The Armenian-Azerbaijani conflict over the region of Nagorny Karabakh has languished in a state of 'no war, no peace' for over a decade since the ceasefire of 1994. The conflict is a central obstacle to the political development of Armenia and Azerbaijan and a key impediment to the development of the South Caucasus region as a whole and its integration into the wider world. It is one of several conflicts between former federal units of the Soviet Union widely framed in terms of a clash between the principles of the self-determination of peoples and the territorial integrity of states. Reference to these principles, understood in absolute terms rather than the more relativistic approach increasingly salient in international practice, continues to dominate the claims of the respective parties to the conflict. Armenia and the Armenians of Nagorny Karabakh insist on sovereign self-determination for the latter, previously an autonomous unit within Soviet Azerbaijan; Azerbaijan insists on territorial integrity within its Soviet-era boundaries, offset by ill-defined autonomy for the Karabakh Armenians.

Overlaying what is fundamentally a territorial dispute are the consequences of the 1991-94 war: a decisive Armenian military victory resulting in Armenian control of Nagorny Karabakh and the further occupation of seven districts surrounding it. Continued occupation or release of these territories forms a key asset to the Armenian side in its attempts to prioritise the determination of Nagorny Karabakh's future status as a precondition for dialogue on other issues. It has thus proved impossible to disentangle negotiations over the consequences of armed conflict from the substantive issues underlying it.

A decade of stagnation has seen international attention to the conflict wane amid exasperation at the lack of progress in the peace process and the more urgent demands of flashpoints elsewhere on the planet. This issue of Accord seeks to refocus attention on the Karabakh peace process and to explore the logic behind the current equilibrium of stalemate. In the first part of the issue the history of the conflict is charted, the competing visions of the parties to it are presented in their own terms and the history of the peace process is documented. Yet beyond the battlefield and the negotiating table, both Armenian and Azerbaijani societies have undergone radical transformations as a result of the conflict. These include the mutual expulsion of Armenians from Azerbaijan and Azeris from Armenia (and thereby the loss of much of both republics' former ethnic diversity), the militarisation of societies and political cultures, a corresponding weakening of democratic impulses, and stunted economic development deriving from blockades and lost investment, exacerbating the already traumatic transition from command to market economies. The articles in the second half of the issue address these broad transformations with which any peace settlement must engage if it is to endure.

One of the first problems encountered in approaching the conflict over Nagorny Karabakh is that the parties to it and external observers construct the conflict, and therefore define its structure, in different ways. De Waal's opening article quickly dispels the popular notion that the conflict can be reduced to incompatible ethnic or religious identities or so-called 'civilisational' differences. As he shows, such explanations neither explain Armenians' and Azerbaijanis' long history of peaceful cohabitation nor the intermittent nature of explosions of violence between them.

The following articles provide insight into the competing visions of the conflict among the parties to it, and how this is reflected in contradictory views of the necessary structure of the peace process aimed at resolving it. Two contributions from the foreign ministers of Armenia and Azerbaijan, Elmar Mammadyarov and Vartan Oskanian, lay out conflicting state perspectives. Azerbaijan's official position defines the conflict as a bilateral interstate conflict between itself and Armenia, an understanding rooted in perceptions of Armenian irredentism, ambitions for territorial expansion and a military invasion of Azerbaijan. In accordance with this view Azerbaijan makes no distinction between Armenia and the Armenians of Nagorny Karabakh; direct negotiations with the de facto authorities in Karabakh are ruled out as a priori legitimating the latter's claim to sovereign statehood. The official Armenian position defines the conflict in terms of a trilateral framework, defining Nagorny Karabakh and Azerbaijan as the key protagonists, with Armenia playing a only a secondary role of interested party and security guarantor for the Armenians of Karabakh. Rejecting claims of irredentism, the Armenian position situates the conflict within the broader discourse of self-determination and decolonisation structuring understandings of the collapse of the Soviet Union as a whole. The Armenians of Nagorny Karabakh support this view and have consistently lobbied for inclusion as an explicitly recognised party to the peace process.

As the following chapters by Baghdasarian and Huseynov show, conflicting conceptual terminology continues in the definition of the respective claims of the Armenian and Azerbaijani populations of Nagorny Karabakh. Informed by hierarchical Soviet understandings of levels of rights inhering in different categories of collective identity, both Karabakh Armenians and Karabakh Azerbaijanis reject the terms 'minority' or 'community' (obshchina in Russian), both claiming Karabakh as their national homeland.

The resilience of the Soviet legacy has at least two implications for outside observers. The first is that that anything less than assiduous use of terminology can inadvertently reflect the bias of one or other side. To illustrate the complexities of the lexicon of the Karabakh conflict and the potential pitfalls in using one or another term this issue of Accord includes a political glossary detailing competing definitions of key terms. The second implication is the thorough discrediting of concepts of autonomy by the experience of dysfunctional federalism under Soviet rule. Secessionist minorities in the South Caucasus regard promises of autonomy with the same lack of seriousness as the leaders of majority groups who make them. Outsiders prescribing autonomy arrangements as solutions to the region's conflicts must take this context into account if they wish their proposals to be taken any more seriously.

Situated at the margins of European, Eurasian and Middle Eastern geopolitical spheres, the Karabakh conflict bewilders the observer in terms of the number of actors involved. The conflict has intersected with and deeply coloured the emergence of a new geopolitical space in the South Caucasus, a region historically and today where great powers compete for influence. Indeed, the competing agendas and unilateral initiatives of international mediators are often held to be a key explanation of the lack of progress in the peace process. Early mediation attempts were initiated by Russia, Kazakhstan and Iran; from 1992 the Karabakh conflict emerged as a key test-case for the validity of the 'New World Order' that many claimed would succeed Cold War geopolitics, and the capacity of the institutions created to administer it: the Conference for Security and Cooperation in Europe, later the organisation for Security and Cooperation in Europe (OSCE), under whose aegis the 'Minsk Group' was given the mandate to mediate in the Karabakh conflict.

Initially debilitated by unfolding events on the battlefield, the resulting OSCE-sponsored process has been oriented towards securing agreement between the sides on both the content and methodology of a settlement. As the chapter by Jacoby shows, however, the process has not vindicated hopes and expectations for greater cooperation among leading powers. Subsequent chapters by Zulfuqarov and Libaridian further provide evidence of how the parties to the conflict continue to fear that peace will institute a new regional hegemony favouring the other side. Their contributions also highlight a central dilemma for many peace processes: the tension inherent in processes managed by states for states in conflict scenarios featuring non-state actors. The de facto authorities in Nagorny Karabakh may constitute a particular kind of non-state actor, one that is highly structured and organised, yet as Libaridian argues, the statist premise underlying international mediation that minorities would relinquish aspirations for independence in return for assurances of stability, economic prosperity and minority rights has proved unfounded. Similarly, as Zulfuqarov shows, Azerbaijan's hopes that a body enshrining the principle of territorial integrity would mediate in its favour have not been realised.

This sense of disappointment should not, however, obscure the underlying point that the current structure of the peace process marginalises precisely those communities that have the most to gain or lose from it: the Karabakh Armenians and the displaced Karabakh Azerbaijani community. Establishing contact between these two constituencies, which must ultimately live in peace in the case of a settlement, must form an integral aspect of future peacemaking efforts, a factor acknowledged in June 2005 by the Azerbaijani Foreign Ministry.

The absence of a clear regional hegemon has also contributed to an idiosyncratic feature of the Karabakh conflict explored in Antonenko's article: the self-regulating nature of the ceasefire regime. The relative stability of the ceasefire indicates that neither Armenia nor Azerbaijan have hitherto been inclined to resume violence. However, as Freizer warns in her contribution, rising ceasefire violations, increased military expenditures and mutual demonisation may reach a tipping point where such an outcome becomes inevitable.

The readiness among some sectors of Armenian and Azerbaijani societies to blame external actors for the lack of progress towards resolution introduces the theme of ownership, and the extent of societal participation in the peace process. As many of the chapters in this issue suggest, only the highest echelons of the political establishments in both countries have been involved in direct contact with the other side: presidents, their aides and foreign ministers. Syndromes of elitism, secretiveness and centralisation inherited from Soviet rule have structured post-Soviet approaches to peacemaking in Armenia and Azerbaijan, as regimes have maintained tight monopolies on the management of the peace process and information about its contents. Autonomous civic initiatives to broaden debate on the conflict ¬ and specifically what can legitimately be said about it in public ¬ are regarded with suspicion, or worse, confronted with charges of 'capitulation'. Elitism in the peace process has forestalled the involvement of wider societies and maintains a huge rift between the rhetoric Armenian and Azerbaijani leaders use to frame the issues for domestic audiences and their positions at the negotiating table. The following series of articles addresses the interface between elites and societies and the impacts of the conflict on different levels of society.

The policing of monopolies on the management of the peace process and contacts with the other side have severely curtailed opportunities for engagement at the civil society level. As Ishkanian and Hasanov show in their article, the number of civic initiatives has dwindled over time and civil society's capacity to broaden the parameters of participation is limited. In Azerbaijan civil society is fragile in the face of a regime enjoying considerable autonomy deriving from its control of oil revenues. In Armenia domestic civil society prioritises other issues over Karabakh, a reflection of the fact that civil society agendas do not necessarily match outsiders' expectations of it as a constituency for conciliation.

Rzayev and Grigoryan chart the role of the media in covering the conflict. Their somewhat pessimistic conclusion is that while the war provided new opportunities for greater autonomy among journalists, the post-war impasse has seen the media in both Armenia and Azerbaijan draw closer to official stances by reproducing militancy and honouring taboos on the subject of potential concessions. In their contribution Baghdasarian and Yunusov chart some of the major social and attitudinal transformations undergone by Armenian and Azerbaijani societies. Their conclusion strongly suggests the paradox of parallel processes in neighbouring, yet mutually isolated societies. Finally Champain considers some of the economic costs of the current stalemate and the prospects for economic development to positively influence the peace process.

Collectively these articles strongly suggest the movement of societies as a whole away from conciliation and towards the internalisation of identities as either victor (Armenia) or victim (Azerbaijan). Mutually exclusive identities constructed since the war pose serious obstacles to a politics of dialogue, yet mirror one another in the narratives they tell. Underpinning these identities are competing understandings of historical justice tightly interwoven with national ideologies. For instance, if some Armenian sources have sought to situate the Armenian-Azerbaijani conflict within a broader narrative of genocide at the hands of 'Turks', official Azerbaijani sources project the problem of refugees and displacement as a solely Azerbaijani problem. There is, moreover, a reciprocal, competitive aspect to such 'symbolic monopolies', as witness attempts in Azerbaijan to construct a narrative of the Karabakh conflict as 'genocide' at the hands of Armenians. On each side there is a tendency to highlight the most extreme instances of violence (for Azeris, the massacre at Khojaly, for Armenians the pogroms of Sumgait and Baku), which are not representative but become remembered as such. Narratives of coexistence and cooperation are lost in this process: as the conflict developed, both communities 'began to remember' historical enmities towards each other, enmities now institutionalised within official narratives reproduced in the media. Plotted along parallel courses, Armenian and Azerbaijani histories do not meet.

If nationalist conflicts defined the 1990s and structured the contours of the re-establishment of independent statehood in the South Caucasus, the current decade is being defined by state-society struggles over the forms that statehood should take. In recent years regimes in both Armenia and Azerbaijan have been confronted with expressions of considerable discontent with the pace and reach of democratisation. Lacking robust mandates, and having promoted a culture of homogenising militancy in part as a result, political elites have left themselves little room for the compromises that any peace settlement must involve. On more than one occasion tentative agreement reached at the negotiating table has foundered when put before domestic audiences. As the articles by Musabayov and Tchilingirian argue, governments in both Azerbaijan and Armenia have sought to manipulate the 'no peace, no war' impasse for internal political gains, rather than to establish preconditions for its resolution. Finally Broers considers the impact of non-recognition on the political development and democratisation process in Nagorny Karabakh itself. In short, throughout the region it seems that only democratic dividends deriving from improved state-society relations may ultimately furnish the necessary resources to break the current impasse in the peace process.

It appears inevitable that any future progress towards peace must involve a widening and deepening of participation and therefore ownership of the peace process. This is possible only in a context where regimes feel more secure in their mandates and no longer feel compelled to reify a hardline stance on the Karabakh issue as a litmus test of legitimacy. Achieving this presupposes a far greater public space for engagement between state and society and the creation of a politics of dialogue within, as well as between, societies. Since its inception, the transformative power of the Karabakh conflict has been demonstrated again and again. The crucial question for the future is how political elites and wider societies in Armenia and Azerbaijan will respond to the potentially far-reaching challenge of transforming politics implicit in the quest for its resolution.