'It has to be a solution that works for the government of Armenia and the government of Azerbaijan, and the people of Armenia and the people of Azerbaijan.'

Carey Cavanaugh, former US co-chair of the Minsk Group, commenting on talks in Key West, Florida, 2001

'To underestimate the position of Karabakh is a major mistake.'

Terhi Hakala, Roving Ambassador of Finland to the South Caucasus

The Conference on Security and Cooperation in Europe (CSCE) began to work on the Nagorny Karabakh conflict in March 1992, soon after newly independent Armenia and Azerbaijan had joined the organisation. This coincided with a unique historical moment in which the iron curtain had fallen and there appeared to be mutual understanding among the CSCE participating states that cooperation was better than confrontation. In this euphoria, it appeared that the Soviet Union’s successor states, especially Russia, could be included in a world system of equals. The CSCE took the first steps to transform itself into the organisation for Security and Cooperation in Europe (OSCE – the name was changed only in December 1994) in an attempt to address issues of common interest in what may now appear a naive spirit of mutual trust and shared values prevailing over narrow national interests.

With its regional remit the CSCE appeared better placed than the United Nations (UN) to deal with the Karabakh conflict, although Armenia and the Karabakh Armenians favoured the UN as a forum for resolution as its historical ‘friends’ France and Russia were members of the Security Council. Azerbaijan, for the same reason, favoured the involvement of the CSCE, of which its biggest ally Turkey was a member. It was also the strong conviction of key state actors that a breakthrough was imminent that led the CSCE to assume responsibility for mediation in the Karabakh conflict.

When the CSCE initiated what was to become the ‘Minsk Process’, it was expected that a conference would be held in Minsk, Belarus, as early as spring 1992 where the details of a peaceful settlement would be determined. The CSCE community believed that only technical details would need to be clarified, the groundwork having been worked out by a preparatory body: the ‘Minsk Group’ of eleven CSCE countries. However, as ever more problems surfaced, the diplomatic preparatory body itself evolved into the forum for negotiations, and the Minsk conference was indefinitely postponed. The participating states accepted this makeshift arrangement without dissent.

By 1994, the CSCE confronted a twofold task: firstly, to mediate, facilitate and support a peaceful settlement of the Nagorny Karabakh conflict, and secondly to negotiate relations between its participating states and determine the role of the CSCE and, specifically, of the Minsk Process within it. Particularly in the initial phases, frictions between key CSCE players complicated both agendas.

Russia has played a dual role as member of the Minsk Group and as a dominant regional actor. Russia has always had its national objectives in its ‘near abroad’ (the fourteen formerly Soviet republics, now independent states), not necessarily shared by other members of the Minsk Group. This contradiction manifested itself in Russia’s competing mediation efforts: it was Russia, and not the Minsk Group, which brokered the May 1994 ceasefire. The other CSCE participating states honoured this, but were reluctant to agree to sending peacekeeping forces. However, all parties to the conflict agreed that multinational troops would be preferable to only Russian ones, and in December 1994 the now OSCE established a High Level Planning Group in Vienna, tasked with preparing the stationing of OSCE peacekeeping forces in the conflict zone.

The US, at least from 1994, developed interests in the region linked to the presence of oil in the Caspian basin and its agenda of diversifying oil production and transportation while circumventing Iran. Tensions between Armenia and Turkey, a Minsk Group member state supporting Azerbaijan, mounted in the aftermath of the Armenian occupation of Kelbajar in March 1993: Turkey declared a blockade on Armenia and admitted to supporting Azerbaijan’s army with military hardware.

In mid-1993, the Swedish Minsk Group chair, responding to conflicts generated by the different national agendas, moved to limit the circle of participants in the peace talks. Minsk Group players seen as less important would be informed but would not take part in the subsequent negotiations. After the ceasefire Russia assumed a role as Minsk Group co-chair with Sweden, and in 1997 a permanent ‘Troika’ of co-chairs, consisting of Russia, the US and France, was formed. This was followed by an active period of shuttle diplomacy to find a resolution, hampered by the fact that Minsk Group co-chairs are less likely to act exclusively as individual mediators than as representatives of their respective states. The only OSCE body in place on the ground is the Personal Representative of the OSCE Chairman-in-Office, a post occupied since 1997 by Ambassador Andrzej Kasprczyk of Poland. His mandate, however, does not include negotiations.

Once more stable working relationships had been established within the Minsk Group its discussions focused on Nagorny Karabakh’s status and security, as well as refugees and internally displaced persons (IDPs) and the problem of the once Azerbaijani-dominated Karabakh town of Shusha. Between 1997 and 2001 four options, representing different methodologies of resolution, were discussed. The first, referred to as the ‘package solution’, favoured talking about all issues, including Karabakh’s final status, simultaneously to achieve the optimum balance. Given the number of issues on the table, this approach would offer more leeway for compromise. The package proposal presented by the co-chairs in May-July 1997 consisted of two agendas: ‘Agreement I’ on ending the conflict, including troop withdrawals, deployment of peacekeepers, return of displaced persons and security guarantees; and ‘Agreement II’ on Karabakh’s final status. The agendas were separate, as the 1997 OSCE Ministerial Council reported, ‘to allow the parties to negotiate and implement each at its own pace, but with a clear understanding that at the end of the day all outstanding issues will have to be resolved.’ Reactions in Baku and Yerevan were encouraging, but Stepanakert rejected it.

The so-called ‘step-by-step’ solution, proposed in September 1997, was premised on sealing Agreement I first before dealing with Agreement II, with the question of the Lachin corridor linking Nagorny Karabakh with Armenia moved to Agreement II. Nagorny Karabakh would continue to exist in its present form until agreement on final status was reached, but would gain internationally recognised ‘interim status’. In principle the step-by-step solution would build a constructive atmosphere in the early stages focused on military aspects, paving the road for negotiations on the more complex political issues.

However, the Karabakh Armenians were not ready to agree to make the first step by withdrawing from the occupied regions of Azerbaijan. Stepanakert argued that this buffer zone was its main source of leverage, which could not be given up without agreement on what concrete security guarantees it would receive in return. Once again, the Karabakh Armenians demonstrated that despite the restricted status accorded to them in the negotiations, Stepanakert wielded significant power of veto over possible settlement options.

The ‘common state’ proposal, presented in November 1998, proposed a vaguely defined common state between Azerbaijan and Nagorny Karabakh, featuring more or less ‘horizontal’ relations between Baku and Stepanakert. It was rejected by Azerbaijan on the grounds of the violation of its territorial integrity and of the principles agreed by the OSCE at its summit in Lisbon, December 1996, where Armenia had been alone in rejecting a statement reiterating principles for a settlement stressing the territorial integrity of Azerbaijan. Finally, President Robert Kocharian of Armenia and President Heydar Aliyev of Azerbaijan discussed a proposal based on an exchange of access to territory in 2001, though this never got as far as an OSCE draft agreement. In the course of the domestic debates launched only after the talks, Aliyev reported (and Kocharian denied) that it had involved Armenia surrendering access to a strip of its southern district of Meghri, offering Azerbaijan direct access to Nakhichevan, in return for accepting Armenian control over the Lachin corridor connecting Karabakh with Armenia.

None of the proposals could bring the sides close to agreement on status by reconciling the needs of self-determination with territorial integrity to the liking of all parties. Being founded on the Helsinki principles (named after the 1975 Helsinki Final Act of the CSCE), the OSCE stands for the inviolability of the frontiers of its participating states. Although the principle of territorial integrity is stipulated with a view to interstate conflicts, how this aspect should be dealt with regarding intrastate conflicts is determined only implicitly. The Final Act speaks of the right of peoples to self-determination “in conformity…with…territorial integrity of States”. This convinces some authors of the OSCE’s inability to be neutral. The Helsinki principles, however, stipulate one important aspect: any decision to alter frontiers must take place ‘by peaceful means and by agreement’. Hence there is no contradiction between accepting the inviolability of frontiers and being neutral at the same time, provided any agreement reached is acceptable to the parties to the conflict.

Azerbaijan perceives the OSCE as an ‘international executor’ that should help it regain at the negotiation table the territorial integrity it lost on the battlefield. This problem has become clearly visible in the discussions around whether Nagorny Karabakh should be accepted as a party to the conflict; at present it is only as an ‘interested party’ with lower negotiating status in the OSCE process than Armenia and Azerbaijan. Within the Minsk Group it was widely believed that Yerevan would have enough influence in Stepanakert to secure the Karabakh Armenians’ compliance with any peace deal reached, thereby obviating the need for their separate and equal representation in the peace process. This has turned out to be a crucial error. On the other hand, elevating Nagorny Karabakh to the status of equal party in the negotiations process is not palatable for Azerbaijan. The Minsk Group so far has not managed to bridge this gap.

The assumption that agreement can be effectively negotiated man-to-man between the presidents of Armenia and Azerbaijan has also proved mistaken. Armenian President Levon Ter-Petrosian was forced to step down in 1998 by his own ministers after publicly advocating making concessions to Azerbaijan. President Heydar Aliyev came under intense pressure in 2001 when he returned home from talks with Kocharian at Key West, and the results achieved in Key West vaporised in the face of domestic criticism. Commenting on what happened in Key West, US Minsk Group co-chair Carey Cavanaugh commended both presidents for being “ahead of their population”. Yet Ter-Petrosian’s resignation and Aliyev’s rapid abandonment of compromise raise fundamental doubts regarding the sustainability of agreements reached by leaders in isolation from their societies. If a strategy to involve the population including the political opposition is absent, peacemaking is likely to fail. The converse argument that it was Ter-Petrosian’s attempt to include and inform the population that led to his downfall is inaccurate. The 1997 press conference where he appeared to attempt to do this was his first press conference in five years. It had been preceded by a long silence and no substantial attempts to get the Armenian population on board his peace project.

The activities of the Minsk Process since its inception have been almost exclusively focused on peacemaking – achieving an agreement rather than a comprehensive solution or a change in attitudes. In contrast, the concept of peacebuilding accepts the need to change attitudes in order to create an atmosphere in which an agreement is feasible as a first step towards a comprehensive solution. Exclusive reliance on political leaders exposes any deals they may strike to the risk of rejection back home. The desirability of complementing peacemaking with peacebuilding is underscored by this need for more communication with the wider societies, without which there can be no sense of public ownership of the peace process. As the Karabakh case shows, no agreement is feasible without popular support.

The problem goes beyond questions of public relations, however, to touch upon fundamental concepts of national identity and interest. For example, among Armenians the differing positions of Stepanakert and Yerevan may in turn each differ from positions originating in the Armenian diaspora. Yet conflict among Armenians is itself a taboo in Armenian political culture, in which attempts are often made to exclude one or other view by labelling it as ‘betraying’ Armenian national ideals. On the Azeri side, there is also a compelling argument for greater internal dialogue among the different stakeholders, and in particular with the Azeri IDP community. A range of conciliation processes within and between all the different social groupings is required before a stable consensus can be reached at the leadership level. Here too there could be a role for the OSCE in supporting forums for conciliatory discussions and in encouraging the parties to the conflict to embark on processes of establishing a societal consensus on what the ‘national interest’ on the Karabakh issue actually is.

The OSCE has so far not taken up this issue nor worked with the leaders to develop such a complementary approach. Though unprecedented for the OSCE, track one diplomacy should be complemented by track two and track three diplomacy undertaken by other actors in an integrated multi-track approach. Obviously, this would mean allowing direct contacts between Armenians and Azerbaijanis. Given the rejection of this idea in Azerbaijan, one possible focus of the Minsk Group could be to convince the parties to agree on the complementary nature of peacemaking and peacebuilding. This would also include accepting direct talks with the de facto authorities in Stepanakert. Closely linked to that could be the role of the Minsk Group in advocating the opening of a direct road link (possibly under international control) across the line of contact, allowing international organisations access to Nagorny Karabakh without violating the de jure border of Azerbaijan.

The OSCE can only be as strong as its participating states allow it to be. Yet antagonisms between the interests of OSCE participating states endure. The OSCE’s experience of mediating in the Karabakh conflict shows there are no grounds to assume that an agglomeration of actors is stronger or more forward-thinking than its individual members. However, mediating in this conflict also poses the dilemma of simultaneous and gradual processes. One process is the Armenian-Azerbaijani peace process, another the development of OSCE capacity within the framework of the conflicting agendas of its participating states, while intra-societal discussions comprise a third. This list is not complete. The interconnectedness of all these processes is evident, yet the conceptual and institutional frameworks at hand to deal with such complex issues are not sufficient. Peace processes elsewhere underscore the importance of third parties maintaining clearly defined roles, and highlight the value of complementary efforts between a range of different state and non-state actors in support of a multi-level peace process. The efforts and constraints of OSCE mediation symbolise the world we live in, and invite us to conceive a completely new, holistic style of politics.