attended mostly by ZANU (PF) members on the site of a prior militia base, participants alluded to their complicity in earlier violence and foreswore future participation. Other workshops have included direct exchanges between perpetrators and survivors, acknowledging the harm done to the community. Conclusion The grassroots approach epitomised by ToL has manifest limitations, and cannot accomplish the full range of transitional justice goals, most importantly individual accountability for violations. Such community-based approaches must at some point be supplemented by some combination of prosecutions, reparations and other accountability strategies backed by a rights-observing successor regime. But in the absence of the necessary political transition, the unavailability of centralised justice processes should not preclude grassroots reconciliation initiatives. The emerging Zimbabwean experience indicates that such initiatives can be successful, and this success may in turn contribute to community solidarity, reducing the scope for future violence instigated by outsiders during elections and other moments of political contestation. The preceding discussion also highlights a deficiency in much contemporary transitional justice debate, which views the functions of the centralised state as the sine qua non of transitional justice processes. International NGOs are still influenced by the paradigmatic model of the South African Truth and Reconciliation Commission and its institutional cousins, and therefore seek to collaborate with successor governments and national-level civic organisations to establish high-profile national-level processes, sometimes at the expense of smaller, grassroots initiatives. Transitional justice practitioners should reexamine their priorities, particularly in protracted “complex emergencies” akin to the Zimbabwean crisis, where a political solution may come too late for many survivors. A renewed focus on grassroots initiatives will allow for greater engagement by victims and survivors in transitional justice, increasing its integrity and local legitimacy. The reconciliation initiatives described above may only be the very beginning of a successful transitional justice program, but they substantially increase the ability of survivors to set the agenda for subsequent centralised processes, if eventually established. Andrew Iliff is pursuing a joint degree in law and African Studies at Yale University. He has worked on transitional justice in the southern African region with Human Rights Watch, the International Center for Transitional Justice, Idasa and the Research and Advocacy Unit (Zimbabwe). Oxford Transitional Justice Research Working Paper Series 5