Even before negotiators for the Ugandan government and the Lord’s Resistance Army (LRA) began to gather in Juba for peace talks in June 2006, negotiations were already destined to be controversial. The previous year the International Criminal Court (ICC) had issued arrest warrants for leaders of the LRA, including Joseph Kony, for crimes committed in northern Uganda.

Uganda, the ICC’s first referral, had yielded the Court’s first arrest warrants. In the atmosphere of uncertainty that followed President Salva Kiir’s announcement that his Government of Southern Sudan would mediate between the Ugandan parties, it took Southern Sudan’s firm leadership to bring the parties, and later international observers and financial backers, to Juba for the negotiations.

Perhaps the most significant overt political support to the process came from the United Nations. The intervention of Jan Egeland, then Under-Secretary-General for Humanitarian Affairs, was crucial in this regard. He capped this by making a high-profile visit to Juba in November 2006 to meet with Joseph Kony on the border between Sudan and the Democratic Republic of Congo (DRC). Later, through the appointment of the UN Secretary-General’s Special Envoy for LRA-Affected Areas, Joaquim Chissano, the UN’s role and endorsement of the peace process was strengthened. Mr Chissano brought lessons from his experience of ending the brutal civil war in Mozambique through negotiations.

Although the UN was motivated by humanitarian and broader political concerns, its involvement in the Juba process helped to silence any lingering doubts about the legality of the process. With the initial hesitations overcome, more international actors came to support and participate in the talks. Over the next two years Juba then witnessed the acute dilemmas confronting those who seek to balance the demands of peace and justice in the new era of the ICC.

From the outset, the ICC and the Rome Statute were planted firmly at the heart of the talks. Although other parties adjusted their positions, the LRA, with the most to lose, remained implacably opposed to ICC trials. Against this clear stance the only issue left to discuss was what form national justice processes needed to take.

But not everyone welcomed the prospect of Uganda re-asserting its right to conduct national trials for ICC suspects. Despite the Rome Statute’s principle of complementarity – which requires the ICC to yield to national proceedings – some supporters of international justice opposed the talks, seeing in them a subversion of the ICC.

Uganda had made the first self-referral to the ICC. The prestige and credibility of the Court was therefore closely linked to the fate of the LRA docket. Talk of national proceedings induced nervousness. Although the Prosecutor was initially careful not to make public pronouncements on the Juba talks, his scepticism about the prospects for an accord became more audible as the process unfolded without a final agreement.

At the beginning of the talks the mediation had persuaded the parties to deal with ‘Accountability and Reconciliation’ as the third of five agenda items. This placement of the issue allowed for a gentle build-up towards the negotiations on justice. More crucially it ensured that criminal justice was located in a more appropriate context amongst the political, historical, social and economic justice issues that also needed to be addressed.

Initially the Ugandan government had preferred to focus on the disarmament and reintegration of the LRA. Using the ICC issue as a bargaining chip, it promised to address the ICC issue only upon the LRA’s signature and implementation of the agreement.

For the LRA the prospect of its leaders being paraded before an international court represented a particularly acute form of political humiliation. Behind the ICC’s intervention LRA leaders saw only the hand of the Ugandan government. Pointing to the charges made exclusively against members of the LRA, its leaders detected collusion between the ICC and the government’s political agenda.

Despite the gaps between the parties’ initial positions and the legal complications and uncertainties on the question of justice, the Chief Mediator Dr Riek Machar guided the negotiators to focus on producing agreements that would stand a chance of resolving the ICC issue – the key sticking point. Only a legally sound text that was consistent with the Rome Statute and which balanced justice with reconciliation could rally the kind of support and momentum needed to secure the full implementation of any agreement reached in Juba.

On the other hand, an agreement that was patently unable to address the justice question and which would be seen as shielding perpetrators would be roundly rejected, and was unlikely to lead to the disarmament of the LRA.

To encourage wider national ownership of the agreements and to allow the parties to clarify their own positions in the light of the realities within Uganda, the mediation required the negotiating parties to undertake consultations within Uganda following the initial agreement.

An agreement was required that would provide sufficient detail on the proposed justice measures. This would not only facilitate implementation, but would give as clear an indication as possible to the LRA leaders of what to expect from justice. A comprehensive agreement would equally serve to reassure international stakeholders about the credibility of the justice proposals.

Two main agreements emerged: the first, signed in June 2007, set out the principles of accountability and reconciliation; it was followed by an annexure of mechanisms in February the next year. These texts confirmed that justice and reconciliation were complementary and would be implemented within Uganda, and that accountability would embrace both formal and traditional justice mechanisms.

The LRA’s insistence upon the need to examine the root causes of all violations committed during the conflict regardless of the identity of the perpetrators was adopted. This idea of a truth commission found wide support within Uganda, as the government discovered during its consultations. LRA negotiators also pressed for and secured a commitment to provide individual and collective reparations for conflict losses, another popular idea.

Mindful of the need to deal with the arrest warrants the parties agreed to a special division of the High Court of Uganda which would try the most serious crimes, including those that had been charged by the ICC. The court would be supported by dedicated investigative, prosecutorial and registry functions. In accordance with the principles of the agreement it would apply alternative sentences and its judges would recognise an individual’s acts of reconciliation, as well as cooperation with justice processes. However, most alleged perpetrators would be subjected to other community-based accountability mechanisms, including conditional amnesties.

During the conflict in northern Uganda sections of the affected communities had become disillusioned with the lack of military success against the rebels and began to advocate for an alternative which they thought might break the military impasse. They found the formula in a return to traditional justice.

In 2006 many of these community leaders were invited to Juba as observers to the talks. In reality they were also co-mediators making visits to the LRA leaders in the DRC. Their influence was reflected in the agreement’s commitment to traditional justice mechanisms.

Through this comprehensive and integrated package of measures the parties sought to establish the domestic processes which would take over the conduct of the LRA cases, while also promoting reconciliation and restorative justice values. This dual approach was reflected, amongst other things, in the emphasis placed on reconciliation, the needs of victims and the rehabilitation of offenders.

The justice measures in the Juba agreements did not emerge in a vacuum but in the context of peacemaking. Any justice proposals therefore needed to be seen to contribute to stability. However, Juba could not provide a definitive answer to the arrest warrants immediately, as that issue was ultimately for the ICC judges alone to determine. And the judges could not do so until actual criminal justice measures had been taken with respect to each individual sought by the ICC.

This in-built delay and uncertainty was bound to be troubling for the LRA. Because the LRA leaders, fearful of the arrest warrants, remained away from the Juba table, opportunities for direct engagement were limited. Whether their day-to-day presence would have transformed the outcome is debatable. Certainly more consistent engagement might have promoted greater trust with the LRA leadership but, equally, such close and intense contact could as easily have skewed the dynamics the often delicate and protracted talks to the detriment of a successful negotiation.

Events took a decidedly sinister turn during the latter part of 2007 when tensions within the LRA mounted and reportedly led to Kony ordering the execution of his influential second-in-command Vincent Otti. After this the LRA delegation was almost entirely replaced and encounters with the LRA leadership virtually dried up. On the rare subsequent occasions that there was contact with Joseph Kony justice was not on the agenda.

Despite these disruptions the negotiations moved rapidly, with the remaining LRA negotiators making the best of the situation. Donors’ patience with the process was running out. The LRA delegation managed to arrange a visit to the seat of the ICC in The Hague where they met with representatives of the registry but not the Prosecutor, returning to Juba with a more sober assessment of the fate of the arrest warrants.

Back in Juba with the key mechanisms of justice agreed, LRA negotiators now focused their attention on securing a commitment that Uganda would seek from the UN Security Council a deferral of the LRA cases for 12 months under Article 16 of the Rome Statute. Although it had first resisted this provision, the government delegation now saw that this would insulate it against conflict with the Court over non-compliance with its duty to arrest and surrender individuals to the ICC. The government ensured that the approach to the Council would be made only upon the establishment of the justice institutions in Uganda and, crucially, after the LRA had taken steps to assemble its forces for in readiness for disarmament.

Joseph Kony failed to sign the Juba agreement and instead asked for clarifications about the relationship between the traditional justice and the proposed special division of the Uganda High Court, established in Kampala to deal with war crimes. However, he did not avail himself of the opportunity to discuss these matters with emissaries or his delegation.

In spite of the LRA’s failure to sign the settlement the government declared that it would press on with the implementation of all agreements. With considerable external input, the war crimes division began to prepare legislation to facilitate national prosecutions and other accountability processes.

Whilst laudable, these unilateral moves on the question of justice might, without the benefit of the oversight mechanisms envisaged by the agreement, risk losing the nuances that were designed to make the agreements palatable to the LRA and to affected communities.

Juba showed that the dilemmas posed by the tensions between justice and peacemaking are real and are sharply accentuated when an international court or tribunal intervenes in a situation. Mediators cannot wish away the legal and political complexity of these circumstances, nor can they be ambivalent about the merits of dialogue. They must be prepared to take a firm lead, prioritising a careful search for a workable settlement. In order to fashion viable solutions such negotiations need the support of sound and dispassionate legal advice on international as well as national legal aspects. In Juba, it took the collective efforts of the team of lawyers from the mediation and the two delegations to disentangle the issues.

On the other hand, legal expertise is of itself insufficient to deliver an agreement and should not dominate the dialogue, as this might marginalise key actors. Negotiations in which justice poses a threat to any of the parties require consistent efforts invested in explaining the options and processes to the party that stands to lose the most. This process should not be rushed.

An inclusive negotiation process in which a range of local, national, regional and international stakeholders participates, either as observers or through other interactions with the key parties, will enhance the chances of the final agreement being respected by all sides. For an agreement involving competing notions of justice, the support of a visible constituency for the process is politically critical and consolidates the credibility of the outcome.

By requiring the parties to undertake additional national consultations on the justice issues the mediation hoped to harness wider national input and support.

Such a broader conversation is particularly important because in negotiations that involve balancing the competing interests of peacemaking and justice-seeking, the affected communities are key stakeholders who bear the brunt and risks of war and flawed peacemaking. Giving them genuine opportunities to engage on the issues provides added legitimacy to the negotiation process.