Towards a restorative approach to post-conflict justice

(1)

TOWARDS A RESTORATIVE APPROACH TO

POST-CONFLICT JUSTICE

Master’s thesis in Political Science, specialising in Political Theory

Name: Joshua Armstrong

Email: armstrongj340@gmail.com

Student no.: 11287470

31 AUGUST 2017

SUPERVISOR: MICHAEL EZE SECOND READER: PAUL RAEKSTAD

(2)

1

Acknowledgements: I would first like to thank my supervisor Michael Eze, for both his teaching of the

class The African Renaissance, during which I learned so much, and his invaluable feedback, without which, the completion of this thesis would not have been possible. I also wish to extend my gratitude to my fellow members the African Renaissance research project group who provided a thought-provoking and inquisitive learning environment. My parents also deserve thanks for the material and moral support they have provided throughout the entirety of my university education. Finally, I would like to say thank you to Maria Maricaș whose support and advice was a constant reassurance throughout the process.

(3)

2

Abstract: The Retributive conception of justice often monopolises the discussion of justice which leads

to a limited understanding of justice without punishment. This limitation is particularly problematic within post-conflict justice theory as employing retributive approaches as a response to conflict often has unsatisfactory results. Attempts to move away from retributive theory as a response to conflict, as in South Africa and Rwanda, are often met with the charge that these efforts amount to a compromise of justice in favour of peace. This thesis will critique retributive theory in order to show that punishment is not a necessary component of justice. Having done so, restorative justice will be presented as an alternative account. This form of justice, through its focus on healing and reconciliation, is better placed to deal with the complexities faced by post-conflict societies and should, therefore, replace retributive theory as the basis of post-conflict justice. In order to show the value of the restorative conception of justice, two case studies will be used. These case studies will discuss the use of ubuntu by South Africa’s TRC as a way of responding to the atrocities of apartheid and the turn to gacaca by the Rwandan government following the 1994 genocide.

(4)

3

List of Abbreviations

ANC – African National Congress ICC – International Criminal Court

ICTR – International Criminal Tribunal for Rwanda RPF – Rwanda Patriotic Front

TRC – Truth and Reconciliation Commission UN – United Nations

(6)

5

1.0 – Introduction

The prosecution of those who commit human rights violations during conflict has been a particularly prominent topic since the inception of the International Criminal Court in 2002. It is argued that prosecution of war crimes perpetrators promotes peace and democracy by strengthening the rule of law, removing violent actors from the conflict, and by reducing the likelihood of future violations through deterrence (Branch 2011). Nonetheless, the efficacy of prosecuting perpetrators has also been questioned within post-conflict justice literature. For example, Synder and Vinjamuri (2003/2004) show that granting amnesty to perpetrators may, in fact, have a better record of supporting peace in post-conflict societies. Moreover, Mamdani (2015), uses the ICC’s 2005 intervention in Uganda to highlight the dangers of pursuing the prosecution of parties involved in conflict. Mamdani argues that after to ICC issued arrest warrants against the Lord’s Resistance Army, this effectively sabotaged both the process towards democracy in Uganda and the overall peace process.

Furthermore, within the academic literature on post-conflict justice theory, there are various examples of scholars discussing a need to balance ‘justice’ – meaning the prosecution of perpetrators – with the overall objective of attaining peace1_{. This blurs the theory of retributive justice with justice}

generally and, therefore, equates justice per se with the prosecution of perpetrators. Consequently, any innovative attempt at securing peace without necessarily pursuing the punishment of offenders is viewed as a compromise of justice in favour of peace. This is not necessarily so. Post-conflict justice should be seen as a more diverse concept than one that can only be understood through the lens of punishment. Indeed, restorative justice provides an account that conforms to the requirements of justice and may allow for the pursuit of peace in post-conflict societies while still enabling punishment to be cast aside.

This thesis will seek to establish restorative justice as a viable alternative to retributive justice within the post-conflict discourse. It attempts to redress the balance and provide further theoretical support for the use of restorative justice in approaches to post-conflict justice in the hope that the use of such theory may become more common in approaches to post-conflict justice.

For the purposes of this thesis, post-conflict justice will be defined as any societal or judicial response to human rights violations which occur during conflict. Post-conflict justice seeks to provide redress

1_{For examples of this see: Mamdani, M. (1996). Reconciliation without justice. Southern Africa Political and}

Economic Monthly. 10, 22-25 and Sarkin, J. (2001). The tension between justice and reconciliation in Rwanda: politics, human rights, due process and the role of the "Gacaca" courts in dealing with the genocide. Journal of African Law. 452, 143-172.

(7)

6 to victims and create opportunities for affected states to transform or adapt their political systems in order to prevent the conditions under which the human rights violations occurred (UN 2008). This thesis also makes reference to transitional justice. Although transitional justice is occasionally taken to refer to particular approaches to justice which favour truth reconciliation and accountability (Clamp & Doak 2012), for the purposes this thesis, the term ‘transitional justice’ is used interchangeably with ‘post-conflict justice’.

The thesis focusses on African societies for two reasons. First, the continent has been subject the majority of ICC investigations and its conflicts are thus most often subject to the perils of retributive justice (Branch 2011). Second, possibly due to some of the risks involved in prosecuting perpetrators, African nations have also been responsible for much innovation by way of incorporating restorative elements into the field of post-conflict justice. This thesis will use the two case studies of the South African Truth and Reconciliation Committee and Rwanda’s use of gacaca courts as examples of these restorative approaches. These case studies will demonstrate the applicability of restorative justice to post-conflict settings. An analysis of each case study will also provide historical evidence for the transformative potential of restorative justice theory, when applied to transitional settings.

In order to posit restorative justice as an alternative to retributivism as a basis for post-conflict justice, while retaining a focus on Africa, the following research question has been developed:

Can restorative justice play a central role in the conflict justice approaches of African post-conflict states?

Within this board research question, there are three sub-questions:

1. Does justice require punishment?

2. Can restorative justice provide an alternative to retributivism as the basis for models of post-conflict justice?

3. How can restorative responses to conflict be justified to victims and wider members of post-conflict societies? (Examining ubuntu and gacaca).

The first sub-question is aimed at showing that punishment is not a requisite component of justice. In order to do so, this thesis will provide a critique of Kant’s theory retributive justice which will highlight the internal weaknesses of his account of legal punishment by demonstrating that Kant does not sufficient justify the practice. Additionally, utilitarianism, the main alternative account of legal punishment, will also be discussed. I will propose a form of rule utilitarianism, which is immune from the general critiques from retributivists such as Michael Moore. This account is not committed to punishing wrongdoers in cases where social benefit will not necessarily be derived from pursuing their

(8)

7 punishment and is therefore able to withstand Moore reductio ad absurdum criticism. This will be done in chapter 2.

Having shown that punishment is not a necessary condition of justice, in chapter 3, restorative justice will be proposed as an alternative theory which still adheres to the requirements of justice in situations where the punishment is an inefficacious approach – i.e. where punishment will not augment social utility. As chapter 2 will have established that in these circumstances, it is acceptable not to punish wrongdoers, retributive justice offers a method to do so but still adhere to the needs of victim, perpetrator and society alike. This makes the theory particularly relevant to post-conflict justice as in post-conflict environments, social and communal bonds have often completely broken down. Chapter

3 will also demonstrate why post-conflict environments present unique challenges which often mean

that the usual assumed benefits derived from punishment are not present. This is another reason why a restorative account of justice, rather than some form of punishment-based account, is particularly suited to post-conflict justice theory. By pursuing this approach, chapter 3 will have answered the second sub-question.

In order to answer the third research sub-question, the final two chapters of the thesis will examine the use of autochthonous philosophies and practices, ubuntu and gacaca, in South Africa and Rwanda, respectively. These were employed as the intellectual/cultural basis for the two nations’ innovative approaches to post-conflict justice. As punishment of wrongdoers (especially human rights violators) is often deeply desired by victims, autochthonous cultural and intellectual resources, which are to some extent restorative, represent a method by which restorative justice may be integrated into post-conflict settings as these practices/philosophies may also prove popular. These practices/philosophies may allow societies to regain ownership of their own post-conflict response, rather than have a Western conception of justice imposed through the international community (through the ICC, for example). Both ubuntu and gacaca will be assessed in accordance to my account of restorative justice outlined in chapter 3 in order to illustrate that each approach was primarily restorative. The use of these case studies will seek to demonstrate that restorative justice can yield positive outcomes and also that African societies are perhaps more receptive than others to restorative justice because of varying conceptions of justice which exists in these societies as demonstrated by the existence of

ubuntu and gacaca. This is not to say Africans have a universal conception of justice, however, as

retributive justice is primarily a Western form a justice, this may make both the theory and practice less applicable to African societies (Clamp 2017).

(9)

8

2.0 – Retributive Justice: A Critique

Breach for breach, eye for eye, tooth for tooth: as he hath caused a blemish in a man, so shall it be done to him again. (Leviticus, 2008, 25:20).

This first chapter of the discussion is aimed at answering the first of my research sub-questions: Does justice require punishment? Clearly, this question is in the background of any debate regarding post-conflict justice, and particularly in this thesis which is advocating a new model without punishment at its centre. I cannot successfully answer my primary research question without first dealing with theory surrounding justice and legal punishment more generally. Therefore, the following chapter will seek to do this by firstly differentiating between the two different accounts which are typically used to justify legal punishment – utilitarian and retributive – in order to highlight that not all philosophical justifications of punishment are retributive. Secondly, this chapter will seek to show that the weakness of the retributive account, before proposing a utilitarian account of punishment which is prepared to allow murderers to go unpunished in cases where no social benefit will be derived from their punishment. This is significant for the rest of the thesis as the following chapter will show that, in many post-conflict cases, many of the perceived benefits derived from punishment are not present. Indeed, pursuing punishment may even come with various risks. As such, utilitarian justifications of punishment may not be present and by my account, punishment should therefore not be pursued. Consequently, an alternative conception of justice will be provided in chapter 3 which addresses the needs of perpetrator, victim and community in instances when post-conflict justice does not follow retributive theory and seek to punish rights violators.

There are two main competing accounts which justify legal punishment: utilitarian and retributive. (Tunick 1992). While these two competing philosophies essentially lead to the same result, that legal punishment is justified, the reasoning behind this conclusion is entirely different for each account. This chapter intends to critique the retributive account and to propose a conception of justice which does not necessarily entail punishment. In order to do this, it is first necessary to distinguish between utilitarian justifications of punishment and retributive justifications.

In its most basic form, utilitarianism is the theory which states that morally correct actions are those which produce the most ‘good’ (Driver 2014). This makes utilitarianism a consequentialist theory as moral evaluations of an action are based on the effects of this action, specifically how much ‘good’ a certain action produces. Within this broad definition, there are various types of utilitarianism. First, there is no specific agreement among utilitarians as to the content of this ‘good’ – what specific ‘goods’ should actions promote? Classical utilitarians take a monist view, stating that pleasure is the sole intrinsic good (ibid). Conversely, G. E. Moore (1988) postulates a pluralistic account of utility, meaning

(10)

9 that elements other than pleasure should be included in utility considerations. Specifically, Moore argues that beauty is another intrinsic good which should be promoted irrespective of any pleasure derived from it (ibid). This sets Moore apart from Bentham’s and Mill’s hedonistic account of utility as Moore expanded the conception of the ‘good’ which moral actions should produce.

Another debate within utilitarianism is between act utilitarianism and rule utilitarianism. Act utilitarianism states that when deciding upon a course of action, the most moral action is that which produces the greatest overall utility and these questions should be answered on an individual basis (Nathanson N.D). On the other hand, rule utilitarianism takes a two-part approach which focusses on moral rules. First, assuming strict compliance, an action is morally permissible if it obeys a justified moral rule; second, a moral rule is justified if its inclusion into our moral code creates more utility than would be created by the inclusion of alternative rules (Harsanyi 1986 & Nathanson). In essence, the difference between these two accounts is that act utilitarianism applies the utility principle directly in order to evaluate actions, whereas rule utilitarianism evaluates rules in accordance to the utility principle before evaluating actions by assessing whether they conform to these rules (ibid). Although the focus of this thesis is not to critique utilitarianism in general, these are important distinctions within the overall theory in relation to legal punishment and retributive justice, as will be shown below.

As mentioned above, classical utilitarians – specifically Bentham and Mill – believe pleasure to be the only intrinsic good (Driver). This is best demonstrated by Bentham’s principle of utility:

By the principle of utility is meant that principle which approves or disapproves of every action whatsoever, according to the tendency which it appears to have to augment or diminish the happiness of the party whose interest is in question (Bentham 1948, p.2).

Mill is also a ‘hedonistic utilitarian’, meaning that, like Bentham, his account holds pleasure to be the only intrinsic good (Driver). Despite this, Mill and Bentham differ in their conception of pleasure. Bentham contends that pleasure can only be quantitively different, while Mill first introduced the notion of higher and lower pleasures into utilitarian theory (Bentham 1948 & Mill 1992). This illustrates a significant theoretical difference between the two accounts as Bentham’s account treats all pleasure in the same way, including pleasure derived from ‘bed desires’. Contrastingly, Mill articulates an account of pleasure which is capable of explaining different types of pleasure, stating that ‘higher pleasures’ should be given more weight in utilitarian considerations.

Mill (1992) states that it is entirely consistent with the utility principle to recognise that pleasures can be of different qualities and therefore some are more valuable than others. Indeed, Mill goes further

(11)

10 by asserting that in assessing the amount of utility produced by a certain action, the quality of pleasure derived should be included as well as the quantity. In terms of assessing which pleasures (or pains) that qualify as more valuable and are thus ‘higher’, Mill uses the following consideration:

Of two pleasures, if there be one to which almost all who have experience of both give a decided preference, irrespective of any moral obligation to prefer it, that is the more desirable pleasure (Mill 1992, p. 120).

Mill’s conception of higher and lower pleasure is a significant addition to utilitarian theory. Using this concept, utilitarianism is able to show why human utility is has more moral weight than the utility of animals, for example (Driver). This is in contrast to Bentham’s account which is subject to the above criticism because of his sole focus on sensual pleasure. Given Bentham’s contention that all pleasure must be treated equally, it stands to reason that the human pleasure should be treated with the same regard as a animal pleasure since, according to Bentham’s account, all pleasure is sensual. While there are various ongoing debates surrounding whether animals should be given equal moral status to human, the purpose here is not to discuss that issue. However, unlike Bentham, Mill’s utilitarianism is equipped to explain the widely-held belief that human pleasure is in some way more valuable than the pleasure of other creatures. Because of this, Mill’s account represents a significant advance within utilitarian theory.

Turning now to the utilitarian account of punishment. Bentham claims that “all punishment in itself is evil” and is thus only justified according to the principle of utility in as far as punishment promises to exclude or prevent some greater evil. Bentham believes that the utility which is produced by punishment is deterrence, namely the prevention of further disutility (Bentham 1948, ch. 12, section 2). In this sense, then, utilitarians see punishment as being only instrumentally valuable, that it has value only insofar as it promotes some other ‘good’, the good here being deterrence and incapacitation.

One significant criticism of utilitarianism is that it takes into account sadistic desires, or the pleasure individuals derive from the pain of others. This criticism is particularly relevant to the utilitarian account of punishment because people may derive pleasure from criminals being punished. Michael Sandel makes this criticism of utilitarianism on the grounds that the utility principle requires the theory to take note of sadistic desires in utility considerations:

If enough cheering Romans pack the colosseum to watch the lion devour the Christian, the collective pleasure of the Romans will surely outweigh the pain of the Christian, intense

(12)

11 though it be. Or if a big majority abhors a small religion and wants it banned, the balance of preferences will favour suppression, not toleration (Sandel 1984, p.2).

One approach to defend utilitarianism against Sandel’s criticism is to claim that Sandel has misunderstood Bentham’s hedonistic calculus. Despite not making any qualitative distinctions between pleasures, Bentham’s hedonistic calculus – the method by which to evaluate the utility created by an action – allows for differing intensities of pleasure (Bentham). Therefore, a classical utilitarian account does not imply that an action should be taken simply because majority of a society would derive pleasure, at the expense of a minority, from this action. This is because differing preferences, pleasures and pains will be correspondingly different in intensity. Therefore, the majority, in Sandel’s example, who derive pleasure from the prohibition of the minority religion will not derive the same intensity of pleasure as the intensity of the pain suffered by the minority because of the prohibition of their religion. In this sense, utilitarianism is not committed to enforcing the religious ban as a way to augment social utility. This is a defence of utilitarianism offered by R. H. Hare. Hare (1986) claims that it is unproblematic for utilitarianism to give weight to bad desires, such as a sadist’s desire to torture, because the pain of the victim will be more intense than the sadists’ pleasure. Here, Hare uses Bentham’s notion of intensity to show that it is unproblematic for utilitarianism to take into account sadistic desires.

Nonetheless, Hare’s account is an inadequate response to Sandel’s critique because Hare’s appeal to intensity of pleasure is presumably only a matter of degree. Applying mathematical logic, even if we accept that the suffering of the minority, in Sandel’s example, is of greater intensity than the pleasure derived by the majority, at some point a large enough majority will presumably tip the balance in favour of banning the religion. This demonstrates that Bentham’s form of utilitarianism cannot properly defend against Sandel’s critique. An alternative solution is to adapt and extend Mill’s aforementioned concept of higher and lower pleasures. Mill only discusses pleasures but once we accept his argument regarding their qualitative nature, this can be extended to pains. This would mean that large amounts of lower pleasures cannot compensate for smaller amounts of higher pains when considering overall utility. In doing so, utilitarianism is not obliged to sanction sadistic desires in instances where they are favoured by a majority.

Returning to Sandel’s examples, it seems easy to conclude that the Romans in the Coliseum would prefer to forgo the pleasure gained by witnessing a Christian be devoured by a lion in order to also forgo the pain of being derived by this lion themselves. If this is the case, according to Mill’s criteria, the pain of the Christian is of a higher nature than the pleasure of the Romans. Thus, the utility principle would demand that the Christian not be persecuted the Christian’s pain would reduce overall

(13)

12 utility to a greater degree than the utility gained by providing lower pleasure to the many Romans. Mill’s concept of higher and lower pleasures provides a solution to Sandel’s critique in a way the Hare does not. This is because Hare’s discussion of intensity does not really solve the problem, it merely increases the threshold required for utility to sanction bad desires. Mill, on the other hand, by proposing an account of pleasure in which ‘higher’ pleasures are deemed more valuable to utility, also provides a way for us to understand extreme pains as having more weight in considerations of utility than lower pleasures.

Returning again to punishment, using Mill’s account of utilitarianism, the pleasure derived by society from seeing a wrongdoer punished, would not necessarily demand that a wrongdoer should be punished because pleasure deprived from vengeance is clearly of a lower quality than the pain caused from a depravation of civil liberties which results from custodial sentences, for example. The utilitarian account of punishment will be returned to later in the chapter but it was necessary to briefly outline key differences between this justification for punishment and that of retributive justice, as well as deal with a significant critique of the theory, in order to ensure clarity in remaining sections of the chapter.

Conversely, proponents of retributive justice have a contrasting view on the justification of punishment. Retributive justice states that punishment is intrinsically valuable – it has value, without appealing to other goods or beneficial consequences. Michael Moore (2010) states that the simplest way to distinguish retributivism from other justifications of punishment is that according to retributive justice, punishment of criminals is justified simply because criminals deserve to be punished. And going further, the theory also “obligates us to seek retribution through punishment of the guilty [Moore’s emphasis]” (p.154).

Immanuel Kant outlines a theory of justice that is purely retributive. In The Metaphysical Elements of

Justice (1965), he endorses punishment but specifically rejects utilitarian grounds for its justification:

Judicial punishment can never be used merely as a means to promote some other good for the criminal himself or for civil society, but instead it must in all cases be imposed on him only on the ground that he has committed a crime (p.100)

By grounding his justification for punishment only in the committing of a crime, Kant is endorsing the retributive notion that desert is a sufficient condition of punishment (Moore 2010). Further, Kant postulates that a criminal “must first be found to be deserving of punishment before any consideration is given to the utility of this punishment for himself or his fellow citizens” (Kant, p.100). Clearly, then, Kant takes a deontological rather than consequential approach to punishment – that is he appeals solely to the good of the practices itself (its intrinsic value) rather than any secondary benefit caused

(14)

13 by the practices. However, by simply grounding his account of just punishment in desert, Kant has not really justified the practice – if deserving a practice is the only criterion for its justification, this appears to be a circular argument. By this I mean that once one has committed to the notion of punishment being deserved, one has (presumably) already committed to the practice also being just (unless one is prepared to accept the claim that people deserve ‘unjust’ treatment and, if so, this is not itself an argument which justifies punishment, it simply accepts that some acts warrant an unjust reciprocal practice). In essence, if the grounding of a justification actually acts as the justification itself – as I believe the notion of desert does when grounding the justification of punishment – then the justification becomes almost tautological. The justification does not really do the explanatory work, it simply replaces the question of ‘why is punishment justified?’ with another, ’why does one deserve punishment?’. That these two questions are essentially the same, illustrates the argument’s circularity.

This is a critique of retributivism also put forward by David Dolinko. In order to shows the deficiencies of the above justification of punishment, Dolinko (1991) breaks down the concept of justification into two sub-questions. The first of these concerns the “rational justification” of a practice – in this case punishment – and the other the “moral justification” (p.539). The rational justification concerns the reason that we punish – why is it valuable or worthwhile – whereas the moral justification answers why it is morally permissible to engage in such a practice. Put simply, Dolinko differentiates these two concerns as “for what reason” and “by what right”. Having this distinction allows us more easily to see the limitations of Kant’s account.

As Kant here employs the desert criterion as a sufficient condition to justify punishment, he has only answered the second of Dolinko’s two questions which together make up justification. To show this more clearly, Dolinko highlights that the state does not usually take actions to give people what they deserve. For example, caring, charitable people are not automatically given awards despite their good actions and nature warranting such treatment (Dolinko 1991). Likewise, those who engage in morally odious behaviour, but do not contravene any laws, are not subject to bad treatment or censure at the hands of the state (ibid). In this context, Kant’s account struggles to articulate what is unique about criminality that demands the state to take action in order to give criminals what they deserve. Dolinko’s critique clearly shows that Kant’s justification simply explains from where the state derives its right to punish, but it does not provide an answer to the rational justification. This explains why Kant’s account seems an insufficient justification of punishment.

Fortunately, Kant adds further explanation to his justification of punishment. He claims that without enacting punishment, members of society become “accomplices in [the] public violation of legal

(15)

14 justice” (Ibid, p.102). This is perhaps best demonstrated if we consider the following Kantian thought experiment:

Even if a civil society were to dissolve itself by common agreement of all its members (for example, if the people inhabiting an island decided to separate and disperse themselves around the world), the last murderer remaining in prison must be first executed, so that everyone will duly receive what his actions are worth and so that the bloodguilt thereof will not be fixed on the people because they failed to insist on carrying out the punishment (ibid).

In this sense, then, we can see how Kant views the justification of punishment. It is justified for without it, the community becomes party to the crimes of the perpetrator and it must be enacted so that public law reasserts itself against the violation it has suffered. Therefore, Kant’s conception of retributive justice is that punishment is necessary – in that we have an obligation to punish – otherwise risk becoming party to this crime itself.

An alternative interpretation of Kant’s theory of punishment is that Kant is, in fact, a partial retributivist. This is the conclusion of Don Scheid (1983) who re-evaluates Kant’s writings on retributivism, claiming that the widely-held view of Kant as a “thoroughgoing retributivist” is incorrect (p.264). Scheid defines ‘partial retributivism’ as a theory that embraces the ‘retributive principle’, but uses crime control (deterrence, incapacitation etc.) as the general justifying aim of punishment. The retributive principle is defined as “[a]ll and only those who commit legal offenses may justly receive punishments so long as the punishments are in proportion to the seriousness of the respective crimes” (ibid p.263). In order to fully understand Scheid’s re-evaluation of Kant, we have to first look at Kant’s theory of freedom and his justification of the state.

Scheid claims that Kant justifies the state on consequentialist grounds that it is the only arrangement which guarantees individual liberty. In doing so, Kant implicitly justifies punishment in the same way. This claim – if correct – makes Kant a partial retributivist as he justifies punishment in accordance to crime control as crime is viewed as an infringement on individual liberty (ibid). In order to understand this more clearly, I will sketch out Scheid’s argument in the following two paragraphs.

Kant views individual liberty as the basic political good. Individual liberty is understood to be the absence of coercion by others. Therefore, any form of coercion must be justified. Also, this freedom must be reciprocal with the freedom of others and is thus specified “in accordance with a universal law” (Scheid, p.268). It is from this position that Kant justifies the state.

The state is justified because of its opposition to the state of nature as only the state can guarantee individual liberty (which is the basic political good). However, the state itself and its legal system

(16)

15 (including punishment) have coercive powers which reduce individual liberty and must therefore be justified. Kant provides this justification by claiming any infringement of individual liberty established by the universal law is wrong. In the state of nature, individual freedom is diminished and unevenly distributed. Therefore, the coercive power of the state is justified as this form of coercion prevents the infringement of universal individual liberty.

From Scheid’s above interpretation of Kant, we can see that Kant’s justification of the state is consequentialist as the state is justified on the ground that it promotes the greatest consequences for individual freedom. Importantly, Kant’s account differs from utilitarianism as the ‘currency’ of Kant’s consequentialism is individual freedom as he understands it, rather than utility or welfare (Scheid).

Moreover, Kant’s justification of the state is relevant to this present discussion of punishment because Scheid uses Kant’s account of the state to show that Kant was a partial retributivist. As the state is justified as the only arrangement which can guarantee individual freedom, therefore, punishment – as a correlate part of the legal system – is likewise justified on the ground that it acts as an external incentive for citizens to respect each other’s individual freedom (Scheid).

Persuasive though Scheid’s account is, it seems does not seem to correspond with Kant’s assertion that punishment should not be used to promote another good. As has been shown, Scheid’s assessment of Kant is based on the premise that Kant makes a consequentialist argument in favour of the state. This may be the case; however, it does not necessarily follow that Kant also endorses a consequentialist justification of punishment. Indeed, by positing that punishment should not be used to promote another good, Kant seems to rule this out. Moreover, by returning to Dolinko, we can see that even a partial retributivist account of punishment is problematic. Dolinko claims that we do not generally believe that the desert criterion makes it permissible to treat someone in a way that violates her rights. Dolinko uses an example of a dying father with two sons to highlight his critique.

The point introducing Scheid into the discussion was not to cast judgement as to Kant’s status as a ‘thoroughgoing’ or ‘partial’ retributivist but to show that either interpretation of Kant is a problematic account. In doing so, I will have further showed why retributivism itself is a flawed account of legal punishment. Assuming Kant is a partial retributivist and by returning to Dolinko, we can see that even a partial retributivist account of punishment is insufficient. In this case, deterrence of crime becomes the rational justification of punishment – Scheid states that, to Kant, punishment is a necessary element of the state. However, Kant also appeals to desert in order to provide the moral justification of the practice – this is why even in Scheid’s account, Kant remains a retributivist. Nevertheless, Dolinko argues that we do not generally believe that the desert criterion makes it morally permissible to treat people in way which would, under normal conditions, be deemed as violating individual rights

(17)

16 (as punishment clearly does) (Dolinko 1991). In order to show this, Dolinko provides the following thought experiment:

Consider, for example, Lear, a rich man with two sons, Jeremy and Howard. Jeremy truly loves Lear and has always treated him with affection and respect, even caring for him (at great personal sacrifice) during Lear's final illness. Howard, on the other hand, is a reprobate who has spent his time drinking, gambling, and chasing women, neglecting his father (for whom, in truth, he feels little regard) almost completely. Perversely, however, Lear has always felt a sneaking admiration for Howard while secretly despising Jeremy as a priggish, unimaginative, overly repressed bore. (This is grossly inaccurate, and unfair to Jeremy, but Lear at some level always wished he himself could have boldly defied the constraints of propriety and convention, as he believes Howard has.) Lear's will leaves Jeremy a comparative pittance and bequeaths the bulk of the estate to Howard (ibid, p.544).

Dolinko states that although most would agree that Jeremy deserves to inherit his father’s estate, the state will not take action to hand over Lear’s will to Jeremy simply because of this desert. To do so would contravene Howard’s right to the will. Similarly, a partial retributivist account of punishment does not explain why the notion of desert makes it morally permissible to violate the various rights which punishment necessarily entails.

Michael Moore (2010), a modern proponent of retributive justice, seeks to highlight that many of us share his philosophy. He outlines a series of thought experiments which are designed to first, disprove utilitarian justifications for punishment and second, to show that retributive justice theory accurately captures our emotional and instinctive responses to punishment. By doing so, Moore seeks to show that retributive justice is the correct justificatory account of legal punishment. In what follows, I will briefly sketch out Moore’s account by replicating his thought experiments and I will then demonstrate that utilitarian theory is not necessarily committed to an absurd conclusion (as he contends), and that my own conception of justice is not criticised by his account.

As Bentham’s utility principle states that utility (the happiness of interested parties) is the guiding force for moral action, and from this principle, Bentham (and other classical utilitarians) justify punishment by the benefit that it produces, chiefly the benefit of deterrence, Moore contends that this theory is therefore committed to punishing the innocent in cases where doing so would reap social rewards. To illustrate this point, Moore uses an example of a judge who knows that an accused man is innocent of a crime, but crucially the judge also knows that this man is a danger to society – for example, he has a high propensity to commit future crimes. In this case, then, there is a net social gain for the judge to find the man guilty, so that the high risk of future crime is negated. Thus, a utilitarian

(18)

17 must support the innocent man’s prosecution because utilitarian’s demand punishment when it augments overall utility (Moore 2010). However, this poses a problem for a utilitarian as inflicting punishment in this case goes against our intuition that we should not punish the innocent. Therefore, most utilitarians will still conclude that the innocent man should not be punished which, in this case, goes against the utility principle and is thus counter to their theory2_{. Moore sets out his critique of}

utilitarianism in three premises:

1. Punishment should be inflicted if, and only if, doing so achieves a net social gain. 2. A net social gain would be achieved in this case by the infliction of punishment. 3. Punishment should not be inflicted in this case (Moore 2010, p.95).

Clearly, this case poses a problem for classical utilitarians as they have come to a conclusion from which their premises do not lead – the conditions for inflicting punishment are met but they are not advocating its infliction because the man is innocent. In other words, they are arguing for a course of action (not prosecuting the man) by appealing to some value outside of utility (the man’s innocence).

This is a valid criticism of Bentham’s act utilitarianism. It is indeed true in the above example that social utility would be enhanced by the judge punishing the innocent man. Therefore, act utilitarianism should sanction this course of action as the act which produces the greatest amount of utility as it is likely that greater disutility would be caused by allowing the man to remain free. Evidently, Moore’s critique is problematic for Bentham’s form of act utilitarianism. However, this is not the final nail in the coffin for utilitarianism per se. Rule utilitarianism offers way to maintain adherence to the utility principle without being compelled to sanction the punishment of innocents in socially beneficial instances. As mentioned above, rule utilitarianism uses the utility principle to guide the imposition of moral rules and then judges an action in accordance with the extent that it complies with these rules. Therefore, it is possible to provide an account of rule utilitarianism which does not sanction the intentional punishment of innocents in any circumstances. As a legal system is a requisite part of any state – it is a tool through which the state exercises its coercive power – rule utilitarians can argue that a legal system which punishes innocents in socially beneficial cases would lose legitimacy if this procedure became a rule. The loss of legitimacy of a state’s legal system threatens to undermine the coercive power of the state and therefore, the state itself. Consequently, a rule utilitarian account of punishing innocents could claim that the disutility created by having an illegitimate and unpredictable legal system outweighs any social benefits created by punishing innocents in individual socially

2_{Jack Smart is an example of a utilitarian who takes the contrary position and is prepared to punish the}

innocent in socially beneficial cases. For his argument, see: Smart, J.J.C. & Williams, B.A.O. (2008). Utilitarianism: for and against. Cambridge: Cambridge University Press.

(19)

18 beneficial cases. Therefore, the correct moral rule is that innocents should never knowingly be punished even in specific individual cases where net social gains might be produced due to the greater disutility created by having an illegitimate legal system.

However, Moore (2010) also turns to another thought experiment which is aimed to show the weakness of what he terms the mixed theory of punishment – namely that we only punish when socially beneficial and where the offender is guilty. There are slightly different justifications for this theory than there are for rule utilitarianism but the general premises of the theories are essentially the same and thus Moore’s experiment aimed at rebuking the mixed theory is also valid for a critique of rule utilitarian’s account for punishment. Moore describes a scenario in which there are no utilitarian justifications for punishment. The murderer is truly repentant and is reformed to the extent that he poses no threat to society. Moreover, the crime can go undetected so that deterrence does not require punishment either. In this case, Moore poses the question of whether the murderer should be punished. Now, here Moore is hoping that we follow the following set of premises and end up with the absurd conclusion as outlined below:

1. Punishment should be inflicted if and only if doing so achieves both a net social gain and gives an offender his just deserts [read here ‘that the offender is guilty’ instead of ‘giving the offender his just deserts’].

2. A net social gain would not be achieved in this case by the infliction of punishment. 3. Punishment should be inflicted (Moore 2010, p.101).

If we do, in fact embrace premise three, Moore is correct in pointing out that the absurdity of this conclusion, given that the correct conclusion that follows from the first two premises is that punishment should not be inflicted on the murderer.

Despite Moore’s attempt to show that adherence to rule utilitarianism results in a reductio ad

absurdum argument, I do not share his conviction. In fact, I dispute the third premise and instead claim

that punishment should not be inflicted in this case. I maintain my adherence to the principle of utility and claim that as no social benefit will be derived from punishing the murderer, as he has reformed to the extent that he poses no threat to society (the actual method by which we are able to say this with certainty is irrelevant to the theoretical argument), especially when the vast cost of incarceration is taken into account, the murderer should go unpunished. This means that a theory of rule utilitarianism is not susceptible to Moore’s critique, provided that one is willing (as I am) to allow for circumstances in which murderers will go unpunished.

(20)

19 With regard to Moore’s argument, I have been able to articulate a theory of legal punishment which, so far, withstands his criticism and I am thus not committed to abandoning my theory and taking up a form of retributive justice. Nonetheless, Moore also utilises the concept of “moral outrage” to show that the retributive account of justice is the more “virtuous” (Moore 2010, p.144). Moore contends that “it is morally odious not to care about others […] then must it not be virtuous to feel such concern?” and therefore “must it not also be virtuous to feel negatively in some way towards flagrant moral violations that hurt others, virtuous to allow such negative feelings to cause retributive judgments?” (ibid). Here, I think Moore is guilty of overstating the point. I entirely agree with the first two elements of his argument. It is, indeed, unvirtuous to not care for others, as it is virtuous to feel negatively towards moral violations that hurt others – all Moore has done here is define the concept of empathy as virtuous and I imagine few would disagree – however, it does not follow that it is also virtuous to have retributive judgments because of our empathetic feelings towards others. I believe that Moore is essentially stating here is that he believes in the expressive function of punishment, that society should feel revulsion at certain acts and should express this revulsion through the practice of punishment (Feinberg 1970). It is clear that Moore does not take this to be the correct justification for punishment – I am not doubting his retributive credentials – yet I am questioning the extent to which the virtue of empathy can act as a proxy for the virtue of retributive judgments. Moreover, it is certainly not evident from Moore’s claims why the virtue of retributive judgments (even if accepted), would confer upon us the obligation to carry out such judgments through the practice of punishment.

There is, however, some merit in the account of punishment which focuses on its expressive function and this account also relates to the virtue of moral outrage as described by Moore. As hinted at above, the expressive function of punishment account is one in which justifies punishment because of its reprobative symbolism, that is:

Punishment is a conventional device for the expression of attitudes of resentment and indignation, and of judgments of disapproval and reprobation, on the part of punishing authority himself or those “in whose name” the punishment is inflicted (Feinberg 1970, p.98).

Feinberg’s conception of punishment has highlighted an important aspect of the practice. Legal punishment is usually carried out by the state which acts on behalf of its citizens. In enacting punishment, the state is engaging in “symbolic nonacquiescence” and the desire its citizens have for punishment stems from the need to see heinous criminal acts condemned (Ibid, p.103). This notion has parallels with Hegel’s obscure concept that punishment is necessary as leaving a crime unpunished would be “to posit it as right” (Hegel 1973 par. 218, addition). What Feinberg is articulating is similar to Moore’s notion of ‘moral outrage’ – because of these outraged feelings, society feels a need to

(21)

20 condemn criminal acts and punishment by the state (on behalf of its citizens) is the method by which this is done. This is certainly a significant element within the concept of punishment and any account of justice which is willing to advocate for the non-punishment of heinous offenders (as mine is) must take this charge seriously. I agree with Feinberg, and to an extent Hegel, in that to simply allow a murderer to walk free would certainly not adhere to any form of justice as this could almost certainly be taken as condoning such an act (and thus posit it as right). However, I believe that public condemnation can be sufficiently articulated through symbolic reprobation so that the need to punish, for the exclusive reason of illustrating society’s repulsion, is unnecessary. Therefore, in the above scenario described by Moore, in which I was happy to leave the murderer unpunished, I also believe that utilitarian concerns should also demand that the murderous act be condemned. As already stated, not doing so would run the risk conferring societal legitimacy upon such acts and this, in my view, would certainly count as a social harm.

Furthermore, Feinberg, in the hypothetical at least, is willing to conceive of a situation in which reprobative symbolism is meaningful enough for there to be no need for ‘hard treatment’ to illustrate condemnation. He sketches a scenario in which a society has within its culture a certain public ritual which is able to demonstrate condemnation so emphatically that there can be no doubt of its genuineness (Feinberg, 1970). He states that “[s]uch a device would preserve the condemnatory function of punishment while dispensing of its usual physical media” (Ibid, p.116). It is irrelevant that Feinberg also states that such an example is perhaps “only idle fantasy” (ibid), he still leaves room within his account of punishment for no actual punishment, in the way that it is normally understood (i.e. no ‘hard treatment’). This will be an important admission when the South African TRC is discussed in later chapters and the concept of ‘reintegrative shaming’ discussed in the following chapter. Another important admission of Feinberg’s is his understanding of the significance of cultural difference: he is aware that under certain conditions, reprobative symbolism can itself be painful without the accompanying “hard treatment” of punishment (ibid), whereas under other conditions, what we would conventionally regard as punishment may not similarly express public censure and may thus not be painful to the same extent (ibid, 114). Again, this will be discussed in later chapters when the importance of autochthonous philosophy to conceptions of justice is discussed.

This chapter has sought to highlight different justificatory accounts of punishment – namely retributive and utilitarian – and has provided a critique of retributive justice that has shown the weakness of the retributive claim that desert itself confers upon us an obligation or duty to punish. Instead, I have provided an account of rule utilitarianism that, as an account of punishment is willing to allow even murderers to go unpunished in socially beneficial situations, and as a consequence does not fall prey to Moore’s reductio ad absurdum argument. Moreover, I have illustrated that this form

(22)

21 of utilitarianism understands the importance of expressing the moral outrage when rights are violated but is not committed to the use of punishment for this purpose. This chapter, then, has laid the groundwork for the rest of the thesis. I have answered the first research sub-question – Does justice require punishment? – by providing an account of justice that is not committed to punishment. It would have been impossible to provide an account of punishment in post-conflict justice, especially one which rejects retributive theory, without first discussing legal punishment more generally and providing a critique of retributive justice at the basic level. It is a further challenge to link my account to post-conflict justice settings but this will become easier after the following chapter, in which restorative justice theory will be discussed.

(23)

22

3.0 - Outlining an Account of Restorative Justice

Learning to forgive is much more useful than merely picking up a stone and throwing it at the object of one's anger, the more so when the provocation is extreme. For it is under the greatest adversity that there exists the greatest potential for doing good, both for oneself and for others. - Dalai Lama (Braithwaite 1999, p.2)

Having established a view which articulates punishment as theoretically and morally problematic, it is now necessary to provide an alternative. Restorative justice represents precisely this. Just as retributivism claims that punishment has intrinsic value, the values of healing, forgiveness and community participation are central to restorative justice. Indeed, John Braithwaite states that restorative justice promotes “healing rather than hurting, moral learning, community participation and community caring, respectful dialogue, forgiveness, responsibility, apology and making amends” (Braithwaite 1999, p.6). It is obvious from Braithwaite that restorative justice has a different set of concerns to retributive justice. However, it is not immediately clear that it is an alternative theory. Within the restorative justice literature, there is disagreement about the precise nature of the concept and also about the extent to which it provides an alternative to retributive justice.3_{Despite these}

debates, I intend to articulate a detailed account of restorative justice in order to critically compare it with retributivism. This account will then be used in later chapters to demonstrate that specific examples of post-conflict justice in Rwanda and South Africa were inherently restorative projects. Moreover, my account, because it rejects the moral concerns associated with retributivism, will provide an alternative to retributive theory, especially within the context of post-conflict justice. I also wish to outline the types of practices that would typically be expected from processes of restorative justice, again with a specific focus on post-conflict settings. I will pay particular attention to the work of Howard Zehr, who is a strong proponent of restorative justice, albeit within the context of national criminal justice, rather than transitional settings. Nonetheless, his criticisms of the failings of Western retributive criminal justice systems, with regard to victims and offenders, are particularly relevant to transitional contexts.

In the most basic form, what differentiates retributive and restorative justice is that retributive justice is backward-looking – concerned with establishing guilt and ensuring that the guilty party receives what she deserves (Zehr 1997). Moreover, retributive justice is primarily concerned with enacting punishment which is an inherently backward-looking as retributive justice views punishment as

3_{For a detailed discussion of these debates, see Clamp K., & Doak J. (2012). More than words: Restorative}

justice concepts in transitional justice settings. International Criminal Law Review. 12, 339-360. Allais, L. (2012). Restorative Justice, Retributive Justice, and the South African Truth and Reconciliation Commission. Philosophy and Public Affairs. 39, 331-363.

(24)

23 intrinsically valuable. Therefore, the forward-looking (instrumental) elements of punishment, such as deterrence, are absent from the theory. Restorative justice is forward-looking, on the other hand, and is “primarily oriented towards doing justice by restoring the harm that has been caused by a crime” (Bazemore and Walgrave 1999, p.48). This means, then, that the central concern of restorative justice is making amends and promoting reconciliation both between individual victims and offenders as well as between the offender and the wider community. Such a notion makes the theory particularly applicable to post-conflict societies as in these contexts, reconciliation will often be a high priority, or even a requirement, so that the newly peaceful state can move towards the establishment/consolidation of stable democracy. Having highlighted the essential difference between these two conceptions of justice, I will now turn to establishing further the tenets of restorative justice.

As mentioned above, there are various values which are said to be central to the concept of restorative justice. An important value is accountability. Howard Zehr offers a different conception of accountability to the one which is usually espoused at various stages of Western criminal justice procedures. In both conceptions, accountability is understood to involve taking responsibility for one’s actions. However, according to Zehr (1997), a second component of accountability is taking responsibility for making things right. This is something which offenders are usually denied during standard criminal justice procedures. Including this notion into an account of restorative justice further sets it apart from any retributive account which may also appeal to a notion of accountability but instead proponents of retributivism seems to understand offenders being held accountable as offenders simply receiving their ‘just deserts’ (ibid). Having the opportunity to make this right will also be central to any process of reconciliation. It is difficult to imagine how any two parties may begin such a process without first the victim feeling as though the offender has made every effort to atone for her crime. As my account of restorative justice is located within transitional contexts, the theory’s ability to promote reconciliation is of central importance, given that reconciliation is key to building a new society following a period of conflict. Therefore, accountability is another key value of my account of restorative justice.

Essentially, what is central to restorative justice is the idea that justice is best served through a commitment to reconciliation rather than retribution. Reconciliation is achieved through community participation, dialogue, truth-seeking, responsibility, apology and accountability. Through these processes, especially by taking responsibility, perpetrators and victims can be simultaneously restored. Once responsibility has been accepted, victims and offenders can move beyond these antagonistic labels and new relationships can be forged. With this in mind, my account of restorative justice is both process- and outcome-focussed. By this I mean that I cannot conceive of restorative

(25)

24 justice without having both a restorative process and restorative outcome. For instance, a community-based forum in which the offender and victim discuss the offence and those affected (including the wider community) come to a consensus regarding how best the offender can make amends would be an example of a restorative process. However, if the recommendation following this process entailed a severe punishment (e.g. a long custodial sentence), I do not believe that this would adhere to the logic or values inherent to restorative justice. All that has happened in this example is that we have simply changed the process by which we have arrived at a retributive outcome. In other words, we have included participation of the community and victim directly into the sentencing process but we still have an outcome that is very similar to what would have been achieved if standard Western criminal justice procedures had been followed. Similarly, following a standard procedure of prosecution, if the court recommended community service and compensation (so the offender has the opportunity to make amends directly) instead of custodial sentences for serious offences, this also cannot be described as an instance of restorative justice. This is because, despite the ‘restorative’ outcome, the whole procedure is lacking any opportunity for community participation, dialogue, truth-seeking, apology, and responsibility and is therefore not a restorative process. This inclusion of both outcomes and processes into my theory sets it apart from other proponents. Morris (1995, cited in Braithwaite 1999) for example, claims that “any outcome – including – a prison sentence – can be a restorative outcome if it is agreed to and considered appropriate by the key parties” (p.86), while Braithwaite (1999) argues that “stakeholder deliberation determines what restoration means in a specific context” (p.6).

Such procedures would, of course, be to some extent restorative – and certainly more so than most Western criminal justice systems – but I do not believe that a collectively-agreed upon custodial sentence fully encapsulates the notion of restorative justice as the procedure will still have retributive logic within it, or at least value the instrumental value of punishment. This is something which restorative justice should be against. The reasons for this are simple. If community-led deliberation becomes the norm for a justice system when deciding upon the fate of offenders, it is easy to see how this process could become corrupted into nothing more than collective vengeance or even a mob mentality. Simply allowing the various stakeholders to administer punishment, rather than a judge, does not result in a significantly different conception of justice to warrant calling it restorative. Moreover, such a procedure will not have altered some of the troubling moral concerns of retributivism as discussed in chapter 2.

The content of restorative justice has been outlined above and I will now turn to explaining exactly why this is the case. The starting point of this is community participation, as the effect of any crime stretches beyond the victim and her friends/family and flows directly into the wider community. By

(26)

25 community participation, I mean that the community should be involved in hearing out the victim and offender as well as playing a role in deciding what the outcome of the restorative process should be (i.e. in deciding the offender can take steps to make amends). Involving the community in every aspect of a justice procedure will have two benefits. First, the community may begin the understand the rational of offenders in committing a crime which will assist towards reconciliation. Second, as an interested party, the community will also have a stake in the outcome and this should provide motivation to ensure that correct outcomes and are arrived at as everyone will benefit from the successful reintegration of the offender into the community. This is true of any crime but it particularly true in post-conflict settings which usually feature antagonistic groups and communities, at least one (but often both) of which has been subject to persecution and harsh treatment. Community cohesion is also a logical way to ensure that individual rights are protected, as conflicts often arise out or inter (or intra) community resentment, for example sectarian conflict, and it is during these conflicts that individual rights are violated. In this context, any attempt at achieving reconciliation is dependent upon these two groups finding a way to live peacefully together. Therefore, it is necessary that communities should be directly involved in the justice process.

Indeed, Zehr (1997 & 1998) strongly criticises Western criminal justice systems for denying both victims and communities any role in justice process. Zehr claims that by way of the crime, the offender removes power from the victim and subsequently, the criminal justice system does the same by not granting the victim a role in the process which renders them “mere footnotes” (Zehr 1997, p.1) in the procedure. Offenders are also only present in the process in order to establish guilt. Retributive justice does not provide any opportunity for the offender to directly make amends to the victim and the offender is, thus, denied the prospect of restoring herself.

Moreover, Zehr states that by leaving communities out of the process, criminal justice systems allow divisions and stereotypes to grow as the system encourages people to view the offender, and criminals more generally, through an antagonistic lens of ‘us and them’ (Zehr 1998). This is particularly significant in divided post-conflict societies where mutual distrust between opposing communities is already extremely high. Therefore following standard retributive criminal justice procedures in these contexts will entrench division and prevent reconciliation as victims and wider communities only see that guilt is established and punishment administered, rather than any effort made to understand the offender or his motivation for committing the crime.

The same is true of dialogue, which is only really meaningful if it occurs between the victim, offenders and the wider community. It is through a process of dialogue that those affected become more willing to understand the offender’s rationale for committing the criminal act. Understanding then lays the

(27)

26 foundations for reconciliation and reintegration. Simply establishing guilt and inflicting pain, as is the case in most retribution-based criminal justice models, cannot provide the environment for reconciliation. This is one criticism that Zehr (1997) levels at retributive criminal justice procedures. He states that standard Western models of criminal justice can be summarised as the “establishment of guilt and the infliction of pain” (p.6). This provides no avenue for the victim of wider community to truly understand the offender and the process also sets up clear division – as the offender is seen as guilty (as well as possibly evil and/or delinquent, depending on the crime) and the victim and wider community as innocent. Evidently, such divisions will harm future prospects of reintegration.

The most fascinating contrast between to the two theories of justice that we are discussing is that retributive justice locates the perpetrator at the centre of the process at the expense of the victim, whereas the central tenant of restorative justice is the community (which includes special place for both the victim and perpetrator). Of interest here is that critics of restorative justice claim that without punishment, justice, from the victim’s perspective, is somehow compromised. The irony, of course, is that in doing so, retributive justice locates the perpetrator at the centre of the process as the need to enact punishment supersedes the needs of the victim or wider community. In other words, retributive justice is perpetrator-focussed, while restorative justice takes a holistic approach, ensuring that victim, perpetrator and community are all provided with opportunities to be restored. In this way, retributivists allow perpetrators to reign supreme over the victim when perhaps the victim’s needs, and those of the wider community, are best served by allowing them to regain a degree of control which has been diminished by the violations of the perpetrator. Restorative justice provides an opportunity for this control to be reasserted by providing space for both the victim and community to play a constructive role the justice process, without diminishing the role of the perpetrator as is done to the victim under retributivism.

The processes of community participation and dialogue, that are inherent in restorative justice, are those which are particularly lacking from retributive procedures and especially those that are aimed at prosecuting offenders for offences committed during conflict. The internationalisation of war crimes justice, evidenced by the creation of organisations such as the International Criminal Court, only heightens this effect. For example, having international tribunals prosecute war criminals takes the justice process away from those affected and denies these individuals or groups any ownership of the proceedings. This can increase feelings of apathy, anger and resentment felt by victims who may believe that they should play a greater role in the process. This was particularly true of the International Criminal Tribunal of Rwanda which took place in Tanzania. Rwandans felt largely ignored by the Tribunal and many were unaware of the verdicts handed down by the ICTR (Graybill 2004). Clearly, such an effect must be inimical to reconciliation efforts and cannot help foster the spirit of

Towards a restorative approach to post-conflict justice