Expanding the Defence: A Second Cut
To recapitulate, amnesties are sometimes morally justified as a necessary means to bring about a justifiedATC peace, albeit at the cost of punitive justice. I would like to end this section with four questions/points.
First, are there nonetheless crimes such that their perpetrators ought not to be granted amnesty? In some cases, notably in South Africa, crimes committed pursuant to a conflict and/or upon order such as interrogational torture were protected, whereas similar wrongdoings committed for personal gain or out of sadistic gratification under the cloak of war were not. In other cases, as articulated by the United Nations, crimes against humanity, war crimes, and (notably in domestic politics) serious human rights violations such as enforced disappearances are not protected. Finally, again as per the United Nations, there is a difference between international and civil conflicts. Whereas amnesties ought not to cover crimes against humanity in the former cases, they may do so in the latter case. According to the 1977 Protocol Additional II to the Geneva Conventions (art. 6.5), in the aftermath of civil conflicts, ‘the authorities in power shall grant the broadest possible amnesty to persons who have participated in the armed conflict, or those deprived of their liberty for reasons related to the armed conflict, whether they are interned or detained’.
Given how difficult it is to discern a torturer’s motives, and more generally to distinguish criminal violence from war violence, I am not convinced that the distinction drawn by the South African’s TRC works. The second distinction seems more promising, accounting as it does for the intuition that certain crimes are so serious that their perpetrators ought to be punished no matter what. However, and to adapt a point made above, if granting amnesty for genocide is necessary for and effective at preventing another genocide, there is a strong prima facie reason for endorsing such a move. Finally, reconciliation in the aftermath of a civil conflict may well require amnesties to a greater degree than reconciliation in the aftermath of an interstate conflict, since victims and perpetrators have to live in closer proximity to one another in the former case than in the latter case. Note, however, that this claim is entirely contingent on the fact of that proximity: there is nothing inherent in a civil conflict qua civil conflict which dictates in favour of amnesties, and there is nothing inherent in interstate conflicts qua interstate conflict which dictates against them. Thus, if the prospect of being prosecuted for (e.g.) the crime of aggression or incitement to genocide leads a negotiating party credibly to threaten to resume full-scale interstate war against its neighbours, or credibly to threaten to undermine reconstruction efforts (for example by withholding reparations), amnesty for those crimes might be justified too. In both cases, what justifies amnesties is the importance of bringing about a justifiedATC peace within and beyond borders.
Second, when conditional amnesties are sometimes justified, on what factors may they be conditioned? Standardly, on telling the truth about one’s deeds, since amnesties conditional on truth-telling can deliver a form of accountability which need not go hand in hand with punishment. Other possibilities include the willingness to go into exile, issuing apologies, and making material reparations. Note that in all those cases except exile, to impose the condition is to impose a burden which the agent is already under a duty to incur irrespective of what victims may demand: wrongdoers are in any case under duties to apologize (as we shall see in s.9.6), to pay reparations (as we saw in ch. 6), and to testify (or so I shall argue in s.9.5).
Exile is somewhat different. To claim that wrongdoers are under a duty to leave, irrespective of whether they are asked to do so, is to claim that they have forfeited their occupancy rights in the community of which they are members or residents just in virtue of having committed war crimes (whether against other members or strangers). But on the assumption that wrongdoers have rights to the necessities of life, they have a right to live somewhere. Unless their fellow community members in general and/or their victims in particular ask them to leave, it is hard to see why they are under a duty to do so of their own volition. It is in that sense that imposing exile as a condition for granting amnesty is not the same as making the latter conditional on apologies, confessions, and reparations. The question then is whether one may justifiably grant an amnesty conditional on exile. It seems that one may. In fact, it seems that exile tout court, unhinged from amnesty, is sometimes permitted as an extreme form of lustration, either on the grounds that wrongdoers must somehow be punished, or on the grounds that their past deeds have rendered them untrustworthy—so long as wrongdoers would not, in leaving and settling elsewhere, be subject to harsher burdens than the punishment which they are liable to incur anyway. If that is correct, exile in exchange for the granting of an amnesty is a fortiori justified.
Third, the rationale I have offered for sacrificing punishment to peace via amnesties has interesting implications for two related issues: the issue of who, of foreigners and compatriots implicated in the same war crime, may be granted an amnesty; and the question of which, of a domestic or foreign jurisdiction, may grant it. Consider the first issue. Amnesties are justified on the grounds that their beneficiaries would otherwise derail the peace process. Suppose that in the course of a civil war in community C, C’s leadership authorize its own military forces to commit grievous human rights violations against civiliansC, with the assistance of foreign actors. Suppose now that, post-conflict, those leaders, and not those foreign actors, pose a threat to the peace process. C’s new regime would be justified in granting an amnesty to the former, and not to the latter. In fact, it would be justified in prosecuting and punishing the latter (subject to other feasibility constraints such as financial costs). This of course is not a case of permissible partiality to one’s compatriots qua compatriots: what permits C’s regime so to act is not the fact that those former leaders are C’s members; rather, it is the fact that, in virtue of their membership in C, they pose a threat to peace. By that token, granting amnesty to some perpetrators but not others simply on the basis of shared nationality, for a crime which they committed jointly, isprima facie wrongful. In a particularly egregious example of such partiality, the French Parliament voted in 1953 to exempt from punishment French, but not German, members of the SS platoon who together had killed almost all inhabitants of Oradour-sur-Glanes in June 1944. Given that neither the French (former) soldiers nor those who opposed their being brought to trial were posing a threat to peace in 1953, this is an utterly condemn- able decision from a cosmopolitan point of view.5°
Consider now the issue of which, of a domestic or foreign jurisdiction, may grant an amnesty. The claim that amnesties are sometimes morally justified, indeed mandatory, for the sake of peace has two variants. In its strong variant, it might mean that their beneficiaries, for all that they have perpetrated serious human rights violations, nevertheless ought not to be punished by anyone, including external actors acting under the principle of universal jurisdiction. In its weak variant, it might mean that reasons not to punish only apply to domestic actors, thus leaving open the possibility of foreign prosecution. The case of Chile’s General Pinochet, who toppled democratically elected President Allende in 1973 and lost power in 1989 is interesting in that regard. His regime was marked by terror and repression in the course of which thousands of people were tortured to death, sent to jail without trial, and/or executed. Pinochet and the military and judicial establishment on which he relied agreed to hand over power to a democratically elected government (no doubt helped along the way by growing international pressure and the withdrawal of American support), but on one condition, couched as a threat which Pinochet himself issued: ‘No one touches anyone. The day they touch one of my men, the rule of law ends. This I say once and will not say again’. 51 Under those circumstances, Chile’s civilian government and its citizenry had very good reasons to grant amnesty to Pinochet and his henchmen.
But now suppose that there had been an international criminal court at the time, with the power to punish defendants in absentia. Indeed, suppose that it would have been possible to kidnap Pinochet and the key members of his regime for trial abroad, either by that court, or in the countries of some of his victims— Spain and the United States, for example. Pinochet’s crimes were such as to fall under the remit of the principle of universal jurisdiction as set out in s.7.6. In principle, then, the fact that amnesties are granted by domestic actors does not
- 5° See S. Farmer, ‘Postwar Justice in France: Bordeaux 1953’, in I. Deak, J. T. Gross and T. Judt (ed.), The Politics of Retribution in Europe: World War II and its Aftermath (Princeton, N.J.: Princeton University Press, 2000).
- 51 In T. Rosenberg, ‘Overcoming the Legacies of Dictatorship’, Foreign Affairs 74 (1995): 134—53, at 134.
preclude prosecution by other parties.52 If Pinochet’s threat would have been empty in the face of those prosecutions, conducted as they would have been by outsiders, the claim that amnesty in that case would have been justified holds only on its weak variant—that is to say, as far as domestic prosecution is concerned. In s.7.6, I argued that it often matters that a state whose members committed a war crime should act as the punishing party. By reverse reasoning, it often matters that the perpetrators’ state should not be the one to do so.
Finally, once amnesties are granted, they should be honoured, for if perpetrators know or even only have reasons to believe that, a few years down the line when their position is much weakened, they will be brought to trial, they are far less likely to accept the amnesty here and now when they are in a position to derail the peace process. However, this creates a real risk that amnesties will lead to collective amnesia, which in turn might be detrimental for peace: in fact, the word itself, ‘amnesty’, has its etymological origin in the Greek a^vrjarla, shares the same root as amnesia, or a^vrjala, and in its original sense denotes oblivion. Now, it is one thing for victims not to see their murderers, torturers, rapists in jail; it is another for those victims to be altogether forgotten. To be sure, many might deeply wish not to relive their own tortured past. But many others might insist on a public acknowledgment of their suffering, and/or private access to some of its details—for example, where their relatives were buried, who tortured them, who assassinated their son, that their father is dead. It is no coincidence, then, that the granting of amnesties has often been accompanied by requests for truth commissions (notably in Chile), or by initiatives on the part of civil society actors such as human rights organizations, local Churches, and victim movements (notably in Brazil and Argentina) to uncover and publicize information about past abuses.
Admittedly, there might be cases where perpetrators will resume war unless the facts of their grievous wrongdoings are wrapped in a veil of silence or publicly replaced by a distorting imaginary past in which they are as much sinned against as sinning. In such extreme cases, collective amnesia within war-torn communities might be the only way to secure peace. But there is no reason why outsiders should remain silent—on the contrary. Thus, the fact that Turkish leaders and citizens are for the most unwilling publicly to admit that their forebears committed a genocide against the Armenians in 1915 is no reason for others—other communities, historians, the world at large—not to say so. In fact, it furnishes them with stronger reasons still to make it possible for the Turks, in time, to come to terms with their past, by preserving knowledge thereof. In s.9.5 when tackling truth commission, and in ch. 10 when offering a cosmopolitan argument for war remembrance, I shall return to the importance of uncovering and recording the past in the aftermath of conflict. Suffice it so say for now that failing to honour those requests, or quelling those initiatives, when doing so would not jeopardize prospects for peace, would add to the injury of being made to come to terms with unendurable loss 
whilst perpetrators go free, the insult of being made to do so purely in private, as if it were no one’s concern but one’s own.
-  For discussions of the United Nations’ position on the one hand and more generally of eligibleand ineligible crimes on the other hand, see, e.g., Freeman, Necessary Evils, 88—108 and 145—51,respectively; see also Hayner, Unspeakable Truths, ch. 8, on the UN and amnesties.
-  L. Mallinder, ‘Amnesties in the Pursuit of Reconciliation, Peacebuilding and Restorative Justice’,in J. J. Llewellyn and D. Philpott (ed.), Restorative Justice, Reconciliation, and Peacebuilding (Oxford:Oxford University Press, 2014); see also Philpott, Just and Unjust Peace, 246ff.
-  2 That point was in fact asserted by the Special Court for Sierra Leone (a hybrid of UN anddomestic actors) in connection with the amnesties granted to perpetrators of atrocities, on both sidesof the civil war, in the Lome Agreement of 1999. I am grateful to Joanna Rozpedowski for bringingthis case to my attention.