Time for radical action on our failing prisons

•Letter to The Guardian, 17 July 2015

Prisons can never provide “purposeful activity” for so many prisoners; the only way is to reduce their numbers. Here are two ways. A high proportion of people sent to prison lack skills; need help with reading, writing or numeracy; have mental health problems; are homeless – the list goes on. To enable them to live productive lives they need appropriate education, training, therapy and so on. But these are more difficult and expensive to provide in prison. The answer is to provide them outside. Where facilities exist in the community, offenders could be required to attend them; elsewhere, courses could be arranged in probation-run day centres, such as existed a few decades ago. These could also arrange restorative justice meetings for victims who want them. Michael Gove, the justice secretary, could pay for them with the money saved by cancelling the secure college and, let’s hope, the mammoth prison at Wrexham.

Second, sentencing requires a complete rethink. Sentences are based on an attempt to quantify the seriousness of the crime, with no regard to the best way to prevent reoffending. The tariff is arbitrary: no sentence can be justified except by comparison with other sentences, and if they were reduced by, say, a sixth they would still be proportionate to each other. Finland progressively reduced sentence lengths over about four decades without affecting the crime rate. We could do the same.
Martin Wright
London       [More letters at http://www.gu.com/letters ]

Sentencing: a return to first principles

If we go back to first principles and ask ‘What is the criminal justice system for?’ we often think, first of all, that it is to protect us from crime. But that thought needs to be qualified.

 Firstly, the most effective measures for crime reduction are not part of criminal justice, but deal with social pressures towards crime. The proportion of prisoners who had been taken into care as a child, truanted from school, were homeless, had run away from home, had below Level 1 numeracy or reading ability, or used drugs, is from two to thirty times the figure for the general population. These factors are nothing to do with sentencing, except where the child’s situation has been aggravated by sending a parent to prison.


Seven out of ten prisoners, both male and female, suffer from two or more mental disorders1. It therefore seems obvious that prison overcrowding should be dealt with by creating more mental health facilities, not necessarily custodial, rather than prisons.

Crime reduction strategy should clearly focus on factors like these, rather than by increasing sentences, which has little effect. A healthy society is not one where people are only kept in order by fear.

When it comes to preventing re-offending, the most important factor is what happens after the sentence: do they have a home, a job, and other forms of support. These too are matters for the community in general, not the criminal justice system.

Secondly, criminal justice should be about helping the victim to recover, as far as that is possible. Many victims are not so intent on seeing the offender punished; often their main concern is that they should co-operate with programmes to change their life, and stop offending. This makes more demands on the offender than sitting in a cell.

Thirdly, we really need to sort out the philosophy of sentencing, especially with regard to prison. Sentencing is currently trying to do several things at once, conveniently ignoring that these largely contradict each other, and could be achieved in other ways. A review of sentencing is welcome, but it needs to be radical. Some people simply assume that it is to punish offenders. Others say that there should be education and training in prison. If you point out that few people receive it, and it’s often outweighed by the damaging effects of prison, it is claimed that it contains people for a period. But they are more likely to re-offend afterwards.

Every time a person is sent to prison, he is more likely to re-offend on release. In one sample, 25 per cent of those with no previous custodial sentence re-offended. After one previous custodial sentence, this jumped to 40 per cent, and so on until 76 per cent of those with 11 or more prison sentences offend again2. So prison seems a short-term solution, just as a credit card is a short-term solution to debt – but it catches up with you, with interest.

It comes down to this: the primary purpose, which trumps all the others, is an attempt to show everyone just how serious the crime was, and to measure this in units of time. It is legitimate for the state to show everyone the boundaries of permissible behaviour, but does it make sense to use this method? It stigmatizes people so that it is harder for them to find a job afterwards; it separates them from people who might be a good influence; and herds them together with others like themselves, with too little to do, often introducing them to drugs, with a high incidence of bullying, self-harm, and too many suicides (often by young people).


Furthermore, the baseline for sentences is quite arbitrary. No sentence is justifiable except by comparison with others. Suppose there are three offences, each more serious than the last: the offenders may be given sentences of, say, 4, 6 and 8 years. But it might just as well have been 3. 5 and 7, or even less, and the point would still have been made. This would reduce the prison population, so that there was a better chance of constructive regimes for prisoners, and there would be no need for the executive to intervene by releasing prisoners early.

Custody should be reserved for the most serious cases of danger to the public, and for enforcing reparation when necessary. The same dilemma about fixing an arbitrary length would remain, but there could be much less custody, and when it was used it would be for a constructive purpose.

Given that a substantial number of offenders are on the borderline between custody and a community measure, and that community measures are generally less costly and no less effective, cuts across the board make no sense. If the probation service can show the need for a particular programme in their area, and persuade the courts to use it when appropriate, part of the resulting saving when the prison population is reduced should be transferred to fund the programme3.

Could we not find a more constructive way of marking society’s displeasure? It sounds very logical to say ‘He has broken the law and should be punished. But surely it is just as logical to say ‘He has caused harm and should make up for it.’ This includes the victim. It enables the offender to make reparation, in some cases by meeting the victim (if he or she wants it) to answer questions and hear at first hand how the victim was affected – a difficult thing to do. In other words, restorative justice, which has been shown to reduce reoffending . It gives the offender a chance to make amends voluntarily, and it means that he or she can return to society with some positive action to weigh against the offence – a new use for the scales of justice.


1 Social Exclusion Unit Report Reducing re-offending by ex-prisoners, 2002, quoted by Prison Reform Trust, Prison factfile, 2006.

2 Of adults sentenced in the first quarter of 2007: Maria Eagle, Parliamentary Answer to Keith Hill MP, 29 October 2009.

3 Martin Wright, Towards a restorative society: a problem-solving response to crime. Make Justice Work, 2010, p. 14.

Is it time to question the concept of punishment ?

Paper to Fifth International Conference, “Positioning restorative justice”, Katholieke Universiteit Leuven, 16-19 September 2001.

This paper will briefly survey the historical use of punishments, and will try to clarify some of the terminological confusion that surrounds the debate. It will summarize some pragmatic and ethical objections to punishment, as reflected in current sentencing practice, and will suggest how a society without punishment could minimize harm and respond restoratively to it when it occurs.

The catalogue of punishments inflicted on human beings, usually by men, is a long and shameful one. A list of measures imposed over centuries and millennia shows that they are often far more barbarically inventive than anything perpetrated by the malefactors themselves. It includes, for example, outlawry, transportation, branding, mutilation by cutting off the tongue or hands, the scold’s bridle, burning, boiling in oil, flogging, and death. All these have been abolished in the civilized world (some not until the mid-twentieth century), but solitary confinement is still widely practised. It is often hard to see where lies the borderline, if there is one, between legally permitted punishments and torture. Reviewing the killing, torture, mutilation, exile, imprisonment, separation of families and general misery inflicted in the name of law enforcement, the Dutch criminologist Louk Hulsman has said that the three greatest causes of human misery throughout the ages have been famine and pestilence, war, and the criminal justice system. Restorative justice is an attempt to find a better way.


Making pain humane

As people began to be squeamish about the grosser forms of physical punishment, the response was not to question the use of punishment but to find a punishment which was doubly invisible, firstly because it was psychological and left no marks, and secondly because it took place in a remote penal colony or behind high walls. The punitive ethic tends to lead to abuses, which are hard to detect because offenders are out of public view. Prison is punishment without bloodletting, but it can be psychological torment imposing silence, degrading uniforms, solitary confinement, hard work specially designed to be totally useless, obedience to silly rules, and sensory deprivation in ‘supermax’ security prisons; and it takes place out of public view. All too often it incapacitates people from being able to earn an honest living, or makes them too angry to want to do so, or drives them to insanity, self-harm or suicide. As the futility of all this dawned on the Victorians and their successors, prisons were given fig-leaves of respectability: the silence rule went, prisoners were allowed to do useful if uninspiring work (such as sewing mailbags and weaving doormats) and given wages for it (usually mere pocket money). Prison regimes began to include rehabilitation, with some education and training for some prisoners, but generally only for a few hours a week.

Prisons are inherently costly (in financial as well as human terms). They require buildings and staff, and prisoners have to be fed; this means that even in rich countries they seldom provide enough work or education, and in less ‘developed’ countries which have imported this institution but cannot afford to maintain it, gross overcrowding is common. Imprisonment is sometimes described as ‘incapacitation’: the word is ambiguous, since it means that the incarcerated person is rendered incapable not only of committing further crimes (except those which can be organized from within prison) but also, all too often, of living a normal working life after release. Another attempt to mitigate punishment is the indeterminate sentence, in which the offender is supposed to be released as soon as he or she is rehabilitated, but this carries the danger that non-conformists and those who protest their innocence may be detained for a period quite disproportionate to their offence.

Just as prison is ambivalent, trying to show a humane side to liberal public opinion while appeasing hard-liners, so is rehabilitation: to satisfy demands for severity, its advocates claim that it includes some punishment because ‘It’s very tough really’, or that ‘anger management courses are very demanding’. It may also be unintentionally punitive, for example through deprivation of liberty in an institution with a euphemistic name: approved school, community home with education, correctional institution. The punitive overtones mean that the authorities are not very receptive to complaints that the ‘medicine’ tastes nasty; while the claimed rehabilitative intentions can mean that fewer safeguards are in place than if the institution were overtly punitive. For example, ‘approved schools’ and ‘community homes with education’ had even less outside supervision than prisons. In closed institutions, security can be made the excuse for restrictions on contact with the outside world, whether from families and friends or from independent observers. Already in 1777 John Howard noted that when a gaoler advised him not to go to one part of the prison because of the risk of gaol fever, he always found squalid conditions, but no fever. According to Garland (2000), however, recent developments have eroded the ‘ “civilised” attitude towards crime of the educated middle classes’. They may feel that they have lost some of their status as experts who determine policy, since policy-making has become more politicized. At the same time they themselves may have been more directly affected by crime, as well as being alarmed by urban riots and crimes of frightening violence which receive much publicity. As a result, Garland argues, the professional middle classes have become ‘less supportive of penal-welfarism and more supportive of punitive responses to crime’ (p. 357). Be that as it may, mass-circulation media, either influencing public opinion or reflecting it, have taken a generally punitive stance, and politicians have felt that tough rhetoric would be popular. This means that their agenda is no longer the same as that of the professional groups who want to produce tangible results (Garland, 2000: 350-1, 368-9). Examples from England are the introduction of minimum sentences for certain repeated offences, in the Crime (Sentences) Act 1997, and the re-naming of community service orders as community punishment sentences, in the Criminal Justice and Court Services Act 2000. These moves fail to take account of evidence that the public is not as punitive as is often assumed, as will be discussed below.

But the more punitive, the less rehabilitative, and vice versa. In other words, the more deliberately painful a measure is, the less likely it is to persuade and enable the offender to change her offending behaviour; and a measure which attempts to help will be less painful, even though it involves an element of coercion which many offenders will find unpleasant. Also, the more punitive, the harder to uncover the truth. The ‘wall of silence’ met by investigators in the inquiry into the Hatfield rail crash is one example (Independent, 23 January 2001; others are given by Wright, 1999: 41-2). Since the result of conviction is pain, it is often thought necessary to inflict even greater pain in order to obtain a confession: thumbscrew, peine forte et dure (crushing with a heavy weight), noise, electric shocks, threats to family, and so on.


Terminological confusion

All punishment, in the normal sense of the word, is intended to cause pain and fear, and in some cases incapacitation. Some writers use the word ‘punishment’ as if it were synonymous with ‘sentencing’ (Daly, 2000); others, correctly in my view, restrict its use to sanctions intended to inflict pain (Christie, 1982).

We need to distinguish, firstly what it is (its definition, and its description, which may not always tally with the definition), and secondly what we call it (its name). The intention of the person causing the injury or imposing the sanction does matter. The infliction of pain is not always punitive. The sufferer may endure pain voluntarily, as in the case of undergoing surgery, taking part in athletics – or writing an article! Even without consent, the way the same sanction is presented can make a difference. An offender will respond differently if he has to work alone, or with other offenders, or with staff or volunteers working alongside him, or if he is told ‘Scrub that floor and don’t let me catch you talking to the patients’, or ‘The hospital is short-staffed, so the patients would really appreciate it if as part of your community service you scrubbed the floor, and you are welcome to have a cup of tea with them afterwards’. In the former, the offender will feel that the intention is primarily to make her suffer; in the latter, the unpleasantness of coercion is still present, but it is likely to be moderated, and possibly outweighed, by a feeling that she is being treated with respect and doing work which is appreciated. Preferably, of course, the work will be of a kind which uses more of the offender’s talents than scrubbing floors or picking up litter.

‘Punishment’ may be used to mean any of the following:

1. a measure primarily intended to hurt, described by Christie (1982) as ‘pain delivery’.

2. a measure primarily intended to rehabilitate or contain. It may often also be painful, but this is not the primary aim. It is sometimes said that rehabilitative measures are in fact punitive, because sometimes they take place in penal institutions and may involve indeterminate sentences which are out of proportion to the offence. This objection is, however, as illogical as it would be to say that imprisonment is rehabilitative because some prisons offer rehabilitative programmes to some of their prisoners.

3. a measure primarily intended as reparation. Whether it is painful or enjoyable is irrelevant. There is much anecdotal evidence of offenders enjoying their community service; a small number even continue it voluntarily because they felt appreciated and their self-esteem was enhanced; this is probably valuable, but to call it punishment seems a clear misnomer.

4. a general term for any kind of sentence. This seems unhelpful because it uses a word with a punitive root to refer to measures which, as we have seen, are not necessarily punitive. Daly defines punishment as anything that is unpleasant, a burden, or an imposition of some sort on an offender (2000: 39). This may well be a widely held concept; one informed lay commentator, for example, expresses what is probably a common assumption when she writes that ‘the guilty deserve to suffer’ but also that punishment ‘can properly take the form of the offender making some reparation to his victim, and/or the community, in order to help restore his relations with them’ (Windsor, 1990: 6, 9).

If all these are described by the same term, confusion is inevitable. ‘Punishment’ is irredeemably ambiguous. It would make for greater clarity in academic debate if we avoided the term because it is not clear what the writer means by it, and used a word such as ‘sanction’ as the general term for any imposed measure, with adjectives to distinguish the various kinds according to their primary purpose; the existence of subsidiary aims or unintended side-effects (when the description of what it is really like does not tally with the definition or the name) could of course be acknowledged. It is true that any measure imposed by authority is backed up by coercion which some will find irksome: it may be more unpleasant than intended, and safeguards are therefore needed. But if the name expresses the intention, it will make clear to those who implement sanctions that this is the ideal to be aimed for. There is no objection to the task itself being burdensome and requiring effort on the offender’s part, but the pain of coercion will be minimized, and the ultimate aim will be voluntary compliance. We could speak, instead, of:

  • ‘punitive sanctions’, to describe measures which intend to hurt, and fail if they don’t. They may be symbolic, when the criterion for ‘success’ is pain or denunciation, or instrumental, where the usual measure is a reduction in re-offending.
  • ‘rehabilitative sanctions’, which are intended to help; if they cause pain it is unintentional, and any loss of liberty should be minimized; safeguards are needed.
  • those which are ambivalent about their aims can be described, for example, as ‘punitive/rehabilitative sanctions’
  • ‘reparative sanctions’, intended to help the victim, and often the offender also. They may involve loss of liberty or money, but this should be by consent if possible (again, safeguards are needed). Since pain is not the aim, they can succeed even when they are enjoyed and continued voluntarily.
  • ‘containment’, intended to protect the public. This may include restriction of liberty, such as disqualification from running a company or driving a car, or deprivation of liberty, i.e. detention. Institutional regimes should as far as possible be reparative, not punitive, and should not ‘incapacitate’ people for normal lives after release.

In the remainder of this paper, ‘punishment’ will be used only where the authors quoted have used it. The word ‘sanction’ implies a measure imposed on a wrongdoer by an individual (e.g. a parent or a gang leader) or an authority (e.g. a school or a court). There are other possibilities, as we shall see, such as voluntary action by the wrongdoer, and ‘natural consequences’ of the deed (Wright, 1982), of which more below.


The desirability of punitive sanctions

This paper will consider two recent writings on punishment: a review article by Duff (1996), and a paper by Daly (2000), drawing on Duff’s work. It will argue that it is impossible to impose punitive sanctions consistently and proportionately, that they are not the only or the most effective way of achieving their aims, and that they have unacceptable side-effects; it is therefore questionable whether an ethical policy for punitive sanctions is possible. Restorative justice should therefore be developed as an alternative.

Duff (1996) is one of those (like Sir Walter Moberly, 1968) who lists many cogent objections to punishment but cannot quite follow them to the logical conclusion and reject it as a basic principle. He rightly points out, for example, that ‘To justify a system of punishment, we must … show not only that it does good (and does more good than harm) but also that no available alternative practice could be expected to bring about as much or more good, at lower (or no higher) cost’ (Duff, 1996: 4-5). At times it almost seems as if he, like Daly, is using ‘punishment’ to refer to any imposed sanction, but he makes clear that he includes ‘the delivery of pain’ (p. 89).

Duff puts forward two main arguments: that punishment is ‘communicative’, and that it encourages repentance and penance, which are conducive to possible reconciliation and forgiveness. We communicate how serious the crime is by how heavily we punish the offender, according to Duff (1996: 56). He concedes that for the state to make detailed inquiries into each offender’s situation is intrusive and expensive, and questions why it should be ‘the state’s task to administer censure through a formal system of punishment: why not just leave it to other individuals (most obviously the victim or her friends) to censure the wrongdoer?’ (pp. 60, 34, his italics). He also admits that setting the absolute levels of punishment is another problem: ‘How long is “lengthy”?’ (p.66).

According to him the response to wrongdoing should be ‘communicative’, involving two-way communication, and forward-looking, aiming to induce repentance and reparation, which ideally the offender imposes on him or herself. Advocates of RJ would certainly agree, because as Daly (2000: 42) says, it sounds very much like the conferencing process; but surely such processes should be distinguished from punishment, not subsumed within that term? As regards communication, punitive sanctions are certainly ‘expressive’ of indignation or condemnation (Duff, 1996: 32); but apart from that, as any child who has been hit on the knuckles with a ruler knows, they are likely to block communication, by making the offender too afraid or too resentful to speak or to listen. Duff says that punishment is to bring offenders to realize the social desirability of obedience to law. The aim is unexceptionable, but is this an effective or justifiable way of trying to achieve it?

The question then is, How best to induce repentance, or metanoia (change of mind)? Here Duff (and Daly, as we shall see) find themselves trapped in the punitive tradition: having said that they would like offenders to undertake penitence or reparation themselves, they propose to impose it as punishment. The trouble with this is that repentance, if it is to mean anything, has to come from within, inspired by empathy for the other person. But punitive sanctions are more likely to produce resistance and resentment, to inhibit learning rather than promote it, and try to avoid the pain. Von Hirsch questions whether the state should ‘use its coercive powers to seek to induce moral sentiments of repentance’ (quoted by Daly, 2000: 47); he might also have asked whether, if it does so, the state is any more likely to succeed than Glendower in Shakespeare’s King Henry IV Part I:

             Glendower: I can call spirits from the vasty deep

Hotspur: Why, so can I, or so can any man;

                              But will they come when you do call for them?

Just as the offender’s repentance cannot be coerced, neither can the victim’s trust that it is freely expressed (Davis, quoted by Daly, 2000: 48).

Duff’s aim is ‘to induce the pain of condemnation and of recognised guilt’ (quoted by Daly, 2000: 42); but in restorative justice, the aim is not the pain for its own sake, but the empathy with the person harmed. It can indeed be painful to feel empathy for someone whom one has hurt; but this is different from the pain inflicted by an authority; when it comes from within, it encourages the offender to relieve it through words or actions, and this in turn can make the victim feel less angry, and more likely to accept an apology or other reparation, sometimes enabling them to forgive.

Duff asks whether censure requires hard treatment: could not the communication be purely symbolic? If its aim is to acknowledge the damage to the victim, this can also sometimes be achieved by apology and restoration or reparation … or by informal community action. Why punishment and not victim-oriented compensation? (Duff, 1996: 34, 39, 80). This is indeed the restorativist case, and it is not clear why Duff suggests that the infliction of pain would assist the process, especially as he quotes Von Hirsch as saying that punishment itself does not profess to deter ‘the most recalcitrant’ but to give ‘ordinary persons good reasons for compliance’, which is just what the restorative process claims to do. He says that it is because the victim has been wronged, and is owed at least an apology, which ‘must involve a repentant recognition of wrongdoing’; ‘communicative punishments’ aim to ‘bring the offender to a repentant understanding of his wrongdoing’ (p. 80-1). He admits that ‘punishment can of course all too often be, in intention or in fact, mere “pain delivery” and that ‘similarly, blame or censure … can too often be merely exclusionary or dismissive’. Duff claims that ‘neither blame nor punishment need be like this’ (p. 82), but the contention of this paper is that the use of the term ‘punishment’ makes this more likely by appearing to legitimate it. If sanctions are clearly labelled according to their intention, as restorative, rehabilitative or indeed communicative, it will be clearer that pain as pain is not acceptable, and pain as side-effect should be minimized.

This is the crux: the restorative case is that the infliction of pain because it is painful is likely to be counterproductive; if the aim is to increase awareness, understanding and empathy, the measures should be designed for that purpose. If they succeed they may prove painful, but the pain will come from within the offender, and that is what is needed in order to generate repentance and the willingness to apologize and make reparation. None of this can be guaranteed, of course, and more thought needs to be given to procedures which are to some degree restorative and can be followed when the full process cannot be used or is not complied with.

Turning now to Kathleen Daly’s recent discussion of punishment and restorative justice, in Heather Strang and John Braithwaite’s collection Restorative justice: philosophy to practice (2000). If I may crudely summarize Daly’s argument, she seems to be saying that:

  • ‘We should embrace (not eliminate) the concept of “punishment” as the main activity of the state’s response to crime’ (p. 34)
  • restorative justice is punishment
  • That’s as it should be

Of course she expresses it much more elegantly than that, in a series of clearly argued propositions. She invites discussion and debate, and this paper is a response to that invitation. She accepts that there are basic differences between traditional and restorative justice, singling out four: victims as peripheral or central, focus on punishing /treating the offender or on repairing the harm, community represented by the state or participating more actively, and an adversarial process or a dialogue (p. 36). But she suggests that the contrast between retributive and restorative justice, with restorative justice as a ‘third way’, while it helped to make the point in the 1970s, is no longer helpful. She suggests that restorative justice processes should be seen not as alternatives to punishment, but as alternative punishments.

So, firstly, why does Daly consider punitive sanctions desirable? One reason is that they are necessary in order to vindicate the victim’s worth (p. 39), and that the offender can atone only by willingly submitting to punishment (Garvey, quoted by Daly, p. 50 n.10). But Daly herself recognizes the value of the restorative justice process (p. 46), which seems to vindicate the victim’s worth much more clearly than inflicting pain on the offender. She is also able to claim that punishment is desirable because she is using the definition of it which includes non-punitive measures; she does have a place for painful measures, but stresses that in her definition punishment doesn’t have to be humiliating, harming or degrading, and need not involve prison – rather like the housemaid who told her mistress that she had an illegitimate baby, but pleaded in mitigation that ‘It’s only a little one, Ma’am’.

Secondly, Daly says that restorative justice is punishment because it ‘combines retribution and rehabilitation’ (p. 35), but there are two problems with this. One is that, if by retribution she means deliberate infliction of pain this is incompatible with restorative justice as most of its advocates understand it; the second is that we cannot ‘serve two masters’: one of the two contradictory aims will have to take precedence. It is true that restorative actions require effort, psychological or physical, and may be ‘burdensome’; but that is not the same as the infliction of pain for its own sake.

Hence, Daly asserts that restorative justice is punishment (in her extended use of the word). She says that it combines retributive, rehabilitative and restorative principles (p. 45). She is referring to the use of conferencing, of which she approves (p. 45-6). We need to separate the processes. If the court orders that a case be referred to a conference, I think it would be clearer to describe that as a ‘restorative sanction’; if the case is routinely referred to a conference, that is not a sanction, but part of the process. Furthermore, the outcome will then normally be agreed by those concerned, so it seems inappropriate to call it a sanction, let alone a punishment.

Daly’s third point is that restorative justice should be punishment. But just as rehabilitation and deterrence are contradictory aims (Wright, 1999 ch. 6), so are reparation and proportionality; the idea of restorative justice is that any reparative acts by the offender are if possible agreed by the victim and the offender. They therefore are not necessarily proportionate to the seriousness if the victim does not feel this to be necessary; it may be thought that there should be a safeguard against demands by victims for excessive reparation, although this does not appear to have been a problem in practice. Restorative justice practices, according to Daly, ‘are concerned with sanctions or outcomes that are proportionate and that also “make things right” ’(p. 35, her italics). But these aims are also in conflict, and certainly the sanction should not be increased in the name of proportionality, as the Clotworthy case described in the same volume shows (Morris and Young, 2000: 11-13; see also Mason, 2000: 4-6). After a very serious offence in which the offender, Clotworthy, demanded money at knifepoint and stabbed the victim six times, he explained to the victim that he had been drunk but had not touched alcohol since the offence. The victim told the offender that he himself had been to prison, which ‘did not do him much good’, and they agreed that instead the offender would pay a large sum towards the cost of cosmetic surgery; but the Crown appealed, and the Court of Appeal ruled that this was ‘inadequate’, and sent the offender to prison for three years, substantially reducing the amount he was able to pay. Thus a punitive sanction replaced a restorative one, and went clearly against the victim’s wishes.

The Court of Appeal stated that it was not opposed to the idea of restorative justice, but that in cases of serious violence it must be ‘balanced against other sentencing policies’, i.e. outweighed by retributive ones. This is an example of ‘authoritarian’ restorative justice (Wright, 2000: 21-23), and could be described as punitive. Similarly, it is reported that in some cases family group conferences in New Zealand include a punitive element, even including imprisonment (McElrea, 1995); to that extent, I would argue, they too are not restorative.

Daly makes the further point that ‘commonsense understandings of a just response to crime’ include ‘punishing wrongdoers’, and that restorative justice advocates would be wise to work with them in building interest in the idea, because to excise the idea of punishment ‘may not be strategic politically’ (p. 45). This is a purely pragmatic argument, not a principled one, and in any case it ignores the fact that ‘commonsense’ is also attracted to the idea that the offender should make amends in a constructive way, as we shall see in a moment.


What’s wrong with punitive sanctions?

In this section we will consider the argument that punitive public opinion requires sanctions to be severe; other conventional justifications for punitive sanctions; and the claim that proportionality and fairness in sentencing are achievable.

Public opinion

First, it is necessary to question the common assumption that public opinion insists on punitive sanctions. Several opinion polls in various countries show that ideas such as reparation to the victim or the community are very popular. To take some British examples: a survey of 988 people by National Opinion Polls found that 85 per cent thought it was a good idea to make some offenders do community service instead of going to prison, and 66 per cent wanted to make them pay compensation to their victim. Another large survey found that 93 per cent thought that offenders should have to ‘make good the consequences of their crime wherever possible’(cited by Wright, 1989). The British Crime Survey found that a minority of victims wanted ‘their’ offenders imprisoned: for ‘mugging’, 31% preferred prison, as against 62% who preferred a non-custodial sanction; 43 per cent would have accepted an opportunity for mediation. For burglary with entry : 34 per cent wanted prison, 48 per cent a non-custodial measure, and 47 per cent would have accepted mediation (Mattinson and Mirrlees-Black, 2000: 69, 70, 42).

It is true however that there is a culture of violence in some parts of British society, perhaps reinforced by, for example, children’s comics such as The Beano, in which children’s escapades routinely end with physical punishment by an adult (see for example Beano 1980 November 15, p. 5; 1981 December 12, p. 18). Several European countries have prohibited physical punishment by parents, since Sweden intorduced the reform in 1979 (EPOCH Worldwide, 1996); in Scotland there is a current proposal to prohibit it up to the age of three, but in England even that small step has aroused strong opposition to such infringement of the parents’ ‘rights’ and the government in 2001 refused to introduce a ban (Independent, 8 November 2001); children’s rights are being ignored by the punitive lobby.

It has to be conceded, also, that through the millennia there has been a persistent tendency of humans to impose punitive sanctions on each other: either inflicting punitive sanctions alone, or combining them with reparation (but in some cases allowing them to be replaced by compensation). In the Judaeo-Christian tradition, there are widely quoted precepts about ‘an eye for an eye and a tooth for a tooth’, and so on, but it is generally accepted that these injunctions are setting limits to the amount of compensation that can be demanded. The ideal of peace and harmony (shalom) is expressed more strongly (Zehr, 1995; Hoyles, 1986). In the New Testament a non-violent doctrine was taken further. To many, ‘Turn the other cheek’ seems at first too idealistic; but when one considers the effects of blood feuds and international tit-for-tat retaliation, it appears obviously practical. Similarly ‘Do not return evil for evil, but drive out evil with good’ is very much in harmony with the principles advocated in this paper.

Hindu teaching also starts from a punitive standpoint, with many punishments from reprimands, fines and shaming to mutilation and death (Tähtinen, 1982: 14-15). But here too there are limitations. There is an element of ‘virtue is its own reward’ and the converse: those whose speech and mind are pure get dharma (socio-ethical good), artha (economic good) and kâma (psycho-hedonistic good), while sin results in karmic social consequences and punishment (pp. 9, 16). There is a need for penance and expiation (prâyaúcitta), but this is seen as replacing or at least mitigating punishment, not as being induced by it:

 It seems that expiation soothes and palliates the retributive punitive feelings in society and modifies the seriousness of punishment. … To a great extent the desire of ‘hitting back’ at the offender has been replaced by the rite of expiation (p. 39, italics added).

If a learned person breaks the law, punishment should not be excessive; Tähtinen notes drily that the rules were mostly written by Brahmins (p. 44).

This tradition has also been developed in a spirit of non-violence, notably by Mahatma Gandhi. He stresses the significance of harm rather than law-breaking: ‘Crime which really matters is a serious moral crime no matter whether it is in harmony with the existing laws or against them’, and the worst is structural violence such as corrupt bureaucracy [or, he might have added, exploitative industry and commerce]. He even questions the state’s right to impose rehabilitation: ‘Has society a moral right to rehabilitate an individual to adjust to an unjust society?’ Punishment aggravates criminal tendencies: ‘Moral crime has to be eradicated by moral means’, and Gandhi reiterates the importance of expiation (Tähtinen, 1982: 47-48).

There are also parallels with Islam. The Quran is explicit about cutting off hands of thieves (e.g. V: 39), but at least one Islamic scholar has argued, with examples from other Quranic texts, that this should not be taken literally (just as, he might have added, Christians interpret the injunction ‘If thy right hand offend thee, cut it off, and cast it from thee’ (Matthew 5: 29) as hyperbole). Certainly the Quran also teaches repentance, mercy and forgiveness (see also Wright, 1999: 167-8).


Conventional justifications for punitive sanctions

The main justifications for punitive sanctions may be grouped under the headings utilitarian, symbolic, or both.

1. Consequentialist, utilitarian This is the basic response, called ‘behaviourist’ by psychologists: if you do that, I will do something that hurts you. The problem with this is that it only works under certain circumstances. When the pain is threatened in advance, the wrongdoer must believe him or herself likely to be detected, and must believe that the pain inflicted will outweigh the ‘reward’ brought by the deed. When it is inflicted after an act, with the intention of discouraging a repetition, it must be prompt, and not so excessive as to engender resentment and anger. These circumstances are not widely found in criminal justice systems.

Even if it works, it does so for the wrong reason: fear. ‘If the law and the authorities win [by using “legitimate” violence], this is a bad victory, because it is achieved by the very means which are pronounced illegitimate then the other side – the criminals – is using it’ (Fatic’, 1995: 201). Or as Marshall Rosenberg, the psychologist and advocate of ‘non-violent communication’ (NVC) has put it, the first reason for punishment is that we want someone to act differently, and for this it sometimes works, but sometimes instead provokes resistance; we should also ask ourselves, however, ‘What do I want this person’s reasons to be for doing what I’m asking?’. We do not want them to act from fear, guilt, shame, or to buy love, but willingly, so as to enrich life for themselves and others. ‘NVC … fosters a level of moral development based on autonomy and independence, whereby we acknowledge responsibility for our own actions and are aware that our own well-being and that of others are one and the same’ (Rosenberg, 1998, 1999). Duff (1996: 5, his italics) similarly queries ‘Is the goal simply that fewer crimes are committed, or does it matter why … : whether for instance people obey the law only through fear … .’

The side-effects of punitive sanctions are well known, though often ignored in theoretical debates. They will be considered below. Also widely ignored is the ethical aspect. Punitive sanctions consist of the infliction of pain by the state on a citizen; if this can be justified at all, it can surely only be, firstly, when it can be shown to achieve a desired (and acceptable) effect, and secondly, when this effect cannot be achieved in another way that does not cause damage.

It is remarkable that the utilitarian argument seldom stops to ask ‘useful for what?’, but assumes that it should aim at the reduction of crime; crime reduction can however be achieved in other ways, and compensating or otherwise helping the victim is just as desirable an intention (Fatic’, 1995: 34-35). For the victim, seeing the truth acknowledged can be more important than seeing the offender dealt with severely, and as we have seen, these outcomes often have to be traded against each other; as Nils Christie (2001) has said, the conventional criminal justice process could hardly have been better designed to prevent the acknowledgement of the facts that are most important to the participants. Even accepting the conventional aim, most penal institutions are far from utilitarian: ‘The reality of prison life has nothing to do with any corrective, training or otherwise rehabilitative policies. Prisoners are beaten, raped, killed and commit suicide in “training” and “educational” facilities while under “protection” of the state’; utilitarianism ‘deliberately shuts its eyes to the social realities’ (Fatic’, 1995: 87). A still more fundamental consideration is the basic moral principle on which social action should be based: according to Ryder (2001) it should be to reduce the pain of others, and to be consistent it would aim to minimize not only crime but the infliction of pain by the state on its citizens. Ryder, however, does not directly address the question of punishment; he states that certain pain-causing actions, such as bullying, should be prosecuted (p. 84), and that wrongdoing by the police should be severely punished (p. 89), which seems at odds with his principle.

Penal measures also have a less desirable utility: crimes sell newspapers and attract television viewers, and it is the exceptional and frightening ones and the apparently ‘lenient’ sentences which are defined as ‘news’ and given the biggest headlines. In reacting to the fear that this produces among the public, politicians have, as we have noted, discovered the usefulness of tough rhetoric in giving the impression that they are ‘doing something’ about it.

2. Punishment as symbol: the impossibility of proportionality and the chimera of consistency The second main justification of punitive sanctions is that they demonstrate that certain acts are unacceptable. The idea is that punishments should be consistent, and their severity should be proportionate to the seriousness of the offence; but this is not possible except in a very rough-and-ready way. Sentencers are supposed to take account of so many competing factors that the scales of justice should have at least four dishes, in which deterrence, rehabilitation, retribution and containment would be weighed against each other; and the coming of reparation orders and victim impact statements in some countries means that a fifth should be added, representing the seriousness of the harm from the victim’s point of view Clearly under these circumstances there is no possibility of achieving consistency, but everyone finds it convenient to ignore the fact that judges are naked under their robes; that is to say, not only are they failing to achieve logic and consistency, but it is unachievable. As I have tried to show (Wright, 1999),

  • Harm – pain It is not possible to quantify wrongdoing, and if it were, there is no way logically to equate it to specific levels of punishment. There is no way of equating the seriousness of a particular crime with a particular punitive sanctions. As Fatic’ (1995: 72) has said, ‘institutions of criminal justice are basically incompetent to determine the exact amount of desert and therefore also the corresponding exact amount of retributive punishment’; in fact he says (p. 75), ‘ “desert” is nothing different from the reproach or dislike that the public feels for the offender.’
  • Pain for A – pain for B The same punishment has different effects on different people, and the system usually exaggerates these differences by sending the more privileged offenders to the less unpleasant prisons. Prison is a trauma for some, an occupational hazard for others. Likewise fines: even with unit fines, the loss of one day’s income has a different impact on a person earning £5 000 and one earning £50 000 (Duff, 1996: 59; cf Wright, 1999: ch. 5 and 6).
  • Short prison – long probation There is no way of integrating different sanctions into a single scale: no one can say that one week in prison is more or less severe than X months or years of probation, or Y amount of fines. Where is the crossover point?
  • Deterrence – proportionality The less a person has to lose, and the more desperate he is, the harsher are the punitive sanctions needed to deter them from even a small crime like stealing food. For a person of status, the conviction alone is deterrence enough. In any case, no one knows how much punitive sanction is necessary to deter a particular person, or the public at large; and to increase penalties is more symbolic than effective.
  • Mercy – proportionality Sentencers often feel the need to reduce the sentence in the light of extraneous factors, such as the fact that the offender had a good army record, or is suffering because she killed her best friend in a road crash, or rescued a child from drowning while awaiting trial, or is pregnant, or to increase it because there have been other similar offences recently, or for reasons they would not admit such as the colour of the offender’s skin.
  • Exemplary punitive sanctions are unjust Even if it were possible to determine a just punitive sanction for a particular offender, to increase it in order to ‘make an example’ of him or her in unjust, and precisely in proportion as it is exemplary (Lord Justice Asquith, quoted in Wright 1982: 191).


3. Both types of punitive sanction A third group of objections apply to both utilitarian and symbolic punitive sanctions. They inevitably tend to make the offender think of him or herself, not of the victim. They are not very effective in deterring offenders, but once the offence has been committed, they deter them from admitting their actions: in order to try to avoid them, offenders are less likely to confess, or if convicted, they try to minimize the harm rather than acknowledge its full impact on the victim.


  • Side-effects of punitive sanctions Even attempts to make punitive sanctions less harsh can have insidious results. It is harder to provide constructive activity such as education and training, and when this is done it is usually available only to a minority of prisoners. But if courts have the impression that prisons are rehabilitative, this can make them more likely to send offenders there. The same applies to facilities such as mother-and-baby units, which serve only to provide a choice of evils: to bring up a child in prison or to separate it from its mother. (Prisons also separate children from their fathers, of course.) It could even be said of restorative justice in prisons. Care must therefore be taken to make all these constructive activities available in the community. Punitive sanctions have other obvious undesirable side-effects: they tend to provoke resentment, and often superficial compliance; they stigmatize people, making it harder to obtain accommodation and work.
  • Collateral damage Punitive sanctions also have an impact on the offender’s family, and sometimes on the victim (as the Clotworthy case, discussed earlier, shows). This is unfair and counterproductive, especially when it has an adverse effect on children.


Restorativists do however need to address some problems, such as whether reparation should (or can) be the proportionate; whether responses to similar kinds of harm should be similar to each other; and, if restraint is needed to protect the public from an offender, how to decide whether deprivation of liberty is necessary or whether restriction is enough; and in either case, how to determine the length.

For all these reasons, despite the best efforts of scholars such as David Thomas, Nigel Walker; Andrew von Hirsch, and others, it is hard to see how the philosophical contradictions can be overcome (Wright, 1999). The conclusion is, as Fatic’ (1995: 192, 255) says, that ‘punishment in principle is morally unjustified’; and ‘the answer to moral concerns about infliction of pain is not to rationalize punishment but to eliminate it.’ Hoyles, writing from a Christian perspective, agrees: ‘The ultimate penal reform must be the abolition of punishment’ (1986: 139). Not only specific punitive sanctions, but punishment itself, would be recognized as ‘cruel, inhuman and degrading’.

The first step, then, is to label sanctions according to their intentions; the second is to eliminate those whose primary intention is pain delivery; and the third is to make the actuality match the name as far as possible, especially by maximizing the degree of voluntary compliance. Restorative measures may well be experienced as difficult and even burdensome, but they should not inflict pain for its own sake.


Managing without punitive sanctions

If punitive sanctions are unacceptable and ineffective, it is necessary to propose how people can be persuaded to behave decently to each other, and how to react when they do not. As Fatic’ (1995: 255) puts it, the theory ‘has to explain what mechanism would provide a social control equivalent for punishment’.

Such a social policy would have four main strands. It would be based on encouragement to behave well rather than threats to inflict pain for misconduct; when a person caused harm the primary response would be to try to repair it as far as possible; in addition people would be made aware of the ‘natural consequences’ which follow from harmful acts, rather than threatened penalties; and restriction or deprivation of liberty would be kept as a last resort when there was a serious risk of serious harm which could be prevented in no other way. Each of the last three becomes necessary only when the preceding one did not succeed.


The first is somewhat utopian: it supposes firstly that a society is possible in which everyone has a fair opportunity to make the most of him or herself and a fair share of material and spiritual resources, and that if such a society were attainable, people would not harm each other or try to obtain more than their share.

But even before that idealistic transformation of society, there is much that can be done to reduce crime by programmes which are not based on state intervention, but involve ‘the organizations, institutions and individuals of civil society’ (Garland, 1996: 451). Garland gives examples of ways of making crime harder to commit, such as employing parking lot supervisors, providing late night buses, and advising retailers about security (ibid. 451; see also Wright, 1982: ch. 9). A complementary approach is the involvement of young people in more constructive activities: early examples were the Wincroft Youth Project (Smith et al., 1972) and the French programme for keeping them occupied during the schools’ summer vacation (King, 1988).

Restorative justice is not limited to the criminal justice system. It is based on thinking and acting for others (and for oneself, as side-effect). There are initiatives to start it in schools, with methods such as circle time, peer mediation, and a problem-solving approach to discipline and bullying (Highfield Junior School, 1997). It can also be used in the workplace, families (not only in relation to divorce or separation), old people’s homes, or when problems arise between doctors and patients or police and citizens.


This is the basic principle underlying restorative justice; it includes making up for the harm as far as possible, dialogue with the victim (if there is an identifiable one), involvement of the community, and feedback to assist crime reduction policies. It will be elaborated below. What many victims want is acknowledgment of the truth of what happened to them; restorative justice, its advocates believe, offers an incentive to confession and disclosure, while punishment encourages concealment and denial.

Natural consequences

The idea that ‘virtue is its own reward’ is mirrored by the idea that ‘wrongdoing is its own punishment’ (Wright, 1982: 255-8, 262-3). As Fatic’ argues, everyday life in society basically depends on each person’s trust in others not to cause harm, and crime is a betrayal of that trust. Initially, therefore, the offender loses people’s trust, but the aim must be to enable him to regain it – not because of a belief in human goodness, but because individuals need trust, and ‘will be best off by accepting the moral norms arising from the principle of refraining from punitive infliction of pain as pain, of justice and vengeance, and from the adoption of reconciliatory and restorative values’ (Fatic’, 1995: 220). At its simplest, if someone tells lies, people will not believe him; someone found acquiring goods dishonestly will have them confiscated. An employee who betrays trust may face dismissal – a preventive measure, although it could also be regarded as punitive – or may be allowed to remain so that they can repay what they took. They may be required to pay the costs of any hearing (for example when charged with attempts, or crimes where no quantifiable loss), although this should not be disproportionate to the offence or to their means. People who cannot be trusted to drive a car, run a business or hold elected office can be disqualified from doing so, but given the opportunity to win back the trust they lost.

Those who have forfeited trust may be placed under supervision (or detention, if there is a substantial risk of serious violence – see below); but as in the other cases they should have the opportunity, and any necessary help, to make amends and regain the trust which they need, and which others need to give them.

Restriction of liberty

Even abolitionists concede that there are ‘a dangerous few’ who ‘have committed a series of dangerous, violent acts [and] need to be protected from their own violent impulses as much as we need to be protected from them’ (R. Morris, 2000). But they stress, firstly, that their numbers are extremely small; secondly that their conditions must be very different from those in present-day prisons; and thirdly that a policy of locking up the dangerous can only be justified if society has done all it can to prevent people from becoming so violent, through measures such as education of children in non-violence, programmes for abusers and gun control (R. Morris, 2000).


Conclusion: Principles and procedures for intervention

From these principles it is possible to develop a model of reducing harm, and responding to it when it occurs (crime-handling, to use Fatic’’s term). There would be an opportunity for those concerned to respond voluntarily before any coercive measures were used. This could include the concept of ‘sanctuary’, as ‘a place of refuge where the perpetrator of a serious offence could go and live for a while in safety until negotiations [about reparation] could begin’ (Bianchi, 1986).

For victims, members of the community would offer support, especially where no offender was known; there should also be financial assistance where they needed it and the offender was unable to provide it. When the offender was known, there would be an opportunity for dialogue between them. The process would not aim at encouraging forgiveness, but at providing conditions in which it could more easily occur.

Offenders would be given the opportunity to take responsibility for their actions and make amends, and to regain trust by undertaking positive action for the victim, for the community or for their own future life as trustworthy members of the community.

Incidents would be handled by community organizations unless it was necessary to involve state agencies, and these would divert cases back to the community if possible. The community groups would use lay volunteers as far as possible. They would help not only in the apprehension of offenders and subsequent processes, but, for example, by forming ‘circles of support’ for vulnerable offenders (Heise et al., 2000; Native Counselling Services, 2001), providing opportunities for them to make reparation, and managing victim/offender mediation services.

Last but not least, in a non-adversarial, non-punitive system people would speak more freely, and much could be learnt about the circumstances which make crime more likely: failures in parenting, in schools, in supervision of buildings and open spaces, in provision of work and recreation, and many more. These could be relayed to local authorities with responsibility for crime reduction strategy, and ultimately to those who influence social policy. Society would ‘eliminate the language of weapons’ (Fatic’, 1995: 233) and commit itself not to fighting crime (or even terrorism) but to building peace. The pillory, mutilation, flogging and the death penalty would be followed into history by punishment itself, and restorative justice would make a significant contribution to social justice.

I am grateful to Margarita Zernova for some constructive criticisms.



Barton, C. (2000) ‘Empowerment and retribution in criminal justice’ in Strang, H. and Braithwaite, J. (eds) Restorative justice: philosophy to practice (Aldershot: Ashgate).

Bianchi, H. (1986) ‘Abolition: assensus and sanctuary’ in Bianchi, H. and van Swaaningen, R. (eds) Abolitionism: towards a non-repressive approach to crime (Amsterdam: Free University Press).

Christie, N. (1982) Limits to pain (Oxford: Martin, Robertson).

 Christie, N. (2001) personal communication.

Daly, K. (2000) ‘Revisiting the relationship between retributive and restorative justice’ in Strang, H. and Braithwaite, J. (eds), Restorative justice: philosophy to practice (Aldershot: Ashgate).

EPOCH Worldwide (1996) Hitting people is wrong – and children are people too EPOCH Worldwide, 77 Holloway Road, London N7 8JZ, and Rädda Barnen, Stockholm.

Fatic’, A. (1995) Punishment and restorative crime-handling: a social theory of trust (Aldershot: Avebury).

Garland, D. (1996) ‘The limits of the sovereign state: strategies of crime control in contemporary society’ British Journal of Criminology, 36, pp. 445-471.

Garland, D. (2000) ‘The culture of high crime societies: some preconditions of recent “law and order” policies’ British Journal of Criminology,40, pp. 347-375.

Heise, E. et al. (2000) Community reintegration project: circles of support & accountability (Mennonite Central Committee Ontario: mcccos@look.ca)

Highfield Junior School (1997) Changing our school: promoting positive behaviour Edited by Priscilla Anderson. Highfield School, Torridge Way, Efford, Plymouth PL3 6JQ.

Hoyles, J.A. (1986) Punishment in the bible (London: Epworth Press).

King, M. (1988) How to make social crime prevention work: the French experience (Nacro, 169 Clapham Road, London SW9 0PU).

McElrea, F.W.M. (1995) ‘Accountability in the community: taking responsibility for offending’ Paper for Legal Research Foundation’s conference ‘Re-thinking criminal justice’, Auckland, NZ, 12-13 May 1995.

Mason, Sir A. (2000) ‘Restorative justice: courts and civil society’ in Strang, H. and Braithwaite, J. (eds), Restorative justice: philosophy to practice (Aldershot: Asgate).

Mattinson, J. and Mirrlees-Black, C. (2000) Attitudes to crime and criminal justice: findings from the 1998 British Crime Survey (Home Office Research Study 200.) London: Home Office. http://www.homeoffice.gov.uk/rds/index.htm

Moberly, Sir W. (1968) The ethics of punishment (London: Faber).

Morris, A. and Young, W. (2000) ‘Reforming criminal justice: the potential of restorative justice’ in Strang, H. and Braithwaite, J. (eds), Restorative justice: philosophy to practice (Aldershot: Asgate).

Morris, R. (2000) ‘But what about the dangerous few?’ in West, W.G. and Morris, R. (eds) The case for penal abolition (Toronto: Canadian Scholars’ Press).

Native Counselling Services of Alberta (2001) A cost-benefit analysis of Hollow Water’s Holistic Circle Healing Process. Lead Researcher: Dr Joe Couture. Aboriginal Corrections Policy Unit, Solicitor General of Canada, 340 Laurier Avenue West, Ottawa, Ontario, Canada K1A 0P8; http://www.sgc.gc.ca

Rosenberg, M. (1998) Expressing & receiving anger compassionately using the principles of NVC. [audiotape] CNVC, P O Box 2662, Sherman, TX 75091, USA. http://www.cnvc.org ; Non-violent communication: a language of compassion. Del Mar, CA: PuddleDancer Press.

Ryder, R.D. (2001) Painism: a modern morality (London: Centaur Press).

Smith, C., Farrant, M.R. and Marchant, H.J. (1972) The Wincroft Youth Project: a social-work programme in a slum area (London: Tavistock).

Tähtinen, U. (1982) Non-violent theories of punishment: Indian and western (Delhi: Motilal Banarsidass).

Windsor, A., princess royal (1991) What is punishment for and how does it relate to the concept of community? Rede Lecture 1990 (Cambridge: Cambridge University Press).

Wright, M. (1982) Making good: prisons, punishment and beyond (London: Burnett Books).

Wright, M. (1989) ‘What the public wants’ in Wright, M. and Galaway, B. (eds) Mediation and criminal justice: victims, offenders and community (London: Sage).

Wright, M. (1999) Restoring respect for justice (Winchester: Waterside Press).

Wright, M. (2000) ‘Restorative justice; for whose benefit?’ in European Forum for Victim-Offender Mediation and Restorative Justice (ed) Victim-Offender Mediation in Europe: Making Restorative Justice Work (Leuven: Leuven University Press).

Zehr, H. (1995) Changing lenses: a new focus for crime and justice (Scottdale, PA: Herald Press).

 Leuven punt 01C recovd rtf 026






Published in: Lode Walgrave, ed. Repositioning restorative justice. Cullompton, Devon: Willan Publishing, 2003.