, however, propose at this point that rewards are not always a matter of ‘celebrating’ worthy achievements — they can also be a form of recognising something as worthy or of value. This change of tack I would readily endorse, but with it we have really moved on to a distinct and, I would argue, deeper and stronger rationale for the remorse principle.

Remorse Should be Recognised by Mitigation

Mitigation of sentence in response to an offender’s remorse is best thought of, then, as an act of recognition or acknowledgment of the offender’s remorse or, better, of the offender in his remorse. This approach first involves viewing punishment as a communicative interaction, whereby the judge (representing the state or the community) reproaches or censures the offender through the imposition of sentence (See von Hirsch 1993; Duff 1986, 2001). In Duff’s view, the sentence is also offered to the offender as a possible penance to be undertaken to express her apology and help make amends. Against this sort of background (though we will see that Duff disagrees) a reduction of sentence can be seen as a recognition of the remorseful offender’s remorse that is communicated to her. Moreover, such recognition can re-figure the sentence’s communicative import, making it morally deeper and more attuned to the person to whom it is primarily directed.

How, though, more precisely, does mitigation of sentence ‘recognise’ or ‘acknowledge’ the offender’s remorse? And, moreover, what is the value in such recognition, especially such as to permit or indeed require a reduction of sentence?

Recognising a Person’s Remorse in Informal, Non-punitive Settings

With regard to the first question, let us first consider a non-criminal and informal scenario in which a person who has done wrong is facing another’s reproach and is feeling remorse. This is, let us say, a purely verbal, face-to-face communicative interaction with no punishment to follow the reproach. Let us accept that the person doing the reproaching is attentive and sees that the other person has heard and understood the reproach, is experiencing genuine remorse, and has expressed that remorse in what appears to be a sincere apology, a rejection of the past conduct and a promise to mend his ways. In such circumstances, other things being equal, the reproacher should properly adjust her speech to take account of—and so recognise—such a response from her interlocutor. Such communicative recognition of remorse may be made manifest in various ways—a lowering of the voice, a looking away slightly, a pause or hesitation in one’s speech, reducing the intensity of the terms used in reproach, and so on. In these ways one can acknowledge the wrongdoer’s acceptance of the moral reality of what he has done.

It is not simply that the reproach has become partly redundant because the wrongdoer’s attention is already focused on what he has done and so it is not necessary for the reproach to perform an attention-directing function. While that is true enough, it is not the key idea behind the modulation of the reproach. Rather, what happens now is better understood as a dialogical response by the reproacher to the wrongdoer’s remorse. That is to say, the attention of the reproacher is itself re-directed in turn, moving from the wrongdoing simpliciter to the wrongdoer’s remorse about his wrongdoing. Moreover, the purpose of the reproach changes from seeking to gain the wrongdoer’s attention and point out what he has done, to acknowledging his remorseful acknowledgment of what he has done.

Such adjustment or modulation in the reproacher’s speech need not, however, involve completely falling silent, or abandoning the conversation, or changing the topic. Nor, indeed, is such acknowledgement to be identified simply as forgiveness. Rather, it may still be appropriate to continue to explain to the remorseful wrongdoer what harms and hurt he caused. But now it should be re-pitched, done in a different tone, in revised terms, after a pause, and so on. It could be that the reproacher’s continuing reproach becomes a more inclusive recounting of the past, not so much as a straightforward rebuke or even a conveyance of hitherto unknown information, but more as a shared, dialogical reflection on and remembrance of the past. The reproacher who was not at all changed by the wrongdoer’s apology and repentance and who kept up the same rebuke regardless would be haranguing the wrongdoer, reproaching him unjustly—even if everything she said continued to be true.

A key aspect of that acceptance by the reproacher is a fundamental respect for the other’s remorse. Respect is one of the fundamental responsive attitudes that the phenomenon of remorse commands or elicits in others. This does not mean that the remorseful person may properly seek to use his remorse in order to induce some sort of subordination or awe in others, for he, too, is in an important sense humbled by or in his remorse. Remorse has a peculiar moral authority which cannot be cashed out as increasing the social power of the remorseful person. Nor, even though it is a virtue of sorts, is it a proper object of pride in the virtue-holder.

Mitigation of Sentence as Recognition of Offender Remorse

Recognising remorse through the softening of reproach may be all very well in informal interactions in the private sphere, but how does this fit with the formal sentencing by the state of a remorseful criminal offender? This situation is not purely verbal, so what, if anything, can modulation mean here? To begin with, the communicative modulations outlined above can and normally should be found in some form in how the judge delivers her sentencing remarks prior to making the formal sentencing orders. This situation is not so far removed from the informal sphere; it is (usually) still a matter of speaking to the offender, and many of the verbal modulations appropriate in the private sphere are still applicable here. This can involve, for one thing, expressly acknowledging the offender’s remorse and choosing one’s words in ways that reflect and recognise that fact.

The next step, of course, is to argue that communicative modulations can also be given effect in a reduction of the sentence severity. That is to say, a reduction of sentence severity can express and symbolise what is expressed and symbolised by modulations of purely verbal communication, namely a respectful acknowledgment of the other person’s experience of remorse. If punishment can emphatically express reproach, it should also be able to express, with its own kind of emphasis, modulations of such reproach.

Another way of putting this is in terms of the remorseful offender’s present culpability or blameworthiness. I would suggest that the present blameworthiness of an offender is not simply a matter of, say, the ‘seriousness of the harm multiplied by the criminal’s degree of responsibility when he committed the crime’. Rather it includes his present worthiness of being blamed (rebuked, censured, reproached) now for what he did then, and that can be modulated by many factors that lie outside the crime itself and relate to the personal circumstances of the offender.

Reduction of the severity of the sentence can be done by reducing the amount of the same kind of sentence (for example, three instead of five years in prison) or by changing to a less severe kind of sentence (for example, unpaid community service rather than prison). However, different kinds of sentence can also address an offender in different ways, some of which will be more relevant to remorse than others. This can mean that recognition of remorse need not only take the form of ‘a less severe sentence’ but may also or instead involve a more appropriate kind of sentence, one which, even if it is not obviously ‘less severe’, more adequately addresses and acknowledges the remorseful offender. For example, in some cases it may be appropriate to substitute a sentence of imprisonment with a period of unpaid community service as a form of reparation, even if the prison term and the amount of community service would be regarded in law as ‘equally severe’. Nonetheless, it would seem inappropriate to increase severity when modulating the kind of punishment to a more appropriate one.

What is the Value of Recognising an Offender’s Remorse?

But what is the value in such recognition of the remorseful offender? And, more pointedly, is that value enough to allow the otherwise just sentence to be reduced?

We place value on recognition as a fundamental form of paying moral respect to another person. Recognition is essentially one person’s acknowledgment and confirmation of another person’s self, especially her self-conception. It is a basic way of treating someone seriously as a rational, autonomous human being (and so a being with a self-conception) because it is through recognition that we accept that person’s self-conception as reasonable or in some sense legitimate or valid. The legal-political sense of recognition perhaps brings this out most clearly. When one state recognises another state, or the government of that state, it is recognising it as what that state or government claims to be (a sovereign independent political entity, or the legitimate sovereign authority within that state, as the case may be), and so confirms its self-presentation. It also undertakes to treat that state as what it presents itself as being. Recognition thus also shapes future conduct toward that state.

This moral valuation of recognition as a form of respect is distinct from but compatible with the metaphysically deeper analysis of recognition as something which is in fact a pre-condition of the very formation of a person’s (or a group’s) self-identity. Many philosophers, psychologists and social theorists since Hegel have pursued this theme (See, for example, Hegel 1977; Williams 1997; Taylor 1995; Honneth 1996). We need not enter into those debates here; it is sufficient for present purposes that recognition manifests a basic moral respect for another person. Still, it is clear enough that sufficient misrecognition and misconfirmation of another’s self-conception can do great psychological harm to him beyond the (bad enough) wrong of disrespect. As Charles Taylor puts it, ‘misrecognition shows not just a lack of due respect. It can inflict a grievous wound, saddling its victims with a crippling self-hatred. Due recognition is not just a courtesy we owe people. It is a vital human need’ (1995, p. 226; see also Laing 1969, chap. 7 on disconfirmation).

Of the many contexts in which a person may receive recognition or misrecognition, that of being subject to the criminal justice system and, in particular, being before a sentencing court is among the most serious and critical. To misrecognise an offender at this juncture can itself be a serious wrong to him. Where the offender is experiencing remorse, he is experiencing a significant shift in his self-perception and a re-orientation toward himself. This makes the situation of the remorseful offender a potentially quite critical one in terms of the re-formation of (at least aspects of) his self-conception. To fail to recognise the remorseful offender as a remorseful offender can thus amount to a basic misrecognition of him which disconfirms the value of his remorse, and so can constitute a significant wrong to him. To try to avoid that wrong of misrecognition, it will normally be the case that a reduction of sentence severity is needed.

Two Arguments Against Remorse as a Mitigating Factor

I turn now to consider two arguments against treating remorse as a mitigating factor in sentencing. As noted earlier, the purpose here is not to defend the remorse principle against all its critics, but rather simply to explore further the recognition rationale outlined above by seeing how it engages with two important lines of criticism. Consideration of these criticisms will help to refine and modulate the remorse principle and its rationale.

‘The Crime, not the Offender, is the Focus of Punishment’

This objection to the remorse principle is that when we punish a person, we punish her for her crime and not for the kind of person she is. This means that a person’s character is not relevant to sentencing and evidence one way or the other about aspects of her character should not be taken into account. Therefore, insofar as an offender’s remorse is regarded as part of her (good) moral character, it is not relevant, the argument goes, to determining the appropriate punishment. This approach is more likely to find support from the more retrospective approaches to punishment, where the focus is on retribution or censure for, or denunciation of, the crime. (In contrast, rehabilitationist, reformist, deterrence and incapacitation approaches, being more ‘forward-looking’, will generally be much more open to post-offence factors that may play a role in determining the causal effectiveness of the sentence.)

Duff argues against the remorse principle along this sort of line, from a censure approach to punishment (2001, pp. 120–121). The offender, he maintains, is to be censured for her crime, not for who or what she is. He focuses here on the question of proportionality. Because, Duff argues, an offender is to be censured for the crime committed and not for her character, the severity of the censuring punishment must be proportionate to the seriousness of the crime. Nothing the offender does or feels after the crime alters the nature and seriousness of the crime committed and, therefore, neither should it affect the severity of the punishment. (Duff does make an exception for ‘immediate repentance’, which, he says, can mitigate the seriousness of the crime itself (2001, p. 121).)

This objection raises the key issue of how we are to understand the communicative, censure-oriented approach to sentencing, which I have argued provides the basis for the better rationale for the remorse principle. Generally speaking, any communicative act will have four basic elements: the speaker, the interlocutor, the subject-matter and the speaker’s purpose in speaking to that interlocutor about that subject matter (cf. Aristotle 1991, p. 80). The nature of the communicative act performed in any particular case will be largely an effect of the combination of these four elements. The (perceived) nature or condition of the interlocutor is thus an essential element of the communicative act, and plays a large part in shaping the purpose of the speaker in performing that act. Variations in any one element, such as a change in or a new discovery about the circumstances or condition of the interlocutor, can produce significant variations in the communicative act as a whole. We saw this earlier in the case of the informal reproach of a remorseful person. Of course, this also means that formal versions of informal communicative acts will be different again, since the shift from informality to formality is essentially a matter of re-conceiving the natures of and the relationship between the speaker and the interlocutor. Almost inevitably, the formal versions of an informal communicative act will lose some of the subtleties and sensitivities that the informal versions possess.

Duff maintains that the formal communication of censure through state punishment is properly limited to being about the crime (as its subject matter), and should not be altered in response to changes in the interlocutor to whom the censure is directed. Duff agrees that a change in the offender (such as her coming to experience remorse) may well mean that in less formal contexts people (such as prison staff, probation officials, and so on) may properly speak differently to her, but he maintains that the formal censure (the sentence) should remain unaltered (2005 and 2007, pp. 383–387). He argues that formal recognition of repentance (equivalent to remorse here) requires the sort of ‘intrusive interest’ in an offender’s inner moral life that a liberal criminal law, respectful of citizens’ privacy, would balk at (2007, p. 385).

Must the necessary formality of state punishment be limited in this way? I would suggest that, while it is certainly true that formal sentencing cannot and should not seek to replicate all that is possible and desirable in informal settings, nonetheless it is sometimes possible and desirable that the sentencing judge, when censuring a particular offender, take into account not just the crime (as subject matter of the communicative act) but also the nature or condition of the interlocutor. This is particularly so when significant changes in or new discoveries about the nature or condition of an offender (as interlocutor) can alter the relationship between him and the judge (as speaker) and, in turn, alter the purpose and hence nature of the judge’s communicative act.

How significant does such a change or discovery need to be in order for it to be taken into account in formal punishment? There would, of course, be severe limitations to any general rule for when and how to register formally in state punishment what is properly registered in informal reproach. However, I would venture that such formal acknowledgement is likely to be appropriate when the particular change in or discovery about the offender is of direct relevance to the purpose and subject matter of the sentence itself. This is particularly so where some measure of agreement or commonality in the understandings of the speaker and the interlocutor has been reached.Footnote 3 Remorse, I would argue, is one such significant feature of the interlocutor; with it the offender and the judge have come much closer to a common understanding of the matter before them.Footnote 4 In this sort of situation, then, I would suggest that the formal punishment may—and normally should—take notice of what an informal reproach would properly note.

A censuring sentence is, then, not just about the crime committed in the past, but is also made to a particular human being at a particular stage in the ‘narrative unfolding’ of her life since the crime. Thus the full communicative significance of the punishment should not be identified solely with the ‘subject matter’ impliedly referred to by a censuring punishment (for example, ‘the grave wrong that you did’). It should also include the speaker’s understanding and recognition of the condition and circumstances of the interlocutor as these bear upon the matter at hand. And that, in turn, has its place in affecting the kind of censuring sentence that is passed.

‘The Truly Remorseful do not Ask for Mitigation’

This argument against treating remorse as a mitigating factor approaches the problem from the offender’s perspective. The genuinely remorseful person normally wishes to apologise, or undergo some sort of penance, or make reparation, or open himself to just criticism by others, and so on. In relevant circumstances, it can also be the case that the remorseful person will positively want to receive his just punishment. To be closed to any thought of apology, penance, reparation, reproach or punishment will usually be taken as a sign that what is felt is not genuine remorse.

Now, on the basis of this feature of the remorseful outlook, it may be argued that the truly remorseful criminal offender would accept his punishment and would not seek to reduce it on the basis of his remorse. Remorse involves humility and so, it may be said, it does not seek to use itself to avoid the full consequences of one’s wrongful actions. There is something self-contradictory about the remorseful person seeking to trade on his remorse in this way, as if, in vigorously pushing his remorse as a mitigating factor, he is thereby failing to be true to it. He is inappropriately putting himself forward, revealing that his remorse has skewed his attention from the victim to himself, that his remorse is really just moral self-indulgence which undermines his integrity. Even worse, it may be taken as evidence that he is not truly remorseful at all.

If, then, the truly remorseful offender would not ask for mitigation and, indeed, would ask that he receive the full measure of the appropriate sentence, then to mitigate sentence might be thought to frustrate him in his efforts to make adequate apology and do adequate penance. If, as Duff argues (2001, p. 111), the court’s sentence is the formal and officially prescribed way for the offender to make a public apology, then to reduce the sentence would be to reduce the capacity of the offender to make that apology adequately, thereby leaving him with a further burden of not having fully redeemed himself.

This general line of criticism is right to urge that the perspective of the remorseful offender is important. However, I would argue that we stop short if we leave things at this point, for there are two further thoughts that need to be pressed here. First, the remorseful offender need not be uncritically and passively accepting of whatever punishment is handed down. Secondly, the sentencer need not uncritically adopt the perspective of the remorseful offender and give him whatever he (the offender) thinks is his just punishment. On the contrary, the punisher, from her distinct perspective, may indeed provide that which the remorseful offender feels (possibly quite rightly) that he cannot and should not ask for himself. I shall explicate these two thoughts in turn.

It is true that the remorseful person typically does not seek to trade off her remorse as if it were some sort of asset in her possession which she may use as a bargaining chip. But neither is the remorseful person expected to be uncritically accepting of anything and everything said and done to her. She should not be seen, by others or herself, as entirely passive in relation to what others may say or do to her by way of reproach or punishment. That would be to make reproach and punishment merely one-way monologues. In contrast, there may be dialogue in punishment, too, though that need not be seen as a matter of negotiation between equals. The humility that is part of remorse conditions the dialogue but need not silence the offender’s voice. This is particularly clear when it comes to legal appeals, for even a remorseful offender may rightly appeal against a too harsh punishment. Just as she can resist or counter unfair verbal reproach, so too can she, in good faith and without betraying her remorse, appeal against an unjust sentence.

If so, may she then even pre-emptively raise relevant matters before the sentence is passed, at the sentencing hearing? It seems fair that she may. But clearly the way in which the claim is advanced can be tricky, for to push it too hard does indeed seem to risk being false to one’s remorse. But it seems wholly appropriate that, before the sentence is handed down, the offender be given the opportunity to express her remorse, in apologies, repenting of the crime, promises not to re-offend, offers of penance and reparation, and so on. And this, of course, will be to present her remorse to the sentencing judge (in an ideal system, at least). But this is not the same as presenting one’s remorse in order to ask for mitigation on its basis.

To be sure, though, this can be a difficult situation, and this can be one reason why it is desirable for criminal offenders to be represented by a lawyer at the sentencing hearing. The legal representative, by being a separate person and indeed an officer of the court and not the mere mouthpiece of the offender, is better placed to ask for mitigation on the basis of the offender’s remorse. Where the offender may find that her own remorseful humility pulls her back from referring to her remorse as a mitigating factor, her advocate is freer to put this argument on her behalf.

Turning now to the second further thought noted above, we may ask whether the sentencing judge is required by the remorseful offender’s humility and quiescence to ignore the fact of his remorse. Some may argue that if a sentencer is to respect the offender and his remorse, then she should not seek to do anything which the offender himself would not seek to do in the name of his remorse.

It is certainly important to respect the offender’s remorse, but that does not entail ignoring it, as if we should join in the remorseful person’s quiescence and not mention that which is the reason for his quiescence. The remorseful person’s perspective is not that of the sentencer and the sentencer is not obliged simply to represent the remorseful person’s perspective in her own deliberations. Again, to do so risks collapsing dialogue into monologue, for the punisher has her own distinct perspective, and it is one which allows her to take into account that which the offender does not seek to put into account. The offender cannot specify that he shall have a lenient sentence; neither can he specify that he shall have a harsher sentence.

But, the objection may be further pressed, the truly remorseful offender will still be frustrated if the sentencing judge limits his capacity for making public apology through undergoing his punishment, by making it seem as if he has less to apologise for, as if the crime were not as serious as it was. If the sentence is the officially prescribed form for what the community will accept as adequate apology, and the remorseful offender has no less need to apologise than the unremorseful one, then he deserves no mitigation of sentence (Duff 2001, pp. 111, 120; Duff 2005).

It is certainly true that a remorseful offender has no less to apologise for than the unremorseful offender guilty of essentially the same kind of offence. However, the sentence need not be merely the prescribed form of apology simply for that sort of crime. It can be the prescribed apology that is now sought from that particular offender for his crime. The court’s discounted sentence can reflect the fact that there is already apology at least latent in the offender’s remorse rather than the thought that he has less to apologise about. That is to say, it is more a matter of him having less apologising to do now, because he has already begun the work of apology in his remorse. (This, of course, depends on the nature of the remorse and what it has entailed in any particular case.) Again, the offender need not be seen as independently setting the agenda for his punishment or his apology. Rather, the court is effectively allowing that the already extant remorse serves as some measure of an apologetic attitude.

There is, however, the problem of how the court would know about the offender’s remorse if he does not seek to lead evidence of it. To this extent, then, the remorseful offender does retain some power here. He could carefully avoid giving any indication of his remorse, precisely so that he receives the full amount of punishment that his crime merits. This, however, would seem to be a masochistic distortion of remorse. If he was truly remorseful, then he should want to express it (through apology, reparation, etc.) even though he may not wish it to be subsequently treated as a mitigating factor. To block any apologetic or reparative action so as not to disclose his remorse and thereby not receive any dilution of his punishment would indeed be a distortion. This is not to say that traditional sentencing practices make it easy to express remorse prior to sentencing, such that the offender would have to work hard at keeping it hidden. On the contrary, the opportunity for the offender to speak at the sentencing hearing is usually limited to giving evidence to the court about his circumstances. It is not a chance to give a direct apology to the victim or other significant persons. (Telling the judge that he has apologised to the victim or will do so is not quite the same.) This is an area in which restorative justice practices involving victim-offender conferencing or mediation can play an important role.

Aggravation of Sentence for Lack of Remorse?

I have now concluded the analysis and defence of the preferred rationale for the remorse principle. Where, however, does this leave the unremorseful offender? In particular, does the above defence of the remorse principle entail that a lack of remorse should be met with an increased sentence? This issue will be more fully addressed in a future work but it may help clarify the nature of the remorse principle to outline briefly here—by way of conclusion—why the remorse principle need not imply that the absence of remorse should be an aggravating factor.

It may be thought that if the presence of remorse is accepted as mitigating sentence, then its absence must aggravate sentence, since, by definition, the unremorseful offender will receive a heavier sentence than the remorseful offender. However, it need not be the case that a heavier sentence is a positively aggravated one. Much depends here on the ‘baseline’ sentence or starting point.

I would suggest that the baseline sentence is that which would be imposed on an offender about whom the sentencing judge knows only the facts proven at trial (or admitted in a guilty plea) which grounded the finding of guilt. These facts are necessary and sufficient for a sentence to be non-arbitrarily imposed. Nonetheless, in most cases it will be highly desirable to supplement those basic facts with further facts, usually proved at a sentencing hearing, which may then alter the baseline sentence. Many of those supplementary facts will serve to mitigate sentence; others to aggravate it. Other facts may do neither but may nonetheless still help the judge to decide on the more appropriate type of sentence.

Now, in the great majority of cases, the basic facts that underlie the baseline sentence will not include any facts about whether or not the offender is remorseful, for the offender’s post-offence emotions will not be among the elements of the criminal offence of which the offender is guilty (unless the offence is a highly unusual one). This means that the baseline sentence is imposed upon an offender whose ‘remorse status’ is unknown, that is, we do not know whether the offender is remorseful or not. Any facts of the matter about the presence or absence of remorse will be supplementary to the basic facts.

The remorse principle holds that if remorse is proved as a supplementary fact, then this should normally lead to mitigation of the baseline sentence. If, on the other hand, an absence of remorse is proved as a supplementary fact, then there seem to be two basic options—both of which are compatible with the remorse principle. On the first option, a lack of remorse will normally serve to aggravate the baseline sentence. On the second option, in contrast, a lack of remorse will have—or, at least, need have—no effect on the baseline sentence. In other words, on this second approach, there need be no added punishment for lack of remorse.

As to which option is to be preferred, I would argue in favour of the second, as it makes fewer assumptions about the reasons why a person is not remorseful and leaves it open as to when and how a particular absence of remorse may reasonably affect the baseline sentence. The possible reasons for remorse’s absence can be quite varied, and it would not be wise to treat them as all morally the same. The wrongly convicted person, for example, who feels no remorse because she did not in fact commit the crime, is clearly different from the defiant gloater who is proud of his deed; and the deluded offender who sincerely but falsely believes he is innocent is morally distinct from the psychopath who is simply incapable of feeling remorse. In the face of this variety we should not adopt a sentencing principle that fails to discriminate between these (and other) diverse cases.

In any case, regardless of how the choice of options is ultimately to be settled, the main point to note for present purposes is that the remorse principle is compatible with both options. That is, we can affirm that remorse should normally mitigate the baseline sentence without being committed to the view that a lack of remorse should normally aggravate it.