"Apology after 21 years? It is so meaningless." This was the response of Kaushalya Kaur, a survivor of the anti-Sikh riots of 1984, to Prime Minister Manmohan Singh's apology to the Sikh community in August 2005. A month later, the Delhi government's announcement of a Rs 7.14 billion compensation package for the riot victims angered the Sikh community and did little to create a sense that "justice had finally been done".

In November 1984, close to 3,000 Sikhs were massacred and thousands injured and displaced from their homes in Delhi. Twenty-one years later, there have been only eight convictions. While it is public knowledge that many politicians and police officers participated in (or supported) the riots, none of them have been held accountable for the roles they played. In spite of judicial intervention and the several inquiry panels that have been instituted, the legal system has failed to deliver justice to the Sikh community.

Recently, the Gujarat police declared that it will reopen nearly 1,600 cases connected to the widespread violence that took place in 2002. Over 1,000 people (largely Muslims) were killed in the violence. Many Muslim women were raped and tortured by the angry mobs led by Hindu fanatics. The Gujarat government's role in allowing the violence to continue is now well documented. However, after four years, the process of minimum legal justice has just begun.

The low rate of convictions in the Sikh riots or in the Gujarat carnage points to the need for a serious national debate on how justice can be delivered in a timely and efficient manner in contexts of social violence.

The timing of justice measures is very crucial. While apologies and compensation packages might have facilitated some degree of healing in the months following the 1984 riots, their issuance 21 years after the violence does little to heal deep wounds.

The shape and content of a justice process require far greater discussion.
The heavy reliance on legal prosecutions, which currently form the core of approaches to deliver justice, has meant that victims have largely been shut out of the process, with little or no space to articulate what measures would help create a sense among them that "justice has been done".

While the complicity of the State and the low rate of convictions in many riot situations have been the subject of considerable discussion in India, little attention has been given to what may be seen as non-conventional approaches to justice. Over the last decade, experiences of horrific violence in various other regions of conflict have given a spurt to research on approaches that transcend the limitations of the legal system.

It is pertinent to note that much of this research points to the widespread dissatisfaction that victims feel with respect to the ability of the conventional legal system to deliver justice. The retributive justice-oriented legal system has been criticized for deepening the already existing divisions between communities.

Added to this, the material obstacles and the destruction of evidence in the chaos that surrounds violent conflicts, such as riots and pogroms, make it almost impossible to achieve justice through the legal system. The deliverance of justice becomes even more difficult in situations where institutions of the State endorse violence against specific communities.

How then does a justice system make democratically elected leaders accountable for their role in communal violence? This is a question that India confronts today, both in the context of the 1984 anti-Sikh riots and the 2002 Gujarat carnage. The widespread dissatisfaction with the slow wheels of justice should lead to greater introspection on "what justice requires". Countries such as Guatemala, Rwanda, South Africa and East Timor, torn by similar violence, have experimented with diverse approaches, ranging from retribution-oriented criminal justice mechanisms to those that focus on truth telling, reparation and healing. Drawing on insights that transcend cultural variables, India could learn some valuable lessons from the experiences of these countries.

First, a justice system must do more than simply "punish" the perpetrators of the riots. Moving beyond the conventional adversarial legal approach, it should generate processes that facilitate individual and collective healing for the Sikh/Muslim community, that build relationships across the divisions of conflict, and that address the injustices that took place before, during and after the riots.

What the State needs to do is organize more programs for psychosocial healing, inter-community trust-building initiatives, victim-offender dialogues (facilitated by trained mediators where the court system serves as a back-up) and reparation initiatives designed with the participation of the victims.

Second, the perspectives of the survivors of violence should form the core of any process that seeks to deliver justice. Victims, not the State, should play a central role in the conceptualization of "what justice requires".

Third, the experiences of many post-conflict countries have drawn attention to the potential of fields such as transitional justice, distributive justice, and restorative justice to deliver justice that is holistic and healing. They all lay emphasis on the participation of victims in a justice process and on the facilitation of the interdependent processes of healing, coexistence and reconciliation.

Each of these fields also lays special emphasis on truth telling as the foundation for any process that seeks to deliver justice. The emphasis on truth and on human encounter ensures that an important need for victims is addressed. When the truth about the nature of the violence is concealed, it further dis-empowers victims, who feel even more voiceless. Despite the various inquiry panel reports on the 1984 riots, this continues to be the reality for many survivors.

While transitional justice became the subject of extensive research following the setting up of the Truth and Reconciliation Commission in South Africa and the ad hoc tribunals in the former Yugoslavia and Rwanda, restorative justice has been a new entrant to the discourse on social violence. Although it traces its roots to the systems of justice and healing of several indigenous communities, such as the Aborigines (Australia), Maoris (New Zealand) and the First Nation People (North America), there are only a few recent examples (Rwanda and Bougainville - an island in the Pacific, East of Papua) of its application in the context of large-scale social violence. Practitioners and scholars are only now beginning to turn to this field to explore its potential in addressing some of the complex challenges posed by mass violations of human rights.

Howard Zehr defines restorative justice as a process that sees "victims'
needs and rights as central, and encourages offenders to take responsibility for the harm they have caused". Not entirely ruling out incarceration, restorative justice sees healing and reparation as central goals. It includes a wide range of methodologies such as dialogue, acknowledgement of - and accountability for - the harm done, apology, remorse, financial reparation, rehabilitation, reconstruction and psychosocial support and trauma healing.

Distributive justice addresses the important dimension of social and structural injustices that often lie at the root of many conflicts and are exacerbated in many post-violence settings.

The time is now ripe to build a synergy between restorative justice, transitional justice, distributive justice and the conventional, state-based legal system, so that victims of the 1984 riots and Gujarat violence can at last play a central role in a process that was created to deliver justice for them.

(The author works with "Women in Security, Conflict Management and Peace".)