Christopher Bennett

Amnesty in the context of national reconciliation involves waiving or cancelling the punishment of convicted or suspected criminals in the name of peace. We can distinguish three positions: amnesty is wrong because it is unjust; amnesty is unjust, but necessary; and amnesty is just because it expresses forgiveness. The third position sounds promising. However, it assumes that when we forgive, we can justifiably waive or cancel the need for punishment. I argue that only punishment that expresses repentance and atonement brings about true reconciliation between the wrongdoer and the rest of the community. If we forgive in the absence of repentance and atonement, we restore our relationship with the wrongdoer, but in doing so ignore the way the wrongdoing conditions the relationship. An adequate, properly reconciled relationship can only be forged on the basis of some agreement on fundamental values, and that requires a change of heart from the wrongdoer. Forgiveness cannot properly be conceived as cancelling the need for repentance, atonement and punishment

Christopher Bennett presents a theory of punishment grounded in the practice of apology, and in particular in reactions such as feeling sorry and making amends. He argues that offenders have a 'right to be punished' - that it is part of taking an offender seriously as a member of a normatively demanding relationship that she is subject to retributive attitudes when she violates the demands of that relationship. However, while he claims that punishment and the retributive attitudes are the necessary expression of moral condemnation, his account of these reactions has more in common with restorative justice than traditional retributivism. He argues that the most appropriate way to react to crime is to require the offender to make proportionate amends. His book is a rich and intriguing contribution to the debate over punishment and restorative justice.

How can a state be morally justified in punishing some of its citizens? In tackling this I shall set aside three important matters: we do not morally approve of all the laws of the land, so that sometimes there is a legal but not a moral case against an offender; we can do more things about crime than just punish the criminals, for example remedying the familial and social conditions that encourage it; and, thirdly, many actual penal institutions do things to convicts that are indefensible by any decent standard. My topic may be humanly less important than any of those three; but I want to discuss it and not them. Like many others, I hold that the punishment is justified only to the extent that it does good, that is, leads to a better over-all state of affairs than would have obtained otherwise. Some versions of this view, however, assign to doing-good a role in which it swamps other considerations which most of us think are also essential to a defensible penal system. My aim in this paper will be remedy this defect - to reconcile a doing-good justification with the rest of what we think about punishment. The fear that this cannot be done has led some to question whether our having penal systems and procedures does any good. Jacobs may be right in stressing how hard it is to discover what the consequences are of any given system of punishment (1999: 540), but it would be lamentable if that led us to omit doing-good from our view of what justifies punishment. I shall return to this topic at the end of Section 7, offering to relieve Jacobs of his fears. I usually think of the good that punishment can do in terms of the deterrence of potential offenders - the punished person or others who learn about his punishment. In all the uncertainty about what our penal procedures achieve, it seems certain that they have a deterrent effect without which our society would be lost. There are other possible goods: making the convict a better person; placating victims and their kin; increasing our sense of the majesty and importance of the law; and so on..

In this paper I am concerned with a problem for communicative theories of punishment. On such theories, punishment is justified at least in part as the authoritative censure or condemnation of crime. But is this compatible with a broadly liberal political outlook? For while liberalism is generally thought to take only a very limited interest in its citizens’ attitudes (seeing moral opinion as a matter of legitimate debate), the idea of state denunciation of crime seems precisely to be focused on the attitudes expressed in action. In this paper I analyse the elements of the communicative theory of punishment, assessing the extent to which they can be considered anti-liberal. I argue that, understood in a certain way, the communicative theory, though in some sense communitarian, is compatible with at least one central and attractive non-perfectionist strand in liberalism. Key Words: communicative theory • Devlin • Duff • Hegel • von Hirsch • punishment.

Our characters are formed, at least in part, by forces beyond our control. Should this lead us to mitigate the responsibility of those who turn out badly? Martha Nussbaum argues that we ought to be merciful to wrongdoers on these grounds. Against Nussbaum, I argue that we have important moral reasons to treat people as responsible for character and hence to eschew mercy. Treating someone as responsible is required if we are to treat them as a moral agent, to treat them as having a moral point of view that is worth hearing. However my point is not that there are no reasons for mercy, but simply that there are other reasons, to do with respect, that pull in the opposite direction. The result is not a decision procedure for all cases but a better understanding of the complex moral geography.

Some philosophers think that forgiveness should only be granted in response to the wrongdoer’s repentance, while others think that forgiveness can properly be given unconditionally. In this paper I show that both of these positions are partially correct. In redemptive forgiveness we wipe the wrong from the offender’s moral record. It is wrong to forgive redemptively in the absence of some atonement. Personal forgiveness, on the other hand, is granted when the victim overcomes inappropriate though humanly understandable feelings of hate or vindictive anger towards the wrongdoer and comes to see them as deserving of a certain unconditional respect ’qua’ moral agent. Personal forgiveness is unconditionally admirable

In my response to Golash I distinguish between two steps in my original argument. The first relates to the special value of conjugal (two-person) love relationships. I defend this step against criticisms, arguing that the two-person relationship provides a form of recognition that is of special importance to us and cannot be found in other sorts of relationship. The two-person relationship is one that, at least as private individuals, we have special reason to pursue. The second step concerns the claim that the special value of such relationships tends to promote the autonomy of those who have them. It is this second step that is important for the argument that a liberal state – one, at any rate, that takes itself to be in the business of safeguarding the pre-conditions of autonomy – could have reason to favour marriage or some form of civic partnership over other forms of intimate adult tie. However, I admit that Golash puts forward plausible – if anecdotal – arguments against this second step. I therefore agree that I need to be more tentative about this step than I was in the original paper

This paper argues that a liberal state is justified in promoting relationships of conjugal love – the form of relationship that is the basis of the institution of marriage – on the grounds that they are essential to the development and maintenance of autonomy. A deep human need is that the detail of our lives be recognized (accepted, affirmed, granted importance) by others (or by an other). Autonomy can be compromised when this need is not met. So a state concerned with autonomy ought to be concerned with relationships in which people can be given recognition. This argument justifies support for friendship as well as conjugal love; why is the latter particularly special? The answer is that in conjugal love partners value each other exclusively (i.e., in a way they do not value anyone else). Conjugal relations therefore recognise the uniqueness and individual value of a person's life in away that friendship does not

abstract In this paper I take seriously von Hirsch's view that sanctions imposed on offenders need to be compatible with their dignity, and argue that some versions of restorative justice — notably that defended by Braithwaite — can put offenders in the humiliating position of having to make apologies that they do not believe in in order to avoid further bad consequences. Drawing on recent work by Duff I argue that this problem can be avoided by conceiving of restorative justice as an apologetic ritual. This view gives some ground to von Hirsch but presents a view of criminal justice that is distinctively restorative. I conclude by drawing out the differences between my account and that of Duff.

Retribution is often dismissed as augmenting the initial harm done, rather than ameliorating it. This criticism rests on a crude view of retribution. In our actual practice in informal situations and in the workings of the reactive (properly called 'retributive') sentiments, retribution is true to the gravity of wrongdoing, but does aim to ameliorate it. Through wrongdoing, offenders become alienated from the moral community: their actions place their commitment to its core values in doubt. We recognize this status in blaming, a withdrawal of civility and solidarity which symbolizes the moral distance wrongdoers have put between them and us. Atonement is the means by which they make themselves 'at one' again with the community. Retribution is properly understood as a cycle which recognizes disruption and alienation, but aims at reconciliation