ZIMBABWE: FROM IMPUNITY TO ACCOUNTABILITY punished at the cost of many others escaping altogether, along with the underlying problems and realities being minimalised and side-lined; vii) the possible high political costs of multiple prosecutions where the country is trying to undergo a ‘healing’ process; viii) a different form of political and social cost (and added trauma for victims) where high-profile prosecutions result in acquittals (as in South Africa, for example). All of these, and others, are factors that Zimbabwe civil society 32 needs to consider carefully. Although there are thus serious arguments that can be marshalled in good faith against prosecutions, there are also strong arguments that can be raised in favour. Some of these are as follows: i) in international law there is a duty on states to prosecute at least the most serious of international law crimes, namely genocide, war crimes, crimes against humanity (essentially widespread and systematic human rights abuses of a gross i.e. horrific nature), and torture; ii) there is a corresponding non-recognition of amnesties and impunity, both de facto and de jure, hence the principle of universal and extra-territorial jurisdiction for these crimes; iii) a basic component of full reparation is the right of victims to see perpetrators tried and punished; iv) to re-establish the rule of law and to build a positive human rights culture requires Zimbabwe to move away from selective justice, de facto as well as de jure amnesties, and the negative culture of impunity for ‘political crimes’ - there is need for all transgressors always to be equally liable before the law; v) prosecutions make perpetrators directly accountable for individual victims whom they have terrorized; vi) justice is done and is seen to be done, strengthening democracy; 32 A useful publication which Zimbabweans could examine is the handbook Reconciliation After Conflict (2003) produced by the International Institute for Democracy and Electoral Assistance (IDEA), which can be accessed at http://www.idea.int REDRESS acknowledges the use which has been made of IDEA materials in producing this paper. 22