The unresolved conflict over Nagorno-Karabakh continues to be the gravest long-term problem for the South Caucasus region and the whole area between the Black and Caspian Seas. Should the conflict re-ignite, it would spread catastrophe over a wide region, impacting not just Armenia and Azerbaijan, but Georgia, Russia, Turkey, Iran and energy routes across the Caspian Sea.

The conflict, which dates in its modern form to 1988, remains at root a dispute between Armenians and Azerbaijanis over the status of the territory of Nagorno-Karabakh, which in Soviet times was part of the Soviet Republic of Azerbaijan, but which had an Armenian majority. The dispute has been costly: during the fighting, around 20,000 people died and more than a million were displaced on both sides. Since a ceasefire agreement was signed in 1994, a situation of 'no war, no peace' has prevailed. By this time the Armenian side had gained de facto control over a territory comprising not just Nagorno-Karabakh itself but a large swathe of land outside the region as well, amounting to just under 14% of the internationally recognised territory of Azerbaijan. Today, the 'Nagorno-Karabakh Republic' is a de facto Armenian entity with functioning institutions, but is not recognised as a sovereign state by any other country, including Armenia. For everyday purposes it has become a province of Armenia, although the Karabakh Armenians continue to insist on their right to a different view on the conflict and to a role in the negotiations, which currently take place only between Baku and Yerevan.

The two sides have adopted different strategies in their attempts to resolve the dispute. The Armenians have focused on normalising the status quo and building up a de facto state in the hope that history will eventually ratify their victory and that Nagorno-Karabakh will follow the path of Kosovo to international legitimacy. The Azerbaijani side has tried to isolate the unrecognised republic and lobby for international statements of support for its own territorial integrity, while also building up its armed forces in an implicit threat that it reserves the right to use military force to re-conquer the territory.

Meanwhile, international efforts at mediation have been modest, stemming from an unspoken consensus that this is not a high-priority conflict deserving of major international resources. The three ambassadors who co-chair the Minsk Group (charged with finding a peaceful solution to the conflict) engage in shuttle diplomacy between the conflicting sides. The so-called Line of Contact that divides Armenian and Azerbaijani forces is monitored by just six observers, meaning that the ceasefire is maintained essentially through the good will of the two parties and their calculation that it is not in their interest to fight again. Although the Minsk Process has appeared poised to deliver success on several occasions, it seems stuck in a perpetual cycle of frustration and disappointment.

The reasons why the conflict over Nagorno-Karabakh remains unresolved despite the existence of a serious and well-tested negotiation process stem mainly from local dynamics and the calculations of local actors rather than the conduct of the international mediators and the format of the negotiations. Re-designing the negotiating process at this stage could be time-consuming and needlessly mire all sides in protracted bargaining over process rather than substance. Nonetheless, negotiations are clearly deadlocked six years after the latest phase of talks, the so-called Prague Process, was initiated, and desperately need reinvigorating. Cynicism about the peace process has reached dangerous levels and is leading the parties down a path which threatens to end in renewed violence rather than peace. The Minsk Group is the international community's instrument of influence on this strategically important dispute and it is worth asking how it can be made to work more effectively, broadened and supplemented by other initiatives so as to attain the long-desired result.

The curious history of the Minsk Group 

It is reasonable to assume that the participants in the Conference for Security and Cooperation in Europe (CSCE) meeting held in Helsinki on 24 March 1992 never imagined that the mechanism they helped to invent that day would still be in existence 18 years later, and still trying to resolve the same conflict. The idea of a 'Minsk Conference' came about because the delegate from newly independent Belarus, then a would-be democratic state aspiring to integration with the new Europe, offered his capital as the venue for a peace conference on the Armenian-Azerbaijani conflict, which was then escalating into full-scale warfare. Because of an upsurge in fighting - in May 1992 Armenian forces captured the towns of Shusha and Lachin - the planned conference was called off, but because it was never formally cancelled international documents on the Karabakh peace process still refer to the 'Minsk Conference'. The first CSCE negotiations eventually took place in the more popular venue of Rome, at which a 'Minsk Group' was created. Belarus has had nothing further to do with the negotiations and the most substantial peace conference to date on Nagorono-Karabakh was held in Key West, Florida, in 2001, but the name 'Minsk' lingers on in ghostly fashion.

It took a long time for the Minsk Group to work effectively and harmoniously. The 1994 ceasefire was negotiated by Russia, working independently of, and sometimes in deadly rivalry with, its Minsk Group partners. Several times, meetings were convened by either Russia or Western countries without one side properly consulting the other. It wasn't until December 1994 that the Minsk Group finally achieved some kind of institutional identity at the Budapest Summit, which transformed the CSCE into the Organisation for Security and Cooperation in Europe (OSCE). Plans were made to raise the OSCE's first-ever peacekeeping force, to be dispatched to the Armenia-Azerbaijan conflict zone. In early 1995 a formal co-chairmanship of the Minsk Group (then held by Russia and Sweden) was established for the first time, which soon came to overshadow the overall group. The Minsk Group still formally exists and today consists of around a dozen countries, but has no formal mandate and plays at best a supporting role.

The mandate of the co-chairmanship itself was set out in March 1995. This now covers the activities of France, Russia and the United States, which have been the three co-chairs of the Minsk Group since February 1997. Interestingly, the full version of the mandate, OSCE DOC 525/95, is not published on the OSCE website, which limits itself to a cursory summary, and is hard to find. It commits the co-chairs to more activities than they are currently engaged in, mandating them specifically to carry out confidence-building measures, help develop plans for a peacekeeping force and maintain contacts with other international organisations.

Thus, it would be fair to say that the Minsk Process consists of a conference which was only occasionally convened, a group which never meets as a group and a co-chairmanship functioning under a barely known mandate, all named after a city where the mediators never meet. This Alice-in-Wonderland backstory means that Ambassador Andrzej Kasprzyk, the OSCE's roving regional official on the ground, has the formal title of 'Personal Representative of the Chairman-in-Office on the Conflict Dealt with by the OSCE Minsk Conference'. While in some ways these structures actually work quite well, they were designed in a different historical era and sometimes look antiquated.

The fact that the Minsk Process was initiated when fighting was still going on and has remained largely unchanged ever since raises the question of whether this is a genuine peace process designed to devise a long-lasting solution to the conflict or simply a conflict-management mechanism, devised in war and updated since, whose primary object is to maintain a ceasefire and prevent a new outbreak of hostilities. Certain peculiarities of the Minsk process, compared to other conflict-resolution efforts around the world, suggest that the political leaders of Armenia and Azerbaijan actually prefer this small and awkward mechanism to a broader and stronger mediation process, which might be better equipped to deliver results but would be less under their control.

Perhaps the strangest feature of the process is that it has no spokesman and virtually no public profile. This means that progress in the negotiations remains confidential, while most media coverage in the region of the Minsk Group is negative. Almost no one speaks up for peace. The leaders of Armenia and Azerbaijan exercise nearcomplete control over the substance of the peace process and the way in which it is perceived by both domestic and international audiences. The presidents want to be in charge not just of the overall decision-making process, which is understandable, but of every detail of the discussions. They are the joint conductors of the process, which proceeds at the tempo they dictate. This allows them the political luxury of doing everything via the Minsk Group co-chairs and, when negotiations go badly, blaming France, Russia and the United States for not doing enough to resolve their conflict. (Recently, Turkish Prime Minister Recep Tayyip Erdogan has joined in this blame game against the Minsk Group, saying in February 2010, 'if Russia, the US and France had worked hard within the past 20 years, none of these problems would have emerged'.1). At the same time, state-controlled television in both Armenia and Azerbaijan, overwhelmingly the main source of news for the populations in both countries, continues to transmit official patriotic messages about either a heroic Armenian victory which will eventually be confirmed by history or the tragic occupation of Azerbaijani land which will one day be reversed. The average viewer barely has an inkling that his country's leader is busy negotiating a document that will make substantial concessions to the other side and, because of the tight cloak of secrecy surrounding the talks and strict media controls, the presidents themselves have the freedom to pass over the subject in public statements.

If the two presidents were more committed to a result, rather than to process for the sake of process, they would set up a permanent channel for bilateral contact and discussion of the conflict rather than relying entirely on the Minsk Group co-chairs to be their go-betweens. There was briefly such a channel between trusted presidential advisers Gerard Libaridian (Armenia) and Vafa Guluzade (Azerbaijan), but it was discontinued after the OSCE Lisbon Summit of 1996. In 2004 two deputy foreign ministers, Azerbaijan's Araz Azimov and Armenia's Tatul Markarian, were entrusted with a bilateral mechanism, but they later transferred responsibility to their foreign ministers. The ministers, in turn, initiated the Prague Process in 2004, but then gave way to the two presidents. Nowadays, the two leaders are rarely in direct touch with each other, even though the Karabakh conflict is the main long-term issue for both countries.

This lack of direct contact puts the onus of mediation on the three-headed Minsk Group co-chairmanship. In practical terms, it is a cumbersome institution. The current mechanism requires a perpetual feat of coordination between Moscow, Paris, Washington, Baku and Yerevan to allow all the relevant officials to be in the same place at the same time. Although for the past decade the mediators have always reported a high degree of harmonious coordination, the current set-up does not make for quick and nimble diplomacy. Meetings are spaced far apart and it is hard to maintain momentum. A single mediator would be able to proceed faster and develop ideas with more agility. It is noteworthy that the only successful mediation effort in the conflict, the 1994 ceasefire, was in fact negotiated by one person, Russian envoy Vladimir Kazimirov, in long and tense but ultimately productive one-on-one discussions with the key actors.

The tri-partite co-chairmanship structure emerged because of the perceived need to balance the competing interests and suspicions of Russia and the Western Minsk Group nations. Today the relationship is much more harmonious, and it is at least worth asking if a single mediator might not be more effective. Moscow, Paris and Washington have a great deal of institutional knowledge invested in this process and will clearly have a major role to play in making a post-conflict settlement work. But they also have bilateral energy, security and trade agendas with Armenia and Azerbaijan that make them political players in the South Caucasus, not just impartial negotiators. Powerful Armenian lobbies in France, Russia and the United States also make relations with Yerevan a matter of domestic politics and not just foreign policy in each country. Perhaps it is worth considering if these three countries could not therefore become the guarantors of a peace agreement, rather than its direct negotiators, and leave the 'coal-face' work of mediation to another Minsk Group country with less of a stake in the region.

If the co-chairmanship is 'one part of the forest', as Shakespeare might say, a bigger problem with the Minsk Process is that no one is even visiting large parts of the forest. At the centre of the Nagorno-Karabakh peace process is a three-page text of 14 or 15 points called the Document on Basic Principles. This text tries in creative fashion to square the circle of the disputed status of Nagorno-Karabakh. Negotiations surrounding this small document have dominated the peace process for more than two years and increasingly resemble the debates of a group of cloistered theologians over an obscure doctrinal text. Yet, vital as this political agreement is, it will be only one part of a peace agreement for the region. A successful peace process will need to involve more than just the two presidents and a handful of negotiators carrying on discussions in private. It will need the support of the Armenian and Azerbaijani populations, including Karabakhis. It will also need international resources in terms of security assistance and economic reconstruction.

The ceasefire and security issues

The Nagorno-Karabakh dispute is not a 'frozen conflict'. Every year several dozen soldiers die in shooting incidents between the Armenian and Azerbaijani armies. As recently as 18-19 June 2010, in one of the worst incidents of this nature, four Armenian soldiers and one Azerbaijani were killed in a clash on the Line of Contact.

A primary function of the OSCE peace process, little noticed outside the region, is to monitor the ceasefire along this line. But this monitoring mandate is extremely limited, and it could be argued that the OSCE presence is basically a fig leaf. Ambassador Kasprzyk and five assistants monitor a ceasefire in which in excess of 20,000 troops on either side face one another across a 175-kilometre line of trenches and dug-outs. Clearly, the monitors are powerless to prevent either side taking military action if it so desires. Furthermore, monitors must give advance warning of several days if they wish to visit a section of the front line. The monitors then gather on either side of the line and file a report. This means that monitoring is less about enforcing the ceasefire than about providing a means of communication between the Armenian and Azerbaijani military commanders and a way for the mediators to keep their finger on the pulse of the two armies and the situation on the front line. That is significant, but should not be confused with a genuine war-prevention mechanism.

The ceasefire was originally signed on 27 July 1994 by three military chiefs - then Armenian Minister of Defence Serzh Sarkisian, his Azerbaijani counterpart Mamedrafi Mamedov, and the 'commander of the army of Nagorno-Karabakh' Samvel Babayan - who put their names to a document which for the first time committed the warring sides to build on the immediate post-war truce and move towards a lasting peace and an international security mission. The three men confirmed their 'desire to intensify the conclusion of a Big Political Agreement in the course of 30 days from August 1994 in which there will be set out military-technical issues, including the interaction of international peacekeeping forces and the observer mission of the CSCE'.2

Of course this did not happen. The two sides in the conflict had a common interest in signing a ceasefire that halted the fighting. Beyond that, their goals were very different. For the Armenian side it was important that the new ceasefire line become as fixed as possible, allowing Armenia to consolidate its conquests in its new 'security zone' around Nagorno-Karabakh. Yerevan therefore broadly welcomed any initiatives for a peacekeeping force and Moscow's efforts to send peacekeepers to the region, as it had done in Abkhazia and South Ossetia. The Azerbaijanis, by contrast, wished to keep the ceasefire line, which runs across what is their de jure territory, as impermanent as possible. They therefore blocked the deployment of a Russian peacekeeping force in 1994 and have consistently opposed any moves to strengthen the OSCE monitoring mandate. Baku regards the ceasefire line as one of its few pressure points against the Armenians and likes to repeat that it is still basically in a state of war. This was the rationale for Azerbaijan's rejection of a call by the OSCE at Helsinki in December 2008 to remove snipers from the vicinity of the Line of Contact.

The situation on the line is a barometer for the health of the peace negotiations as a whole. In 2009, when peace talks were intensive and frequent, things were relatively quiet on the line, and there were only 19 casualties. In the first half of 2010, however, casualty figures have already matched that, and peace talks are deadlocked.3 Although the risk of a resumption of full-scale war across the Line of Contact is small, a more plausible threat is that one side - almost certainly the Azerbaijani side, given the balance of power - will seek to upset the status quo by using heavier weapons, such as mortars, causing greater casualties to its opponent. Should either side choose to do this, there is very little that the OSCE could do. This is one reason why strengthening the ceasefire mandate should be a top priority. Such a move would face resistance in Azerbaijan, but a quid pro quo might be that the monitors would also have a mandate to report on activities inside the Armenian-held territories outside Nagorno-Karabakh.

The fact that plans for the overall security arrangements following any agreement are still extremely sketchy is another cause for concern. The lack of any credible proposals is weakening the peace process, because it makes the two presidents even more cautious about signing an agreement without knowing what kind of security arrangements will be put in place to underpin it. The OSCE High-Level Planning Group, resident in Vienna and consisting of six military officers, was intended to devise the eventual peacekeeping mission for the conflict, but very rarely visits the region and has little or no contact with the Minsk Process negotiations.

Both Armenians and Azerbaijanis have reservations about the potential composition of a peacekeeping or policing operation in the region. When asked for their opinion on the issue, Karabakh Armenians routinely say that they trust no one but themselves to look after their security. They also mention the UN mission in Srebrenica as an example of a force that was designed to protect a civilian population but failed to do so. Azerbaijanis are similarly wary about allowing international forces on their territory and delegating their security to others. Azerbaijan is likely to veto any Russian involvement in a peacekeeping force, while Moscow almost certainly sees Russian peacekeepers as a way of maintaining its influence in a post-conflict region. Given these sensitivities, it is troubling that, despite the existence of a structure specifically mandated to deal with this issue, there is little sign that anyone involved in the Karabakh peace process is working on either the design or the politics of a peacekeeping force. For the negotiations to have real teeth, a substantial conversation needs to take place; and Russia, the United States and the European Union need to decide who is prepared to commit what and on what terms to the post-war security force for Karabakh.

A weak Track Two process

A peace process is more than two signatures on a piece of paper. If there is no broad public support for peace, any agreement is almost bound to fail. Those shut out of the process or who feel they have no stake in its success are more likely to become spoilers. Yet the negotiations over Nagorno-Karabakh must rank as some of the most secretive and least inclusive peace talks in the world. The mediators in the Track One process give no support to a complementary Track Two process and are often actively hostile to other peace practitioners.

There are several reasons why there is no dynamic Track Two process on Karabakh. In the first place, both parties to the conflict are still living with the black-and-white attitudes of the late 1980s and early 1990s, which cast the other side as the enemy. For an Armenian or Azerbaijani to engage in grassroots contact across the Line of Contact takes an effort of will in defiance of the national consensus. Moreover, in a region that is still getting over its Soviet legacy, independent person-to-person contact remains a novelty, and is something authorities often seek to discourage. The semi-authoritarian governments in both countries are suspicious of the Western-funded civil-society activists who have attempted to promote dialogue. The same activists also tend to criticise the governments for their record on democracy and human rights, and are perceived as being stooges or agents of Western interests. Thus, the highest-profile initiative the two governments have supported outside the official Minsk Process has in fact been semi-official: two visits by the Armenian and Azerbaijani ambassadors to Moscow and intellectuals to Baku, Karabakh and Yerevan.

There are also important differences on either side of the conflict that work to prevent a Track Two process from developing. The Armenians, as the 'winning' side in the 1991-94 conflict, can afford to take a more magnanimous stance on the issue of people-to-people contacts. To put it another way, grassroots contacts do not contradict the official government ideology that the conflict has been resolved and victory won. So Azerbaijani experts and journalists, albeit in small numbers, are invited to visit Armenia and even Nagorno-Karabakh from time to time. The Azerbaijani authorities are much more suspicious of these contacts, worrying that they legitimise what they regard as the unacceptable status quo in which Armenian forces remain in control of Azerbaijani land. There have been occasions when contacts have been authorised, as when, around a decade ago, several delegations of Armenian experts and journalists visited Baku. More recently, however, the Azerbaijani government has been much more negative about civil-society dialogue initiatives, and Azerbaijani activists who have met with Armenians have been harassed or arrested. Thus, while civil-society activities supported by NGOs such as Conciliation Resources, International Alert and Pax Christi are undoubtedly helpful, in this difficult environment they have not catalysed broader-based grassroots activism.

This picture presents an interesting contrast with that of Armenian-Turkish relations. Given the history between these countries (in 1915, more than one million Armenians and hundreds of thousands of Turkish Muslims died in what is today known as the Great Catastrophe or Armenian Genocide), Armenians and Turks arguably have much more to fear from one another than do Armenians and Azerbaijanis. Yet the passage of time and Turkey's more democratic society have made people-to-people contacts much more common. Even if the current Armenian-Turkish Protocols process fails, Armenians and Turks will remain connected by a thick web of civil-society interaction.

It is not the job of international mediators formally to arrange civil-society dialogue, still less to stimulate the kind of debate inside Armenia, Azerbaijan and Nagorno-Karabakh that is a prerequisite for society 'buying in' to a peace process. But the mediators can and should demonstrate a greater willingness to interact with and support civil-society initiatives. One part of Kasprzyk's mandate is to 'assist the parties in implementing and developing confidence-building, humanitarian and other measures facilitating the peace process, in particular by encouraging direct contacts'. 4 Ten years ago, with a more tolerant approach from Baku, he could do more of this than he has been able to recently. Yet more could be done even now if there were a stronger show of support from the OSCE for parallel initiatives. Moreover, for immediate pragmatic reasons, the co-chairs need to engage in a better dialogue with citizens on both sides of the conflict to counter conspiracy theories that the negotiations are some kind of international plot to impose a settlement on unsuspecting Armenians or Azerbaijanis.

Civil-society initiatives are also important because they include the chief missing element in the Minsk Group talks, the Karabakhis. Another strange anomaly of the Minsk Process is that, although the negotiations are about Nagorno-Karabakh, they take place between Baku and Yerevan without any direct participation by Karabakhis themselves. Much of the wrangling over the years has been over how many sides there are to the conflict. The current situation is that Karabakh Armenians are not directly represented at the negotiating table. The same is true of Karabakh Azerbaijanis, who constitute a minority community in Karabakh and do not have an elected leadership or formal structures in the way that the Karabakh Armenians do, but who still deserve a (rather quieter) voice in the negotiations. (Karabakh Armenians understandably resist efforts to label them a 'community' equivalent to the Karabakh Azerbaijanis, but their efforts to reserve the term 'Karabakhis' for themselves amount to an unhelpful attempt to erase the Karabakh Azerbaijanis from history.) Unfortunately, the question of Karabakhi participation has become an issue for negotiation in itself. Everyone knows that the Karabakh Armenians at least should have a say in the peace process that concerns their very existence, but the issue of how to include them is much harder to solve than it was in 1994.

Earlier, the Karabakh Armenians did have a place in the negotiations. The original 1992 text calling for a Minsk Conference spoke of inviting 'elected' and 'other' representatives from Karabakh, which covered both Armenians and Azerbaijanis from the territory.5 (This raises the interesting question of why, despite referring to elected officials from Nagorno-Karabakh, the international community still routinely refuses to recognise elections there, in contrast to, for example, its more pragmatic approach to elections in Northern Cyprus.) During the war, there were direct contacts between Azerbaijani and Karabakh Armenian officials and, as we have seen, Karabakhi military commander Samvel Babayan signed both the ceasefire document in May 1994 and the subsequent document reaffirming the truce. Karabakh Armenians, along with Karabakh Azerbaijanis, attended talks for two years after the war. This involved much wrangling about language - whether, for instance, a declaration should use the word 'among' or 'between', because 'among' implied there were three parties to the conflict while 'between' implied there were two - but at least the Karabakh Armenians, who are central to the dispute, were at the table. This changed after the December 1996 OSCE Lisbon Summit, when all the OSCE states, barring Armenia, signed a declaration affirming Azerbaijan's territorial integrity, and Baku toughened its stance. Armenia's first president Levon Ter-Petrossian did try to keep the Karabakh Armenians inside the process, but clashed with the Karabakh Armenian leadership and was forced to step down when he tried to press ahead with a plan they opposed. One of the chief plotters against him was Robert Kocharian, who became president in his place. He had been the war leader of Nagorno-Karabakh and then its elected 'president', and had calculated that he could negotiate on behalf of both Yerevan and Stepanakert. Thus, it was a Karabakh Armenian who formally helped shut Karabakhi residents out of the process.

Clearly, there are many issues on which Yerevan and the Karabakh Armenians hold identical positions. For example, the armies of both territories function as one unit and all military issues are decided jointly. But as long as they are shut out of a broader process, Karabakh Armenians will be fearful and regard the peace process as something directed against them rather than for them. Baku is concerned that if it allows the Karabakh Armenians into the process they will use it as a propaganda victory, and negotiations will become a case of 'two against one'. Involving the Karabakhis in a Track Two or 'One and a Half' process would be a good way of easing them back into formal negotiations.

Broader international engagement

The early phase of mediation efforts over Nagorno-Karabakh in 1991-92 suffered from the phenomenon of 'forum shopping' in which the conflicting parties could choose among different mediators, including Russia, Kazakhstan, Iran and the United States. The involvement of the CSCE from March 1992 began to regularise the process, although Russian envoy Kazimirov operated in a state of suspicious rivalry with his Western counterparts. From 1997 the co-chairmanship of France, Russia and the United States, all of which happen to be permanent members of the UN Security Council, gave the OSCE Minsk Group undisputed authority. There was no question of the United Nations taking a mediating role; Azerbaijan has occasionally sent out feelers, but eventually it too came to the conclusion that it prefers the Minsk Group co-chairs to be the mediators.

Today, the co-chairmen risk becoming, instead of the leading international players in this conflict, the only international actors. The mandate of the Minsk Group co-chairmanship charges members to 'maintain necessary contacts with the ICRC [International Committee of the Red Cross], UNHCR [United Nations High Commissioner for Refugees], and other relevant international and regional organizations and institutions', but so far, almost no other international organisation has played a role in the peace process, with the exception of the Red Cross, which occasionally deals in prisoner exchanges.

It could be argued that there is not much for other international organisations to do. There is little movement in the peace process and the OSCE can take care of what is going on. But this interpretation overlooks the untapped potential for constructive involvement of numerous relevant organisations, not to mention states. A whole range of international actors who are not formally involved in the peace process, from the EU to Turkey to Iran to the US Congress to BP, have a stake in a peaceful outcome of the Karabakh dispute. But as they are at one remove from the Minsk Group - some of them are not even briefed on its discussions - they contribute nothing to the peace process. Most of them do not even bring up the subject, or do so only in the most pro forma, perfunctory way, in bilateral discussions with Baku and Yerevan. (Of course, they never travel to Nagorno-Karabakh itself.) A vicious circle is created in which actors like these believe they know nothing and have nothing to contribute, and therefore are not able to exert their influence over local actors. One reason why Turkey miscalculated in its stalled rapprochement with Armenia was that it mistakenly believed that progress on the Armenia-Azerbaijan dispute was more advanced than it actually was. Had Ankara had a better understanding of the state of negotiations, it might have acted differently.

A second group of actors, including the UN agencies, the World Bank and other development agencies, have expertise to bring to a post-conflict settlement. The needs here are vast. The seven Azerbaijani regions outside Nagorno-Karabakh that are wholly or partially controlled by Armenians are completely devastated, stripped of even their most basic infrastructure, and littered with mines. It will take years of work following any peace agreement before they are fit for habitation by returning refugees. There is an expectation that these agencies, with the skills to do this work, will be asked to contribute once a political agreement is signed. But their non-involvement at this stage contributes to the impression of drift in the negotiations. Why should President Ilham Aliev of Azerbaijan or President Serzh Sarkisian of Armenia put his signature to the Document on Basic Principles today when neither knows what kind of international effort will begin tomorrow to underpin the peace process? In 2001, an international donors' conference was held in Geneva prior to the Key West peace talks in Florida, but so far no plans have been made for a follow-up meeting. It would be too ambitious to plan such a conference now, when talks are deadlocked, but agencies could at least start to make contingency plans for their later involvement as a show of their serious commitment to a post-war settlement.

The most glaring absence from the Karabakh peace process in this regard is the European Union. The EU has vast resources and expertise it could bring to bear, as demonstrated by its successful stabilisation and reconstruction efforts in the Balkans. The EU has expressed interest in devoting more resources to Nagorno-Karabakh, and EU Special Representative for the South Caucasus Peter Semneby has tried to secure a greater role for Brussels in the peace process. The default position of the Minsk Group, however, continues to be that the EU has sub-contracted its role to France, even though France plays a very different role in the process, that of conducting high diplomacy. It makes sense for the EU, in coordination with the UN and other agencies, to begin planning now for a post-conflict reconstruction programme for Karabakh and the surrounding areas. Yet to date it has not found a useful role in this conflict on its borders.

The Nagorno-Karabakh conflict remains a low priority internationally because of a consensus that there is no immediate danger of war and no prospect that greatly enhanced international pressure will force the local leaders to become more amenable to peace. Those assumptions need to be constantly revisited, particularly given the example of the August 2008 war over South Ossetia. If it is to be more than just a conflict-management exercise, the Minsk Process needs an injection of new life and ideas. The local leaders are understandably cautious about taking bold moves that could shift the ground under their feet and usher in a new era of concessions on territory and sovereignty in the name of long-term peace. For them, a wrong move could entail losing their job, or worse. The co-chairs of the Minsk Group have done good work on trying to bridge differences on difficult issues, but do not have enough sticks or carrots to persuade their interlocutors to change their calculations and 'go the extra mile' in accepting a peace agreement.

The relationship between the Minsk Group co-chairs and the Armenian and Azerbaijani presidents has become too close. The co-chairs grind away at the almost thankless job of mediating between two very suspicious leaders, who reject almost all proposals for constructive engagement with the other side. But the presidents themselves bear no cost for their lack of positive thinking and the process drags on as before, often more pretence than reality. The mediators should consider being more honest about how difficult their task is. In July 2007, the co-chairs did release an unusually candid statement in which they said their capacities were limited and 'if the parties to the conflict choose … to avoid making courageous decisions now, we as mediators cannot make these decisions in their place'.6 But not long afterwards they distanced themselves from the implications of that courageous statement and continued to engage as before in secret shuttle diplomacy in pursuit of elusive signatures on the Basic Principles Document.

If, as seems likely, it is too difficult for the co-chairs to bridge the differences between the two sides on the Document of Basic Principles, it makes sense for them to broaden the peace process and work on creating an enabling environment, both internationally and domestically, which will make it easier for the political leaders to close the gap at a later stage. That means working on three aspects of the peace process that are receiving far too little attention: the ceasefire and plan for an international security force; the Track Two process; and the role of international organisations and other interested actors. This can be done without abandoning the Minsk Group format and while keeping the co-chairmanship as the prime negotiating mechanism. Now is a good time to engage in a re-think. To stick with the current Minsk Process in all its details, despite the negative omens, runs the risk of discrediting all the good work that has been done so far. The greatest danger is that the international community only wakes up to the importance of the Karabakh conflict when it is already sliding back into full-scale war.


An earlier version of this paper was presented at the Association for the Study of Nationalities conference in New York in April 2010. I thank a number of people involved in the Minsk Process, most of whom prefer to remain anonymous, for their useful comments on the paper.


1 'Erdogan Vows to Abolish EMASYA Protocol, Revise Security Priorities', Today's Zaman, 1 February 2010,

2 The agreement can be found in Tatul Hakopyan, Zelyonoe I Chernoe [Green and Black] (Yerevan: Anatares, 2010), pp. 475-6.

3 Casualty figures obtained in author conversations with OSCE officials.

4 OSCE, 'CiO Representative on Minsk Conference',

5 'Helsinki Additional Meeting of the CSCE Council: Summary of Conclusions', Helsinki, 24 March 1992, available at

6 Yury Merzlyakov, Bernard Fassier and Matthew Bryza, 'Statement by Co-Chairmen of the OSCE Minsk Group', Vienna, 13 July 2007,


This article was originally published in Survival, Volume 52, Issue 4 August 2010, pages 159 - 176.