Mediation Attempts into Nagorno Karabakh Conflict: A Case Study


Untitled Document

Executive Summary

The case study analyses the mediation attempts into Nagorno Karabakh conflict between Armenia and Azerbaijan.  The paradigmatic model consisting of three models – Rational Actor, Organizational, and Governmental Politics - borrowed from Allison (1971), is utilized to analyze the mediation attempts from three different viewpoints.  Two periods of conflict are analyzed – 1988-1991, when Nagorno-Karabakh conflict was an internal conflict of the Soviet Union, and 1992-1998, when, immediately after the collapse of the Soviet Union, the conflict acquired an inter-state character.   During the first period, the ultimate source of mediation were the Soviet authorities, while during the second period, different actors – Russia, OSCE, Iran – attempted to play a mediating role.  The application of Allison’s paradigmatic model to both periods allows for making several judgments and recommendations for further action: first, one of the reasons for no tangible success of mediation attempts was the fact that each mediating party had an agenda specific to it, and only to it.  Second, civil society, which has the advantage over the states and international organizations of not having any geopolitical agenda and being much more flexible organizationally, did not play any significant mediating role during the conflict.  The recommendations this study proposes are that in order to a mediating attempt to achieve its goals, (i) a mediating parties should incorporate the interests of the conflicting parties to its own geopolitical agenda, (ii) that each mediating party should ensure at least as much organizational dynamism and variety, as the conflict itself, (iii) that the mediating party should parcel out the problem of mediation to as few of its constituent agents, as possible, and (iv) civil society should exploit its advantages over the states and international organizations and become actively engaged in conflict resolution.
Keywords: Nagorno-Karabakh conflict, mediation attempts, rational actor model, organizational model, governmental politics model
...blind men lead equally blind
and cause them to plunge
into the abyss
The Name of the Rose
Umberto Eco


Two decades of confrontation, dozens of thousands of victims, and three presidential resignations – these are just a few of characteristic features of the conflict around Nagorno-Karabakh – the first, and most decisive breach on the monolith of the Soviet Empire, which is still smoldering in South Caucasus.  The conflict, which emerged as an internal Soviet Union conflict, led to a de-facto war between Armenia and Azerbaijan, which lasted for almost two years; and both before the war, during it, and after the ceasefire established in 1994, continuous mediation attempts were initiated by a number of parties – unfortunately, with no tangible success.  Today Armenia and Azerbaijan seem to be (at least for a detached observer) no more closer to its peaceful resolution, then a decade ago.  

Hence, it is never too early to reflect back and to analyze the reasons of the failure (or, one should say – lack of success) of mediation attempts to understand the underlying mechanisms responsible for their performance; on this basis, on might attempt to extrapolate recommendations for the further improvement of mediation attempts.  This is why this research into Nagorno-Karabakh conflict mediation attempts is a case study, in the sense that it examines one unique practice-based problem, where the importance of both the experiences of actors and the contexts of action are critical (Bonoma, 1983). 

In order to better understand and present valid recommendations, one needs to scrutinize in an informed way – not merely closely, but also from different viewpoints.  This article will attempt to analyze the mediation attempts into Nagorno-Karabakh conflict, using a scrutinizing device, tested to examine another classical case study – Graham Allison’s insight into the Cuban Missile Crisis (Allison, 1971).  Allison proposed and successfully utilized a three-lens approach, consisting of the Rational Actor, Organizational, and Governmental Politics paradigms, thus explicating numerous critical, yet implicit mechanisms of decision making in a crisis situation. 

This research will utilize this already tested approach within a new application domain – mediation attempts into the Nagorno-Karabakh conflict resolution.  The three paradigms will be applied to mediation attempts in two periods: first is 1988-1991, when the Nagorno-Karabakh conflict was an internal Soviet Union conflict, which limited the sources and structure of mediation into internal Soviet Union actors, and the second is 1992-1998, when, after the collapse of the Soviet Union, Karabakh conflict transformed into a de-facto war between Armenia and Azerbaijan, followed by a ceasefire, and the mediation attempts, now international by definition, led to political shakeups in both countries.  This study covers one decade when there was a diversity of mediation attempts, which allows comparative analysis based inference building.  The period after 1998 is marked with one dominant mediation format, which does not allow for comparisons; this should be a matter of a separate study, with appropriately adjusted research methodology.

The next section presents a brief of Allison’s (1971) three models.  Then the models are sequentially applied to the two periods of mediation attempts – 1988-1991 and 1992-1998.  The concluding section discusses the findings of the analysis and makes some recommendations. 

The Conceptual Framework

As already mentioned above, methodologically this inquiry into the mediation attempts in Nagorno-Karabakh conflict is a case study, for this research strategy is particularly appropriate when researching such practice-based problems as mediation in conflicts and negotiations, where the importance of both the experiences of actors and the contexts of action are critical (Bonoma, 1983).  Furthermore, the case study research in researching mediation and negotiations have been mentioned elsewhere (Yin, 1984) as an exemplar of how this approach can be utilized to explain phenomena; the particular example quoted is Graham Allison’s Essence of Decision (1971), which analyses the Cuban Missile Crisis of 1962 through the lenses of three different paradigms: the rational actor, organizational, and bureaucratic or governmental politics models.

Allison’s classical text has proven to be fruitful in explaining such a complicated and dynamic situation as the Cuban Missile Crisis, with its numerous channels of negotiations and many interests on the stake.  This encourages a researcher to apply the same paradigmatic construct to analyze the mediation attempts in Nagorno-Karabakh conflict and to assess their performance.

Allison (1971) starts with noting that the majority of policy analyses (at least at his times) were explaining and predicting the behaviours of governments in terms of the Rational Actor model, the main implication of which is that actions of nations and governments are aimed at maximizing strategic objectives, being ‘rational’ in this context.  Allison then stresses the need to balance the Rational Actor model by such models, which would allow seeing the decision-makers different than just ‘monoliths [that] perform large actions for large reasons’ (1971, p. 5), for these monoliths cover various gears and levers in a highly differentiated decision-making structure and … large acts result from innumerable and often conflicting smaller actions by individuals at various levers of bureaucratic organizations in the service of a variety of only partially compatible conceptions of national goals… (Allison, 1971, pp. 5-6).

Allison achieves this balance by introducing two alternative paradigms: the Organizational model and the Governmental (Bureaucratic) Politics model.  The former sees governments as organizations, functioning based on standardized patterns of behavior, with policies and actions being the outputs of this functioning; the latter stresses the politics within a government, with actions and policies resulting from bargains between different actors in the government. 

It shall be noted, that Allison saw the Rational Actor model to be ‘supplemented, if not supplanted’ (1971, p. 5) by the Organizational and Government Politics models.  In this research, the three models will be assumed to have the same level of epistemological validity, in order to elicit as many as possible aspects and mechanisms of the Nagorno-Karabakh conflict mediation attempts.

Allison presents the three models as analytical paradigms; for this, he relies on Merton’s (1968) definition of the structure of analytical paradigms.  Thus, each paradigm consists of basic units of analysis, organizing concepts, the dominant inference pattern, general and case-specific propositions, and type of evidence, suitable for that paradigm.  On this basis Allison presents in detail all three models and ways of utilizing them in a particular research.

Below the three models are presented as analytical paradigms.  It shall be mentioned that different elements of all three Allison’s paradigms are selectively utilized in this research; some elements are ignored for they have had no explanatory power in the particular case of Nagorno-Karabakh conflict.

The basic unit of analysis of the Rational Actor model, is the governmental action as a choice of nations or national governments; the choice is assumed to be maximizing strategic objectives.  The explanatory focus of this model is aimed at ‘solutions’ to strategic problems.  The organizing concepts of the Rational Actor model are the national actor (be it the nation or the government) as a rational and unitary decision maker; the problem, which calls for a response, Action as Rational Choice, which includes goals and objectives, options, consequences, and a value-maximizing choice.  The inference paradigm, generating the explanatory power of this model the assumption that ‘if a nation performed a particular action, that nation must have had ends toward which the action constituted a maximizing means’ (Allison, 1971, p. 33). 

The general propositions of this model are that an increase in the costs of an action, which can be expressed as a reduction in the value of its consequences or as a reduction of a probability of getting fixed consequences reduces the probability of selecting that action – and vice versa: the decrease in the costs of an action, expressed as an increase of the value of its consequences or of the probability of getting fixed consequences, increases the probability of its selection. 

The Organizational Model’s basic unit of analysis is the governmental action as organizational output – a result of routine organizational processes.  The organizing concepts are organizational actors – non-monolithic constellations of loosely allied organizations with leaders on the top, factored problems and fractionated power, parochial priorities and perceptions of different parts of the governmental system, and the action as organizational output – a concept consisting of goals as constraints defining the limits of acceptable performance of an organization, standard operating procedures (SOPs) as lower tasks allowing to perform strategic functions, programs and repertoires as sets of rehearsed SOPs, needed to perform actions with precisely coordinated behavour of many individuals, problem-oriented search as a process in which organizations engage when facing non-standard situations, and organizational learning and change – new routines resulting from problem-oriented search, as responses to non-standard problems. 

The other organizing concepts are the central coordination and control over decentralized responsibility and power within organizations and decisions of government leaders, who sit atop the organizations and can decide which of the programs or repertoires, currently under an organization’s disposal, will be utilized. 

The dominant inference patters on the organizational paradigm is that ‘if a nation performs an action of a certain type today, its organizational components must yesterday have been performing (or have had established routines for performing) an action only marginally different from today’s action.’ (Allison, 1971, p. 67).  The general propositions of this paradigm are that activities are not far-sighted adaptations to a problem, but are results of organizational implementation through SOPs, programs, and repertoires.  Organizations have options – limited menus of possible alternatives, developed in sufficient detail.  Organizations have limited flexibility and any organizational change is incremental – organizational behavior at any moment is only marginally different from its behavior at the previous moment.  Long-ranged planning in organizations is institutionalized, which allows for it being disregarded.  Organizational goals are formulated as constraints, tradeoffs are neglected, and incompatible constraints are attended to sequentially (satisfying one and ignoring another).  The evidence base of this paradigm is behavioural observation.

The Government (Bureaucratic) Politics model’s basic units of analysis is the government action as political resultant in the sense that happenings and actions are not selected as solutions aimed at solving particular issues, but are results of a process consisting of compromises, conflicts, and confusion of actors with different interests and unequal influence; this process is characterized as bargaining among political decision-makers, which takes place along regularized channels.  The organizing concepts of this paradigm are players in positions, who are independent individuals, a group of which is the agent for a particular decision, determinants of players’ stands, which are the parochial priorities and perceptions, players’ goals and interests, stakes and stands, and deadlines and different faces of issues, the determinant of players’ impact on result – the power (influence on decisions and actions) – that each player exercises, the game itself, consisting of action-channels (regularized means of action), rules of the game (can be explicit or implicit), and action as political resultant (since power is shared, but judgments of different actors are separate, hence, resultants of politics are compromises by definition).

The dominant inference pattern of the governmental politics model is that ‘if a nation performed an action, that action was the resultant of bargaining among individuals and groups within the government’ (Allison, 1971, p. 173).  The general propositions of the model are as follows: stands and preferences of individual players, their advantages/disadvantages, and the mix of players between and along the action channels affect the political resultants; intention does not presuppose action, because the majority of resultants are generated by games among actors with different perceptions of the issue, because there is rarely an agreed doctrine of action, and because actions rarely present a coordinated strategy of the government; problems are rarely considered to be strategic to the players, who rather focus on a decision to be made today; ‘where you stand depend on where you sit’ (Allison, 1971, p. 176), which implies that both vertically and horizontally, the demands upon the players shape their priorities; misperceptions matter – players play in the situation of incomplete information; misexpectations of players from each other, resulting from the pace of often multiple games, and the limited attention to each of these games from each actor; miscommunication, resulting from the pace and the noise level of games; and style of play differs for careerists and political appointees. 

The issue of evidence in this paradigm is far more problematic, compared to the other two ones; detailed information on perceptions’ and priorities’ differences, as Allison notes, is rarely available.  He argues that the source of information could be the participants themselves, each of whom, on the other hand, knows only a small part of the story, and hence information from public documents, press, participants, and expert community should be synthesized to serve as a basis for insightful inferences.

Analysis of Mediation Attempts

In what follows, periods of 1988-1992 and 1992-1998 of Nagorno-Karabakh conflict will be briefly analyzed through the lenses of each of the paradigms presented.  The results of this analysis will allow us to attempt to make conclusions on the reasons of mediation performance by now and to attempt to extrapolate these conclusions into some recommendations for further action.

1988-1991: Mediations into an Internal Soviet Conflict

As already mentioned above, during this period the Nagorno-Karabakh conflict has been an internal conflict of the Soviet Union, and the almost exclusive source of both arbitration and mediation were the central, or Moscow authorities.  Below we analyze these mediation attempts from all three models’ viewpoints.
Rational Model viewpoint: The question we seek an answer to is why the Soviet authorities selected to undertake the particular set of mediation attempts, which could be labeled under the umbrella concept of retaining the Union’s internal borders’ status quo.  We will focus here on Soviet choices and utility functions of these choices, the Nagorno-Karabakh problem, as it appears to the mediating party, and the goals and objectives of the mediating party, options under disposal, and consequences of each mediation action.  We shall bear in mind that the rational choice of each party is value maximizing for it, and that behind each action there are certain, rational for the acting party goals. 
During this period the almost exclusive source of mediation were the authorities of the Soviet Union, although, as mentioned above, the Soviet mediation attempts at that time included also the 1991 mission by presidents Yeltsin and Nazarbayev of (respectively) the Russian Federation and Kazakhstan – constituent parts of the Soviet Union at the moment.  We will refer to these attempts as to Soviet, once again recalling that this period of the conflict is remarkable for it has been an internal conflict of the Soviet Union.
What was the governmental mediation action that the Soviet authorities selected to implement, when faced with the problem?  The action can be characterized as pushing the conflicting parties to retaining the existing administrative status of Nagorno Karabakh in particular, and any change of border within USSR, in general.  The particular expressions of this action, in chronological order, are as follows:

  • The very early decision to solve the conflict by addressing ‘practical problems’ (Gorbachev, 1996); the underlying assumption here was that the socio-economic situation and the assumed under development of Karabakh was the reason for desire to secede from Azerbaijan.  An example of this is Gorbachev’s 1988 decision to provide assistance to Karabakh worth 400 mln soviet rubles; even ten years later, Gorbachev believed that had this type of assistance been provided ten years before the conflict started, it could have been prevented.
  • The mission of Politburo representatives to assist the local leadership (Gorbachev, 1996) – Yegor Ligachev and Georgi Razumovski were sent to Azerbaijan, while Alexander Yakovlev and Vladimir Dolgikh arrived in Yerevan.  The main message they brought was that any border change was unacceptable, however, concrete actions would be done so that economic, cultural, and social situation in Karabakh was improved.  The same message about inadmissibility of border change and necessity of improving the social, economic, and cultural situation in Karabakh was reaffirmed by Gorbachev during his meeting with Armenian poetess Silva Kaputikyan and writer Zori Balayan, which took place on 26 February 1988 (de Vaal, 2003).
  • The next mission of Politburo representatives (Yegor Ligachev and Alexander Yakovlev) to Armenia and Azerbaijan respectively brought the same message: Karabakh shall remain the part of Azerbaijan.  Although Alexander Yakovlev tried to promote the compromise of Armenia renouncing its claim of unification with Karabakh in exchange of Azerbaijan relinquishing claims of retaining Karabakh, Ligachev in Baku insisted on Karabakh remaining part of Azerbaijan, and the final Moscow position was his – not Yakovlev’s – one (de Vaal, 2003).
  •  The next Moscow mediation attempt was raising the status of Nagorno Karabakh Autonomous District to an Autonomous Republic, again remaining a constituent part of Azerbaijan.  Thus Karabakh would acquire the right to have its own constitution, government, and parliament.  Gorbachev (1996) mentions that Politburo was in favor of this solution; de Vaal (2003) notes that the leader of Nagorno Karabakh Autonomous Region Henrik Poghosyan, who had replaced the notorious Boris Kevorkov in this position, was very close to accepting this option, and only at the last moment inclined to the option of secession from Azerbaijan.
  • A unique in its sort – a one time – mediation event was the meeting of the Presidium of the Supreme Council of the Soviet Union dedicated to the Karabakh conflict; the meeting took place on 18 July 1988 in Moscow. During the meeting, which was broadcasted on Soviet television, the theme of importance of friendship amongst nations as the basis of the Soviet Union was the main context, into which Gorbachev tried to put the Karabakhi issue; hence, he made notorious intrusions into and interruptions of the speeches of delegates from Armenia, when they would try to speak about the principle of self-determination of nations, and its place in the Constitution of the Soviet Union.  The meeting finished with a resolution stating that Karabakh should remain the part of Azerbaijan. 
  • On 24 July 1988, Gorbachev appointed Arkadi Volski as a representative of Politburo (officially – as a representative of the Central Committee of the Communist Party and of the Supreme Council of the Soviet Union) in Karabakh, with an authority to put a veto on any decision of Azerbaijani authorities (de Vaal, 2003).  Volski's mission lasted for one hear and a half, and, as de Vaal mentions (2003), Volski was actually a general governor of Karabakh, ruling on behalf of Moscow.  The focus of his activities was the economic and social recovery of Karabakh.  Volski's mission lasted till January 1991, when, after the bloodshed in Baku, he left the region.
  • On 14 January 1990, just in the midst of massacre of Armenians in Baku, Moscow sent another mediation mission, headed by Yevgeni Primakov, then time member of Politburo.  The Soviet Minister of Defence Yazov was also part of the delegation.  The immediate focus of this mission's attention was the resolution of the situation in Baku; at the same time, as one of the leaders of the People's Front of Azerbaijan Etibar Mamedov mentions (in de Vaal (2003)), that Primakov stressed that Moscow was not going to tolerate Azerbaijan's any move away from the Soviet Union. 
  • Victor Polianichko’s rule in Azerbaijan and Karabakh can be considered as another mediation attempt.  Polianichko was the second secretary of the Central Committee of the Communist Party of Azerbaijan, and, after Volski left Karabakh, Polianichko became the head of Karabakhi administration, known as the Organizational Committee.  He was in charge in 1990-1991, and Polianichko’s main objective was to keep Karabakh as part of Azerbaijan, using political, economic, and military means; his strategy in Karabakh was to provoke tensions and spread dissention amongst Karabakh’s Armenians, as his aid Mirzoev recalls (in (de Vaal, 2003).   
  • And the last, but maybe the most potentially successful meditative action of this period was the so-called ‘Zheleznovodsk mission’ of September 1991.  The President of the Russian Federation Boris Yeltsin (who, after August 1991, exercised more de-facto power than Gorbachev as the President of USSR), together with the President of Kazakhstan Nursultan Nazarbaev visited Baku, Stepanakert, and Yerevan; the visit took place on 20-23 September 1991.  The results of this visit were summarized in the Zheleznovodsk joint communiqué, which stated that Armenia and Azerbaijan should announce a ceasefire, cancel all 'non-constitutional' decisions of the parties regarding Karabakh, withdrawal of all armed forces from the region, except for the USSR Ministry of Interior and Ministry of Defense forces, establishment of a task force consisting of Russian and Kazakhstani representatives, return of internally displaced people to their native lands, and start the process of peaceful resolution of the conflict.  However, due to a helicopter crush in Karabakh, which took lives of about 20 Azerbaijani high officials, Azerbaijan withdrew from the Zheleznovodsk process (Zolyan, 2001).  Kazimirov (2005) provides with another, more general and attributive explanation of the reasons of failure of Zheleznovodsk process: the ‘mutual fault’ of the parties.

What were the goals, objectives, consequences, and choices of the mediating party?  From the very first days of the conflict in February 1988, Moscow authorities, who already had been operating in the rather new – and unpredictable - context of Gorbachev’s perestroika, perceived the tension in Karabakh as a new, quite dangerous challenge for their already non-absolute power in the country.  The problem for the Moscow authorities was the problem of preservation of the internal order within the Union; as Gorbachev (1996, pp. 333-334) recalls, at the very first meeting of the Politburo dedicated to the tensions in Karabakh (21 February 1988, before the Sumgait massacre), the Chairman of KGB Chebrikov reported that Karabakhi events ‘were having a bad effect in other republics’, referring to Estonian sentiments in favour of independence, tension between Tajikistan and Uzbekistan around Bukhara and Samarkand.  Hence, the general goal of the Soviet mediation was to retain the Union, to avoid its collapse.  The objective, achievement of which would ensure that the Union does not collapse was not to allow for a precedent of changing administrative borders within the Union.  The consequence of keeping the internal borders unchanged was supposed to be a relatively stable Union, going through perestroika; the consequence of changing the belongingness of Karabakh would have been the chain reaction of similar processes all over the Soviet Union; at a certain point, these processes would have immediately transform from the change of internal border into a dissolution of the Union (as it in fact did happen). 
Therefore, mediation in order to leave the border unchanged, while changing the status of Karabakh within Azerbaijan, investing into Karabakh’s economic system, or protecting the rights of Armenian population, all these mediation actions were serving the strategic objective of the Soviet authorities – to preserve the Union.  And even the Zheleznovodsk communiqué, facilitated by presidents Yeltsin and Nazarbayev, could be argued to be aimed at the same goal: indeed, the communiqué envisaged a key role for the armed forces of the Ministries of Interior and Defense of the Soviet Union, after 1 January 1992, meanwhile, the Union ceased to exist (with active participation of the same president Yeltsin) on 25 December 1991 – a little than three months after the Zheleznovodsk mission. 
Organizational Model Viewpoint: The question we seek to answer here is what the organizational process were that determined the particular mediation attempts.  Mediation attempts were all organizational outputs, emerged as a result of organizational actors dealing with factored problems and fractionated power, under the conditions of parochial priorities and perceptions.  The actors had certain goals, they paid sequential attention to the problems, they had programs consisting of standard operating procedures, and government leaders’ decisions sometimes shifted program-based organizational decisionmaking. 
Within this theoretical frame, who the organizational actors (who were subjects of mediation) during this very first phase of the conflict were?  A rough and approximate estimation would include the following ones:

  • The Political Bureau (Politburo) of the Central Committee of the Communist Party of the Soviet Union;
  • The Presidium of the Supreme Council of the Soviet Union;
  • Volski’s administration;
  • Polianichko’s administration; and
  • Presidents Yeltsin and Nazarbayev.

The Karabakh problem was parceled out to various organizations, subsequently, power and responsibility were also fractionated.  Indeed, whilst, from the very beginning, Politburo, chaired by Gorbachev, was dealing with perhaps the most important part of the Karabakh problem – border change and Karabakh status, Soviet missions in Karabakh (such as Volski’s mission) dealt mainly with socio-economic issues of Karabakh, while Polianichko’s ‘Organizational Committee’ was tasked with restoring order (in Soviet sense of the word) in Azerbaijan and Karabakh; not surprisingly, military power was parceled to Polianichko, for he was to solve a political problem in a region with frequent outbursts of violence. 
The parceled problems and fractionated power are responsible for the parochial priorities and perceptions of the mediators.  The parochialism of the Politburo sounds paradoxical, however, it makes sense when the Politburo is seen as the governing body of the Communist Party of the Soviet Union, which was gradually losing power and credibility at the time of the conflict.  The parochial priority of Politburo were to retain the power within the Soviet Union, the necessary condition for which was retain the Soviet Union itself.  The Karabakh conflict was seen inside Politburo as a factor potentially undermining the existence of the Union by sparkling a number of similar conflicts (Gorbachev, 1996), therefore, the priority of Politburo, within the context of this perception, was to resolve the conflict as soon as possible, by means which would not serve as an example for the other potentially dangerous parts of the Soviet Union. 
Volski’s administration in Karabakh became very much associated with the social and economic situation in the region; de Vaal (2003) stresses the respect that Volski enjoyed from both parties of the conflict.  Volski’s mission’s parochialism meant this being associated with mainly social and economic problems of Karabakh; his priorities, subsequently, were the economic and social recovery of the zone of the conflict.  In contrast, Polianichko’s regime, which, as de Vaal (2003) notes, was closely associated with the hawkish wing of Politburo (Minister of Defense Yazov, Chairman of KGB Kruchkov, etc.), shared these hawks’ parochial perceptions of Karabakh as a separatist threat to the Soviet Union, and the priority of Polianichko’s regime was to undermine the sources of such a threat in all possible ways.  Indeed, as Polianichko’s assistant Seyran Mirzoev recalls it was a conscious and planned action that Polianichko's administration implemented, to introduce discord and conflict to the Karabakhi 'separatists' (in de Vaal (2003) – an aim actually achieved by the end of 1990.
Presidents Yeltsin and Nazarbayev had their own parochial agenda; Yeltsin was in opposition to Gorbachev, who, during the Zheleznovodsk mediation, was still formally the head of state.  Nazarbayev had his own agenda, with the vision of either a greater political role within USSR, or leadership of independent Kazaskhstan.  Both men perceived Karabakh conflict as an opportunity to show the ability of independent from Gorbachev troubleshooting. 
Action, according to Allison’s presentation of the organizational model, is an organizational output; this means that it is a product of a programmed and routine character.  In this context, the goals of organizational actors are defined by a set of constraints, defining acceptable performance.  For the Politburo the goal was to preserve the Soviet Union, and the constraints, defining the acceptable performance were policies that could have resulted, in principle, in the dissolution of the USSR. Politburo, as outlined in the model, paid sequential attention to the emerging issues – the perceived importance of the issues was determining this sequence, hence, Karabakh was a matter of discussion both in Poliburo and in the Presidium of the Supreme Council, when the challenges brought ahead by Karabakh seemed to be undermining the supreme goal of the Soviet authorities – the preservation of the Union.  When, in 1990-1991, more and more internal conflicts emerged within the Union, and Karabakh ceased to be the main challenge to the Union’s preservation, the attention to it from Politburo and other central government bodies gradually decreased. 
Poliburo dealt with Karabakh using certain programs, which are sets of standard operating procedures; the latter are routine operations that serve as basis for higher level activities.  The general propositions of the organizational model say that these standard operating procedure based programs are not the strategic ones, adopted for the particular situation, rather, they are standard responses to standard situations, and as such they allow for regularized performance, which is achieved at price of standardization – in critical situations these programs are unable to provide with flexible responses.  This is the reason why organizational change is incremental, and even in non standard situations the organizational learning / change, is based on existing procedures (Allison, 1971). 
Politburo’s standard operating procedure was discussions where the Secretary General had the right of the final decision – a role that is envisaged in the organizational model, where the governmental leaders, sitting on the top of the organizations, can shift policy, without much changes in standard operating procedures.  Another standard operating procedure was replacing party leaders in the problematic areas.  The third and last (for our purposes) standard operating procedure was to appoint a representative of Moscow as a supreme decision maker in the site of the conflict. 
Politburo discussions on Karabakh with Gorbachev summarizing and making the final decisions, replacing of Boris Kevorkov, the party leader of Nagorno Karabakh Autonomous Region, Kiamran Bagirov, the party leader of Azerbaijan, and Karen Demirchyan, the party leader of Armenia, and Volski’s and Polianichko’s administration in Karabakh are respectively examples of the most important standard operating procedures of Moscow trying to solve an internal conflict.  All the attempts of the Soviet authorities to resolve the conflict were programs, based on different combinations or sets of these standard operating procedures.  However, these procedures and programs were not suitable for the new challenge of raising national identity and centrifugal forces within the Soviet Union.  Gorbachev’s attempts to solve the conflict through the investment of 400 mln rubles in the economy of the region, consideration of raising the status of the region up to the autonomous republic, envisaging a separate budget for Karabakh within the budget of the Soviet Union (de Vaal, 2003), Volski’s attempts to revitalize the economy of the region, Polianichko’s tough rule and suppression of the local movements – these all are examples of the Soviet authorities engaged (with little success) in organizational learning in a non-standard situation. 
However, the Soviet authorities did not succeed in actually changing as an organization – the same standard operating procedures and programmes based on them were continuously used throughout 1988-1991; when the Soviet party found itself in a novel, ambiguous, and challenging situation, the search, in which it embarked, did not yield any novel programmes, based on new standard operating procedures.   The programmes the Soviet party was implementing in 1988-1991 were only marginally different from the programmes and standard operating procedures it had had before; and even the Zheleznovodsk communiqué, which called upon the parties to return to status quo and to allow for the central (Moscow) military forces to restore the order, show that even Yeltsin and Nazarbayev acted based on the standard operating procedures and programmes based on these, which were not at all different from the ones used decades ago.  Therefore, the Soviet mediation in Karabakh conflict can be explained by its routinized organizational behavior, and the Soviet mediation failure in Karabakh contexts is directly inferable from the failure of organizational learning.
Governmental Politics Model: The question we seek to answer within the context of the Governmental Politics Model is as follows: which results of which bargaining processes among which players resulted in particular mediation attempts.  Mediation action is seen as a political resultant, not selected for a certain solution, but rather appearing as a result of a conflict, misunderstanding, and/or compromise amongst political players with differing political agendas and varying influence.  The characteristic process here, according to Allison, is bargaining. 
Players in this mediation game are individuals within the government and their groups; these are

  • Gorbachev;
  • ‘Conservative’ members of Politburo – Ligachev, Yazov, Kruchkov;
  • ‘Liberal’ members of Politburo – Yakovlev, Shevarnadze;
  • Volski and his administration;
  • Polianichko and his administration; and
  • Presidents Yeltsin and Nazarbayev.

Players’ stands are determined by certain factors – parochial positions and perceptions, goals and interests, stakes, and deadlines, and different faces of issues.  Gorbachev’s position as the leader of the Soviet Union with his perestroika, glasnost, and new thinking sorts of concepts, not fully understood and supported within the country, was a position of a leader with not much support, ambiguous future, and number of novel challenges – Karabakh being one of them.  Gorbachev’s goal was to preserve his power and to prove the rightness of the perestroika discourse.  His whole political career was on the stake, for Karabakh was the first and the biggest challenge for it.  The ‘conservative’ members of Politburo, who would had perfectly suited the political context of early 1970s, rather than late 1980s, were in positions of men who constantly had to defend the organizations and procedures they represented, in the context of rapid changes with unclear direction.  Their goal was the preservation of the existing power system with the Communist Party on the top.  On the stake were their power positions.  The ‘liberal’ members of Politburo were in positions when these rapidly changing context was seen to present them opportunities for acting as independent political figures and gaining own political dividends.  Their interest was to create better positions for further governmental career – on the stake was their public image of liberal, perestroika-generated politicians.    
Volski as a player is rather interesting in the picture.  Not a characteristic man in the Soviet hierarchy – a self-made industrialist with a very rich experience of management in machine industry.  He had a brilliant career in the Communist Party apparatus, dealing with industry and economy related issues. In post-Soviet years he continued to play a significant role in Russia’s business community’s life, becoming the founding member of the Russia’s main business club – the Union of Industrialists and Businessmen (Institute of Problems of Commerce NGO, 2004).  His ‘parochialism’, so to say, was determined by his previous experience in economy and industry – his main strategy during his 18 month mission, as de Vaal mentions (2003), was to improve the socio-economic situation in Karabakh.  His goal was to restore the economy and to solve the social problems in the region.  His interest might have been to fulfill successfully his usual function of a manager (with the goals pre-given).  On the stake was Volski’s reputation of a manager and troubleshooter.  
Victor Polianichko’s background also was a defining factor for his parochialism: this man, with close links with the ‘conservatives’ Yazov and Kruchkov, had been the second secretary of the Central Committee of the Communist Party of Azerbaijan. His parochialism was Soviet-minded, and the main strategy (as mentioned above) of his Organizational Committee was to try to divide the Armenian Karabakhi movement.  Also, as de Vaal (2003) mentions, Polianichko had come to Azerbaijan, and, later, to Karabakh from Afganistan, where he was the Soviet shadow ruler of the country.  Polianichko’s Organizational Committee replaced Volski’s administration in January 1990, and its main strategies were to use tough political and military means against Karabakhi Armenians, using terror and arrests of Karabakhi Armenian activists.  Polianichko was responsible for the formation and dislocation in Karabakh and adjacent areas of Azerbaijani special police forces, counting up to 10 000 men.

His goal was to ensure the loyalty of Azerbaijan to the Soviet Union, as this loyalty was seen, as de Vaal (2003) mentions, to be critical for the preservation of the Union.  Ensuring the loyalty of Azerbaijan apparently implied suppressing the movement for independence in Karabakh – exactly what Polianichko’s rule was aimed at.  On the stake for him was his future career on a Moscow level.  
A very special case of parochialism can be found when considering Presidents Yeltsin and Nazarbayev’s Zheleznovodsk mission; these two players had had their roots in the highest circles of the Communist Party bureaucracy, however, being charismatic and flexible men, by Fall 1991 they have both expressed claims for a greater role rather than merely heads of units in a confederation, which the Soviet Union had de-facto become by then.  Their ‘parochialism’ is a parochialism of leaders ready to act and solve problems without Gorbachev.


The goal for each of the presidents was to get domestic and international recognition as of successful peacemakers (possibly, as opposed to an unsuccessful Gorbachev).  On the stake was their reputation of successful leaders of the new, post-Gorbachev era.
We have discussed the determinants of political stand of each mediating player involved.  The next question is what the determinants of their necessarily varying impact on results – power.  Gorbachev’s power as of the leader of the USSR was far from being absolute – he had to deal with a Politburo, which was, as the future showed, not wholly loyal to him.  His power was big – yet, it was not absolute.  The ‘conservative’ members of Politburo exercised significant power within the apparatus and local structures of the Communist Party, however, the power of the Party itself within the Soviet Union was steadily decreasing.


The ‘liberal’ members of Politburo, being promoted from republics (as Shevarnadze), industry (as Rizhkov), or diplomatic service (as Yakovlev) had less power within the Party structures, yet, they had better public image, compared to the ‘conservatives’.  Volski’s power was in his capacity of a representative of the Moscow government, however, as his disagreements with Gorbachev on Karabakhi issues rose, his support from Moscow – and consequently his power – decreased.  Polianichko’s power – again as Moscow’s representative – remained stable during the period of his rule, due to his close links with the ‘conservatives’, towards whom Gorbachev was leaning more and more in 1990-1991.
What the action-channels – ‘the regularized means of taking governmental action on a specific kind of issue’ (Allison, 1971, p. 169) – were for each of the major mediation attempts?  The action-channels that Gorbachev and Politburo took from the very beginning of the conflict as a mediating party were the replacement of local party leaders (Demirchyan, Bagirov, and Kevorkov), use of military force, sending short-term mediating missions consisting of members of Politburo, direct contacts with the representatives of intelligentsia of the conflicting parties, and imposing direct rule of Moscow in Karabakh.  The action-channels of Volski’s administration was the improvement of the socio-economic situation in Karabakh; one particular action, as de Vaal (2003) mentions, was supporting Karabakhi industries in establishing direct links with Russian partners.  Polianichko’s action-channels were much more straightforward: use of military force to ‘restore order’, arrests of Armenian Karabakhi activists, and creation and deployment in Karabakh of Azerbaijani special police forces.


President Yeltsin and Nazarbayev’s action-channel was probably the closest to the conventional understanding of mediation – the two men visited separately the conflicting parties, and then summarized their recommendations in a communiqué.
The next, critically important organizing concept of this model is the rules of the game – the implicit or explicit canons, whereby the positions of all parties involved are defined, the range of acceptable actions is constricted, and moves of certain kinds are allowed, while others are prohibited.  During the 1988-1991 phase of the Nagorno-Karabakh conflict, the main rules of the mediating game stemmed from the Constitution of the Soviet Union, which reserved a leading role for the Communist Party, hence, the role of Politburo in the first phase of the conflict.  Since the Secretary General could delegate the member of Politburo to the conflicting republics as designated mediators, it is of no surprise that in Baku the territorial integrity of Azerbaijan was stressed by the Moscow delegates, while in Yerevan and Stepanakert the right of self-determination (with the status of an autonomous republic, at least) was being stressed.  Prohibited were any promises of independence for Armenia and Azerbaijan themselves.


The position of Moscow – central Soviet authorities – as of the exclusive source of mediation was the basic rule of the game – hence, the Volski’s and Polianichko’s administrations can be seen as action-channels for Gorbachev and Politburo.  And even Yeltsin and Nazarbayev’s mission at the final stage of the Soviet Union stuck to the basic rule of the supremacy of the central mediation and exclusive right of even military mediation by the Soviet Ministry of Defense and Ministry of Interior only.
The Governmental Model strictly differentiates between an action and intention: the very idea that the governmental actions are resultants of bargaining among different actors implies that actors’ intentions not necessarily translate into actions.   Indeed, Gorbachev’s intention could be to lead the Soviet Union through the period of change, however, his actions, aimed at the preservation of the status quo, had the opposite result: military conflict which accelerated the dissolution of the Soviet Union.  The same intention can be attributed to Polianichko – the means he selected to preserve Karabakh within Azerbaijan pushed back with a tremendous force.  Volski’s intentions and actions could be seen as to be more aligned – restoring peace through socio-economic measures.  Yeltsin and Nazarbayev’s intention was to solve the conflict and get the whole credit for it; their action was aligned with it – shuttle diplomacy and signing of a concluding communiqué.  However, for the reasons presented above, they did not succeed.
Finally such propositions as misperceptions (games are played in the situation with imperfect information) and miscommunication (the pace and the noise level making the communications difficult) had significant effect on the actual failure of all the mediation attempts of this period.  Indeed, as powerful as Politburo was, its members did not seem to have clear understanding of the reasons why Karabakh demanded independence from Azerbaijan.  Gorbachev’s main misperception, general to his entire vision of the country’s challenges – not to Karabakh specifically – was that his policies of perestroika, glasnost, and new thinking were supported by the majority of population; therefore, these policies could serve as sufficient basis for resolving the conflict.  Miscommunication in Gorbachev’s case has played a critical role – the noise level created by the ‘conservatives’ in Politburo, who later were going to attempt a coup d’état against him, made his position more and more detached from what was really happening in Karabakh, and his actions based on this miscommunication, completely eliminated the last pieces of Gorbachev’s credibility in the region.  As in his interview with de Vaal (2003), the last Communist Party leader of Armenia Aram Sargsyan recalls that in July 1991, something like six months before the collapse of the Soviet Union, during the session of the Soviet parliament, he saw Gorbachev speaking with the Minister of Defense Yazov, Chairman of KGB Kryuchkov, and party leader of Azerbaijan Mutalibov; when Gorbachev noticed Sargsyan, he asked him what the Armenians were doing in Karabakh – blasts, terror…  Apparently, this is what Gorbachev had just been told by the ‘conservatives’; interestingly, this conversation took place after the ‘Kolco’ joint operation by the Azerbaijani and Soviet armed forces, and right during the time of another joint terrorist operation by the same actors against Armenian villages in Shahumyan region.
In the case of Arkadi Volski, the concept of miscommunication describes his relationships with Gorbachev.  Volski went to Karabakh per Gorbachev’s request, however, Volski’s position as being too much associated with what Karabakhi realities resulted in his gradual loss of credibility with Moscow.  Also, he recalls situations when he was unable to reach Gorbachev on the phone in critical situations (in de Vaal (2003).  Polianichko, on the other hand, had the misperception that he would be able to suppress the Armenian movement in Karabakh in the same way he had acted in Afghanistan.  His even bigger problem was his failure of communication with the Armenian party – his Organizational Committee in Stepanakert was operating in complete isolation from reality, the Armenian members of the committee (except for one) did not participate in its sittings at all, and his only real lever were the Soviet Armed forces deployed in Karabakh (de Vaal, 2003).

1992-1998: War, Ceasefire, and Mediation

This period of the conflict is marked by the simultaneous independence of Armenia and Azerbaijan, almost immediate outbreak of the Karabakhi war, ceasefire, start of the international community’s involvement, and loss of power by three presidents, because of the conflict.  The two almost exclusive sources of continuous mediation were the Russian Federation, and the Conference on Security and Cooperation in Europe (CSCE), renamed in 1994 into Organization on Security and Cooperation in Europe (OSCE).  We proceed now with again sequentially applying the three models to each of the mediating attempts of this period.
The Rational Actor Model: The question we seek to answer is what the reasons were for the Russian and CSCE/OBSE particular mediation actions and proposal.  The unit of analysis is the mediation action as a choice of a set of activities, selected by the mediators to maximize some strategic goals or objectives.
The particular mediating actions of this period are as follows:

  • OSCE-led mediation attempts

Nagorno-Karabakh conflict comes to CSCE attention in January 1992, when during the CSCE Council of Ministers Second Meeting in Prague ten former Soviet republics were invited to join the organization as participating states (Organization for Security and Cooperation in Europe, 2009).  In March 1992, the Foreign Ministers of the member-states decided to hold a special conference on Karabakh in Minsk, which did not take place due to the escalation of the conflict (de Vaal, 2003), however, Kazimirov (2005) puts it more directly, mentioning that the Conference, which was scheduled for 23 June 1992, failed, because on 19 May 1992, the Azerbaijani party laid as a precondition the withdrawal of Armenian forces from Shushi and Lachin.  Instead of the Conference, the Minsk Group of states was formed in order to mediate the peaceful resolution of the conflict, with the first chairman being Mario Rafaelli from Italy. 

The first Minsk Group negotiations took place in Rome in June-August 1992 (de Vaal, 2003), with no substantial effects.  During the Stockholm Meeting of the Council of Ministers of Foreign Affairs, as de Vaal (2003) mentions, the mediation by the members of the meeting succeeded in almost achieving an agreement, but the Azerbaijanis stepped back at the last moment, however, Minsk group was not very much involved in this, for its Italian chairman did not visit Stockholm at all, Russia was not represented, and the United States sent only John Mareska – its representative to the Group. 

On 20 June 1992, the Minsk group made its first proposal on de-escalation of military actions and offensive operations, however, as Kazimirov (2005) mentions, at that moment there was no demand of such de-escalation on both sides of the conflict.  On 3 July 1992, the group proposed that the conflicting parties suspend the military actions for a month; the Azerbaijani party agreed, but there was no reply from Yerevan and Stepanakert, who, allegedly, did not receive the proposal (Kazimirov, 2005).  Another similar proposal to suspend the military actions was made on August 5, and Yerevan and Stepanakert immediately agreed, while Azerbaijani president Elchibey spent months to reply (Kazimirov, 2005).  In December 1992, during the Minsk Group meeting in Geneva, the Armenian party proposed to announce a ceasefire, which was not accepted by the Azerbaijani party (Kazimirov, 2005).  In April 1993, after the Armenian forces conquered the Kelbajar region, the Azerbaijani party refused to participate in the Minsk process, putting as a condition of their return the withdrawal of the Armenian forces from Kelbajar (Kazimirov, 2005).  On 18 April 1993, a temporary cease fire was established for the period of the CSCE mission visit to the region, however, the proposal to extend the ceasefire till the end of April – for the period of the CSCE CSO meeting was rejected by the Azerbaijani party (being accepted by Yerevan and Stepanakert). 

In 1993 the chairmanship of the Minsk group was passed from Mario Rafaelli to Ian Eliason from Sweden; this was followed by a decision (de Vaal, 2003) to shift from meeting-style of mediation to more direct negotiations and frequent regional visits – a format, which has remained prevalent since.  The first visits took place in the midst of the war – in February 1994.  However, the Minsk Group played no critical role in the establishment of the ceasefire in May 1994, which was due to the Russian and Commonwealth of Independent States’ mediation attempts.

On 5-6 December 1994, during the CSCE Summit in Budapest, the Conference was reorganized into an Organization of Security and Cooperation in Europe.  The final document, entitled ‘Towards a Genuine Partnership in a New Era’ of the Summit stressed the leading mediating role of Russia (in co-operation with the Minsk Group) in the achieving of the ceasefire (Organization for Security and Cooperation in Europe, 1994).  This document also envisaged the appointment of co-chairmen of the Minsk Group, which should work towards reaching a peaceful political agreement on Karabakh, also, the creation of multi-national peacekeeping forces was covered (Organization for Security and Cooperation in Europe, 1994).

However, despite of the moderately optimistic note of the Budapest summit, no real tangible advancement in OSCE mediating attempts took place before the Lisbon summit of 1996; in fact, as de Vaal (2003) mentions, during these two years, the parties were at their furthest point to reaching a political agreement, due to radical disagreement of the parties involved on returning to Azerbaijan the territories outside of Nagorno Karabakh, controlled by the Karabakhi forces, Lachin corridor, and refusal of Azerbaijan to negotiate directly with Nagorno Karabakh.  It shall be mentioned that the international community was at that moment much more preoccupied with the Russia’s internal war in Chechnya, the situation in the Balkans, and the re-election of Boris Yeltsin as the President of Russia.
Also, in December 1996, France became a co-chairman of the Minsk group, with the US following as the third co-chairman in February 1997.  This tripartite co-chairmanship led to more active work of the Minsk Group – the first result was the co-chairmen’s proposal of May 1997, which implied deployment of OSCE’s peacekeeping forces in the zone of the conflict and determination of the legal status of Karabakh as a result of a mutually acceptable compromise, to be reached during the OSCE Minsk Conference on Karabakh, to be convened in three months after signing the peace agreement (OSCE Minsk Group Co-Chairmenship, 1997).  In September 1997 the co-chairmen presented a reworked proposal, which ensured the security of Karabakh from the very first stage of the peaceful resolution process, with demilitarization of Karabakh and further negotiations on its legal status following this first stage.  This project was welcomed by the Armenian president Ter-Petrosyan, who was convinced, that time was on Azerbaijan’s side (de Vaal, 2003), however, Ter-Petrosyan’s acceptance of this proposal was countered by the key men in the Armenian government who saw this proposal as not favorable for Armenia, and in the beginning of February 1998 Ter-Petrosyan resigned from the presidency.  The next president of Armenia became Robert Kocharyan, who previously had been the leader of Karabakh and who had been one of the key critics of Ter-Petrosyan’s positive attitude towards the Minsk Group Co-Chairmen’s recent proposal.
Russia-led mediation attempts

  • Russia has been engaged in the mediation attempts both as the key player in the Minsk Group, and separately.  The first of its independent mediation attempts were led by the Defense Minister Pavel Grachev, who organized a meeting of Ministers of Defense of Armenia and Azerbaijan in Sochi, in September 1992.  The result of the meeting was a ceasefire agreement allegedly signed (de Vaal, 2003) by the Ministers and which implied Russian control over Lachin corridor, which connects Karabakh and Armenia, however, Azerbaijan denounced the signature of its Defense Minister. 

Another Russian mediation attempt were the two meetings organized in September and October 1993 by the Russian Foreign Ministry between the President-elect of Azerbaijan Aliyev and the leader of Karabakh Robert Kocharyan.  The meetings did not produce any tangible agreements (de Vaal, 2003), however, they allowed to establish working relationship between the two men. 
The real ceasefire in Karabakh became possible after the signing of the Bishkek protocol of the Parliamentary Assembly of the Commonwealth of Independent States’ meeting, which established a ceasefire from 9 May onwards (Parliamentary Assembly of the Commonwealth of Independent States, 1994).
From 1993 onwards, when Russia officially announced its claims of being a privileged warrant of security and stability in the former Soviet Union region, Russia’s attempts to play the leading role in the resolution of the conflict and specifically to deploy Russian peacekeeping forces became continuous (de Vaal, 2003).  After Bishkek agreement, Russia proceeded with attempting to deploy its peacekeepking troops in the region; in May 1994, Russian Defense Minister Grachov again tried to have the Defense Ministers of Armenia and Azerbaijan sign a ceasefire agreement – despite the Bishkek Protocol had already been signed – with Russian peacekeeping troops’ employment as a part of it.  Azerbaijan did not agree with this (de Vaal, 2003).
Iranian mediation

  • Although very short-lived, this one-time mediation is significant, for the mediator was Iran – a Muslim state with huge Azerbaijani minority on one side, yet with cultural ties with Armenia lasting for many ages.  After the collapse of the Soviet Union, Iran has had close relationships both with Armenia and with Azerbaijan, and Iran had expressed its concern over the conflict from the outset.  In May 1992, the Presidents of Armenia and Iran Ter-Petrosyan and Hashemi-Rafsanjani, as well as the acting President of Azerbaijan Mamedov conducted negotiations in Tehran, signed a Communiqué, the main implication of which was supposed to be the ceasefire and opening of all transportation roots.  However, the seizure of Shushi by the Karabakhi forces on May 9, by which Armenian President Ter-Petrosyan seemed to be caught by surprise, cut short of this Iran’s first and last mediation attempt.

Let us now consider, in concordance with the logic of Allison’s Rational Actor model, to which strategic ends the choices of action of each of the mediating actors served as means.  The problem that OSCE faced was the conflict amongst its two member states, which had emerged before Armenia and Azerbaijan joined the organization, but which escalated into a real war right during and after the two countries joined the organization.  De Vaal (2003) describes how during the CSCE Council of Ministers Second Meeting in Prague in January 1992 one of the British delegates raised the attention of the participants on the fact that two of the states that had just joined the organization were in war with each other.  The goal of the OSCE was to be able to find and enforce a peaceful solution to the conflict, thus getting the credit so important for the organization, facing the challenges of dealing with post-Soviet realities.


OSCE’s rational option, maximizing value for the organization in achieving its goals was to organize as many opportunities for a dialogue as possible, coupled with generation of possible conflict resolution scenarios and putting them on negotiations table.  The positive consequence in the case of success of this option would have been the credit and the increase of the OSCE’s role; the consequence of the converse – non involvement – would have been the increase of the main competitor-in-peacemaking – Russia’s – role in the region. 
Indeed, one cannot overlook the passiveness of the OSCE Minsk Group in the periods of critical escalation of the conflict, and its active involvement during periods of relative stability.


After all, the main achievement on the road of peaceful resolution of the conflict cannot be credited to the OSCE – the Bishkek Protocol was signed as a result of mediation of CIS structures (although Vladimir Kazimirov, Russia’s representative to OSCE Minsk Group, played personally a great role in the signing of Bishkek Protocol).
After Lisbon, when the positions of one of the parties – Armenia – became weaker because of internal political shake ups, Minsk Group drastically activated its works, which resulted in the number of proposals, put on the table, again, it is characteristic of Minsk Group to become active (to attempt to get credits) during the period of relative stability.  The entire year of 1997 was marked with OSCE attempts of getting a peaceful resolution to the conflict and, subsequently, increasing the role of the organization.  However, the rush with which OSCE presented proposal after proposal, with no enough attention to the internal political predisposition in Armenia led to the change of power in Armenia, and the hitch in the entire resolution process.
Therefore, one can claim that in terms of Rational Actor model, OSCE’s mediation attempts were aimed at increasing the organization’s role and significance in the region and competing with Russia as the main mediator.
What about the Russia-led mediation attempts?  The problem for Russia was a conflict between two nations within the sphere of Russia’s traditional interests.


Russia’s goal was retain its influence in South Caucasus – a region that had been controlled by Russia or the Soviet Union for almost two centuries.  The objective was to become instrumental in establishing the ceasefire, thus claiming the position of the primary geopolitical factor in South Caucasus.  The option that Russia chose to achieve its objectives was getting the conflicting parties agree on the ceasefire first, with the discussion of the remaining issues of the problem afterwards – indeed, the achievement of the ceasefire with Russia mediation would almost completely serve Russia’s goals, outlined above.  The consequence of a possible alternative action – getting the parties negotiate the whole scope of conflict resolution – could have resulted in CSCE/OSCE Minsk Group becoming the main facilitator of the ceasefire, thus minimizing Russia’s claims to play a key role in the post-soviet space. 
What was the problem for Iran?

  Again, a conflict between its two neighbors, within a very complicated context with Azerbaijani minority in Iran and US-Azerbaijani and US-Armenian close links.  The potential penetration of American influence in South Caucasus led Iran to attempt to solve the problem and fill in the space of the regional power – this could be Iran’s goal.  The objective was the enactment of the ceasefire and the opening of transportation infrastructure, in order to ensure that the ceasefire will last for a while.  The consequence of not acting as a mediator was that an unwanted mediator – the US - could become instrumental in achieving a ceasefire.  However, as already mentioned, the seizure of Shushi cut off Iranian mediation attempts from the very start.
Organizational Paradigm – Let us once again recall that the question we seek to answer here is what the organizational process were that determined the particular mediation attempts by the OSCE Minsk Group, Russian Federation, and Iran.  Mediation attempts were, in Allisons terms, all organizational outputs, emerged as a result of organizational actors dealing with factored problems and fractionated power, under the conditions of parochial priorities and perceptions; the actors had certain goals, they paid sequential attention to the problems, they had programs consisting of standard operating procedures, and government leaders’ decisions sometimes shifted program-based organizational decision making.
The organizational actors during this phase of the conflict were as follows:


  • CSCE/OSCE and OSCE Minsk Group;
  • Russian Federation, represented by Vladimir Kazimirov and Pavel Grachev; and
  • President of Iran Hashemi-Rafsanjani.

Let us now consider how the problem was factored and power was fractionated within each of the actors.  CSCE/OSCE itself was a loosely coupled alliance of countries, where not every country was equally interested and involved in the issue.  For this reason the problem was factored and delegated to the special Minsk Conference, which never took place, and to the special body – the Minsk Group.  The problem was factored per the following decomposition: all the ground principles and strategic directions were set during the CSCE/OSCE summit in Budapest, while the tactical implementation was laid with the Minsk Group.  This naturally implied the fractionated power within the organization – indeed, the Minsk Group Co-Chairmen never had the power and leverage of challenging any of the grounding principles, such as the territorial integrity or self-determination of nations; their mandate covered only preparing a basis for the parties’ negotiations, encourage the parties’ contacts, inform and consult the Chairman-in-Office of OSCE and the Security Council of the United Nations, and other mainly communicative functions.


The main parochial priority of OSCE was showing that that and only that dynamically developing organization was able to deal with conflicts between its member states.  Hence, all mediating actions by the OSCE Minsk Group were more aimed at immediate signing of peace agreement, without often paying attention on the internal dispositions of the parties involved: the February 1998 resignation of President Ter-Petrossian, who, having accepted the OSCE Minsk Group Co-chairmen proposal was not able to persuade influential men in his own government, can serve as an example.  The goal – in the organizational paradigm’s terms, the constraints defining acceptable performance – of the Minsk Group was the preservation of the ceasefire; the standard operating procedures of the Minsk Group were results of their mandate, outlined above; these procedures were (and still are) consultations with the Armenian and Azerbaijani leadership, organizing meetings of Presidents and Foreign Ministers, shuttle diplomacy, and infrequent consultations and meetings with Karabakhi authorities.  Different programs, based on these standard operating procedures, varied depending on the concrete period under consideration: as already mentioned, the frequency of the visits of the Minsk-Group Co-Chairmen drastically increased after Lisbon Summit, and the main program of that period was shuttle diplomacy coupled with drafting and presenting to the parties of proposals, summarizing the previous period of negotiations.


Rather interesting is to follow the CSCE/OSCE’s and the Minsk Group’s organizational learning during this first period of their involvement.  As Allison (1971) notes, organizational learning takes place in non-standard situations, when the given agent’s existing programs are unsuitable for solutions.  The very first example for the CSCE/OSCE was the Minsk Conference, which did not really take place.  Why a conference?  Because, per organizational model, any organizational implementation is not strategic, rather, it is based on the existing standard operational procedures and programs based on these, so any organizational change is incremental and has limited flexibility.  The implication for the CSCE/OSCE first mediations is that since the organization itself was a Conference, with main decisions being made on during gatherings of senior officials on a regular basis, a conference was proposed to deal with the conflict at once.  However, this program, based on the standard operating procedure of gathering and deciding, was too inflexible both in time and in terms of its toolkit, with such a dynamically developing conflict.  Hence, CSCE/OSCE found itself in deep in the process of organizational search and learning – the result was a new programme – creation of a more dynamic Minsk Group working, based on the old standard operating procedure – regular meetings and discussions. 

Hence, even the organizational learning was rather incremental and inflexible – of which Minsk Group’s performance by now can serve as a good example.  Indeed, the program, on which the format of the Minsk Group rested, was based on the same standard operating procedure, as the format of the never-happened Minsk Conference, or CSCE gatherings – meetings and discussions.  The organizational learning resulted in an organizational change, which is incremental by its nature.  The core principles of the Minsk Group operations (the standard operating procedures, in the organizational model’s terms) made its work ineffective – De Vaal (2003) quotes the US envoy to the Minsk Group John Mareska recalling that the Italian Chairman Mario Raffaelli’s request to have all the communication translated into Italian resulted in the same request from the French and German representatives; as a result, the Minsk Group transformed into a slow and absurd negotiation machine, composed of representatives of eleven countries, two regions, chairmanship, secretariat, and five translating boots.  Former Armenian Foreign Minister Vartan Oskanyan recalls (again in de Vaal (2003)) that in 1992-1993 the participants of the negotiations process were more preoccupied with the positioning of the parties around the table, with the place where the Karabakhi representatives should be sitting, with the placement of commas in the text.
The conclusion in the terms of this model would be as follows: mediation attempts of the CSCE/OSCE and OSCE Minsk Group were resultants of the meeting/consultation-based working procedures of the organization, the main goal of which was to appear as a potent peacekeeper and to disallow the re-starting of the war.  When faced with failure to deal with the conflict via its main procedure – Minsk Conference, in this case – OSCE incrementally shifted towards the more dynamic Minsk Group format; the frequent visits, consultations, and proposals summarizing those are examples of this new program, which did not achieve tangible successes due to the OSCE’s organizational incapability of taking into account the internal political configuration of the parties involved in the conflict.


Rather a different picture we see when we analyze the mediation attempts by the Russian Federation.  Here we again have a loosely coupled system, with main actors being the Minister of Defence Grachev and Russia’s representative to the OSCE Minsk Group, Ambassador Vladimir Kazimirov.  The Karabakh problem is factored amongst them: Grachev is dealing with military aspects of it, while Kazimirov dealing with the diplomatic facet.  The power is also fractionated: Grachev’s power as of the Minister of Defense is explicit during the periods before the ceasefire, while Kazimirov’s power of a diplomat with a mandate to deal with the issue becomes vivid especially in the framework of the Minsk Group.  Grachev’s parochial priority was the deployment of Russian military force in the region (particularly, in Lachin corridor), while Kazimirov is more preoccupied with Russia’s leading diplomatic role especially in international frameworks. 
Grachev’s mediation actions are organizational outputs, specific and characteristic for the ‘rituals’ of the former Soviet Union military system – September 1992 meeting of Ministers of Defense of Armenia and Azerbaijan in Sochi, as well as the May 1994 meeting in the same format in Moscow, during which Grachev was trying to push a truce, using his informal superiority in the military hierarchy.  Grachev’s goal – the set of constraints defining the acceptable performance – was the deployment of Russian troops as peacekeepers in the region; his standard operating procedure was to push, with his ‘military’ straightforwardness – the deployment of Russian troops as a decision by the three Defense Ministers.  Grachev’s program based on this standard operating procedure was more than simple – to proceed pushing the deployment of forces regardless of what had been achieved on the diplomatic frontier – this is why his last attempt of imposing this decision took place after the Bishkek ceasefire: as de Vaal (2003) mentions, thus Grachev simply chose to ignore the fact that a ceasefire had already been achieved in Bishkek.  Thus Grachev also showed his inability of organizational learning – he could not generate new programs and standard operating procedures to push the deployment of Russian troops in Karabakh, provided that a ceasefire had already been achieved.  In twenty months that divide Grachev’s two mediation attempts, he showed no signs of even incremental organizational change – his proposals of 1994 hardly differ from those of 1992.  The concluding inference on Grachev’s mediation attempts would be as follows: September 1992 and May 1994 mediation actions were resultants of Russia’s Minister of Defense attempts to enforce deployment of Russian troops in Karabakh as peacekeeping forces.


The case of Vladimir Kazimirov is far more interesting; a career diplomat, who became very much associated with Karabakh conflict’s peaceful resolution attempts.  His goal, both in his capacity of the Minsk Group Co-Chairman, and as simply of Russia’s representative was to make sure that the ceasefire and the long-term resolution of the conflict will be results of Russia’s attempts, and Russia would increase its political influence in the region.  Kazimirov’s standard operating procedure was negotiations in any format – be these face-to-face negotiations, shuttle diplomacy, or even ‘fax’ diplomacy – a term coined by Kazimirov himself (Kazimirov, 2005) to describe the process by which the Bishkek Protocol was sent to different parties by fax to get their signatures.  Kazimirov’s programs of action were much more flexible, allowing him to actually achieve his goal – Kazimirov was instrumental in achieving a ceasefire in Bishkek.  Kazimirov showed his ability of organizational learning by the mere fact of organizing the ‘fax’ diplomacy process, when the rush for signing the document as soon as possible would not afford time to send it sequentially to all the missing parties.  Kazimirov’s actions were resultants of the organizational routines and organizational learning of Russian diplomacy in a new and dynamic post-soviet environment, aimed at ensuring the leading political role of the Russian Federation in the former Soviet Union.
The last significant player to be considered through the lenses of organizational paradigm is Iran; its goal was to achieve reconciliation between the two neighbors, thus getting political influence in the region.  The standard operating procedure and the program based on it was facilitation of a meeting between the leaders of Armenia and Azerbaijan and, in the potential future, shuttle diplomacy by the Special Representative of the President of the Islamic Republic of Iran.  Nothing can be said about organizational learning, incremental organizational change, and parochial priorities and perceptions in this case, for this was just a one-time mediating attempt, which ended before actually starting – on 9 May 1992 the local Armenian forces seized the fortress of Shushi on the same day the Communiqué was signed.  Thus, it is hard to make any organizational inferences about this case.
Governmental Politics Model – Through the lens of this paradigm, we try to understand which bargaining processes among which players and how resulted in particular mediation attempts; the mediation attempts are seen as political resultants, being generated by a conflict, misunderstanding, and/or compromise among players involved.  As mentioned above, the characteristic process in this model is bargaining.
The most important players in the mediation game here are as follows:

  • OSCE key member states involved – Russia, USA, and the OSCE structures;
  • OSCE Minsk Group chairmen and co-chairmen;
  • Russian Defense Minister Grachev;
  • Russian Envoy Kazimirov; and
  • The President of Iran Hashemi-Rafsanjani.

Let us consider the factors determinant for the players’ stands: the parochial positions and perceptions, goals and interests, stakes, and deadlines and different faces of issues.  Russia’s parochial position (as part of OSCE) was its claim to play a key role in the former Soviet Union region, and president Yeltsin’s administration’s perception of Karabakh conflict was as of an arena where Russia can both successfully mediate and use the opportunity to ensure military presence.  Subsequently, Russia’s goal was to be the party who would resolve the conflict, with the interest attached to this being the main troubleshooter’s role in the post-soviet space.  On the stake was post-Soviet Russia’s newly emerging role as of the de-facto successor of the Soviet Union.  The deadline here was rather interesting – it was relativistic: both for Russian administration, and for the United States the deadline was the moment when the other player would have been too close to becoming a successful mediator.


The position of the United States within and as a part of the OSCE framework was again determined by its geopolitical agenda: attempting to counter Russia’s influence in the South Caucasus in the context of its oil interests in Azerbaijan – as John Mareska notes in de Vaal (2003), for these ends Russia used its covert threat to increase its assistance to Armenia, to pressure Azerbaijan.  The US goal was the same as Russia’s – to be the conductor of the conflict resolution, thus claiming the role of the new player in Caucasus; the interest of the US was control of the oil supplies from the Caspian, and, more generally, control over infrastructure and communication roots in the region.  The parochial position of the OSCE/CSCE was determined by its claimed role of the primary political arena in the post-Soviet space.   The goal of the OSCE/CSCE was again the resolution of the conflict, with the interest being its own – establishment of the organization as the key decisionmaker in the Eurasia region.  On the stake was the organization’s credibility as the key forum for decision making in the newly formed post-Soviet Eurasia; the many deadlines the OSCE/CSCE faced were tied with the Russia’s mediating interventions – each party tried to outrun the other.  The conflict’s face for the organization could be the one of a challenge to its troubleshooting ability.  Therefore, the OSCE/CSCE mediation attempts were political resultants of bargaining between Russia, US, and the OSCE structures, each of these agencies following their own parochial goals, rather than a common organizational objective.


What can we say about two other important players – Minister of Defense Grachev and President of Iran Hashemi-Rafsanjani?  Grachev’s parochial position was determined by his perception of the necessity of retaining the Russian military presence in South Caucasus.  His goal was to inscribe the deployment of Russian peacekeepers to any conflict resolution recipe.  Grachev’s apparent interest was in getting political credit and leverage from being able to ensure Russian military presence in South Caucasus.  Grachev’s actions and attempts to inscribe the Russian peacekeeping to the final conflict resolution formula impacted the Russian mediation as a political resultant – the resolution was promoted within a CIS – not OSCE – framework.
President Hashemi-Rafsanjani’s parochial agenda was determined by both the presence of a significant Azerbaijani factor in the country on the one hand, and the desire to keep the positive relationships with Armenia, on the other.  Succeeding in peacemaking would have meant for Iran a significant increase of its role in South Caucasus and ability to counter the growing US influence; Rafsanjani’s interest could be both political credit and greater role in the region.  On the stake was Iran’s role in South Caucasus; the different faces that the conflict had for Iran were determined by the Azerbaijani minority in the region on the one hand, and the US-Azerbaijan close links, on the other.  The governmental politics model helps little in explaining Hashemi-Rafsanjani’s mediation attempt as a political resultant – what can be hypothesized (and this contention definitely needs separate consideration and valid proofs) is that the Iranian president’s mediation attempt could be a political resultant of bargaining between the interests of Azerbaijani minority in Iran and countering the US potentially growing interests in the region.

Conclusions and Recommendations

The conceptual model applied in this research and borrowed from Allison’s classical Essence of Decision cannot claim to provide with comprehensive and multi-faceted insights about the case; its three paradigms are not mutually exclusive – many concepts, such as parochialism, appear in more than one of them.  Nor these paradigms are comprehensively exhaustive.  However, this conceptual construct illuminates the most important aspects of political decision making – strategic, organizational, and political. 
The brief scrutiny into the two periods of the conflict showed as the following: the mediating parties always had an agenda specific to them, and only to them.  Indeed, let us consider the inferences we deduced, applying the Rational Actor, Organizational, and Governmental Politics paradigms: in terms of strategy, in 1988-1992, all the mediating parties, be that the Politburo, Moscow emissaries in Karabakh, or even presidents Yeltsin and Nazarbayev, had in mind the preservation of the existing borders in order not to challenge the current structure of the state or in order not to sparkle similar conflicts elsewhere.  Organizationally, the mediation attempts of this period lacked the flexibility and dynamism they needed to match the pace and complexity of the conflict’s development. From the political point of view, the mediating attempts of this period were resultants of the Gorbachev’s struggle to retain the USSR, power struggle within Politburo, and claims for a leading role after Gorbachev’s era de-facto ended in late August 1991.  This is why these mediation attempts did not produce any tangible results – indeed, being resultants of strategic, organizational, or political processes within the gradually collapsing Soviet Union, they did not incorporate the real desires of the conflicting parties.


The picture is similar when we consider the agenda of the mediating parties in the period of 1992-1998.  Indeed, in terms of strategy, CSCE/OSCE and its Minsk Group was trying to author a feasible solution, thus instating itself as the main troubleshooting organization in the post-Soviet Eurasia; Russia tried to claim a privileged role in the post-Soviet region; Iran attempted to enter the South Caucasus arena as a regional power.  Organizationally, CSCE/OSCE mediation attempts were resultants of rather inflexible organizational routines, based on communication reinforcement – meetings, consultations; Russian mediation attempts were resultants of routines of Russian Defense and Foreign Service systems, with the aggressive straightforwardedness of the former and dynamic pro-activity of the latter.  As already mentioned, the organizational paradigm is incapable of providing with insights about such a one-time mediation attempts as the Iranian one.  From the viewpoint of political processes, OSCE/CSCE mediation attempts were political resultants of bargaining between Russia, US, and the OSCE structures, each of these agencies following their own parochial goals, rather than a common organizational objective; Russian mediation was a political resultant of the non-coordinated and competitive struggle between the Foreign and Defense Ministries; President Hashemi-Rafsanjani’s mediation action could have been the political resultant of internal pressures of the Azerbaijani minority and Iran’s leadership’s desire to counter the growing American influence in the South Caucasus, determined by the close US-Azerbaijan relationship.
Another important point is stunning by its absence – all the mediating parties were states, officials, or inter-governmental organizations – Karabakh conflict is notable with the fact that civil society mediation attempt have never had any significant impact.  Meanwhile, the civil society has certain advantages over state or inter-state players – the civil society does not have strategic geopolitical goals, which any state has and which inevitably affect the state or international organizations’ mediation.  The civil society is rather flexible in terms of organizational standard operating procedures and routines.  Finally, the civil society is less disposed towards parochialism and political bargaining, because the stakes within the civil society are not as high as in the states or international organizations.
What are the possible inferences or recommendations, based on this analysis?

  • Any mediating party must make sure to inscribe or incorporate the goals and interests of the conflicting parties into its own agenda, if the mediation is to be successful;
  • Any mediating party shall ensure that its standard operating procedures, organizational programs, and repertoires allow for at least as much variety, as the dynamism of the conflict itself;
  • Any mediating party should attempt to parcel out the problem to as few of its agents, as possible, for the parochial perceptions positions inevitably complicate the subsequent bargaining within the mediating party; and
  • Civil society mediation – be this the civil society organizations of third nations or joint civil society initiatives of the parties involved – have the advantage of neutralizing any strategic considerations’ influence over the mediation process, are more flexible organizationally, and do not have (as a rule) any political agenda; the final recommendation, therefore, incorporates the previous three.

This analysis of mediation attempts is not full by definition – for more than a decade after 1998, OSCE Minks Group worked actively to generate and promote resolutions to the conflict.  The summit at Key West, the Prague process, initiated in 2004 by the Armenian and Azerbaijani foreign ministers, and the Madrid proposals are probably the most important benchmarks on this way.  However, the conflict is still unresolved, and hopefully the recommendations presented above could be of a certain usefulness for those involved in the process.

Works Cited

Allison, G. T. (1971). Essence of Decision: Explaining the Cuban Missile Crisis. Boston: Little, Brown and Company.
Bonoma, T. V. (1983). A Case Study in Case Research: Marketing Implementation. Working Paper 9-585-142. Boston: Harvard University Graduate School of Business Administration.
de Vaal, T. (2003). Black Garden: Armenia and Azerbaijan through Peace and War. New York and London: New York Univeristy Press.
Gorbachev, M. (1996). Memoirs. London: Doubleday.
Institute of Problems of Commerce NGO. (2004, September 24). www.ippnou.ru. Retrieved February 27, 2009, from http://www.ippnou.ru/biography.php?idarticle=000356: http://www.ippnou.ru/biography.php?idarticle=000356
Merton, R. (1968). Social Theory and Social Structure. New York: Free Press.
Yin, R. K. (1984). Case Study Research. Beverly Hills: SAGE Publications.
Zolyan, S. (2001). Nagorno-Karabakh: The Problem and The Conflict. Yerevan: Lingva.

The detailed description of the analytical paradigm can be found in (Allison, 1971, pp. 32-35).

The detailed description of the organizational process paradigm is presented in (Allison, 1971, pp. 78-96)

Detailed description of the Government Politics model is presented in (Allison, 1971, pp. 163-181)

The conceptual vocabulary of this chapter is borrowed from Graham Allison (1971).

Formally, this was the highest decision making body of the Soviet Union; however, throughout the history of the Soviet Union, the Presidium was mainly a formal body, which approved as state decisions the policies and decisions made by the Communist Party leadership.

The conceptual vocabulary of this chapter is borrowed from Graham Allison (1971).

The conceptual apparatus used in this section is borrowed from (Allison, 1971, pp. 162-183)

Whilst first secretaries of the Central Committees of the Communist Parties of the former USSR republics usually represented the titular nation of a given republic, second secretaries, like Polianichko, were usually Russians, who generally acted as representatives of Moscow in republics.

The Conference on Security and Cooperation in Europe (CSCE) was created in 1973, as a ‘multilateral forum for dialogue and negotiation between East and West’ (Organization for Security and Cooperation in Europe, 2009); from 1995, it has been renamed in an Organization for Security and Cooperation in Europe (OSCE).  OSCE is the largest regional security organization in the world, with 56 States from Europe, Central Asia and North America (Organization for Security and Cooperation in Europe, 2009).

Conference on Security and Cooperation in Europe Council of Senior Officials.

We refer to the players here as to the countries, which is characteristic to the Rational Actor model; however, let us not forged that here we speak of actors within the framework of the OSCE and its Minsk Group.  Therefore, speaking of Russia and US we speak within the context of Governmental Politics model. 

This recommendation is actually the paraphrase of Ashbey’s law of requisite variety – see Ashbey, R. (1957) An Introduction to Cybernetics, London: Chapman & Hall.



Call for Expert on Policy Communication and Instruments

This announcement is available only in Armenian. 


“Green light” for environmentally neutral business development from Lisbon to Vladivostok

On March 3, 2021, the first meeting of the GreenDeal Task Force created under the Initiative Lisbon-Vladivostok was held. In the videoconference format, more than twenty authoritative experts in the field of ecology and business from Austria, Armenia, Germany, Italy, Kazakhstan, Russia, France, as well as the representatives of the largest business industry associations supporting the Initiative Lisbon-Vladivostok, discussed common approaches to harmonizing the activities implemented by the EU and the EAEU on the path to sustainable development, including a radical reduction in greenhouse gas emissions into the atmosphere by 2050.

 more >>



Youth Expo Catalogue

The Publication is available only in Armenian.