COLLECTIVE VICTIMHOOD AND MEMORY IN AZERBAIJAN

COLLECTIVE VICTIMHOOD AND MEMORY IN AZERBAIJAN: THE CONSTRUCTION OF NATIONAL IDENTITY THROUGH THE NAGORNO-

KARABAKH CONFLICT AND ITS POLITICAL IMPLICATIONS

by Emma Pérez

A thesis submitted in partial fulfilment of the requirements for the degree of

Master of Arts in Eurasian Studies

at

NAZARBAYEV UNIVERSITY -

SCHOOL OF SCIENCES AND HUMANITIES

2023

Abstract

The present thesis examines the transmission of memory and political implications of victimized identity in Azerbaijan. It focuses on the strategies of victimization that shape society’s discourse of Nagorno-Karabakh, which in turn underpins national identity.

Collective remembrance strategies place Azerbaijanis at a point of impasse in relation to the conflict and pose a further obstacle to post-war peace negotiation, thus contributing to the intractability of the conflict. Using ethnographic interviews and incorporating evidence from official government bodies, this study demonstrates that the Azerbaijani discourse on the Nagorno-Karabakh conflict has been articulated through strategies of self-victimization that are transmitted through collective memory. These strategies function in two directions:

selective and biased information-processing, justification and rationalization of negative group behavior, self-pity, and denial are directed at the in-group; whereas attribution of blame, moral superiority and paternalism, and moral disengagement are addressed to the out- group. By analyzing the disparate treatment of the Sumgait pogroms and the Khojaly

massacre in collective memory, this study offers empirical evidence that the Azerbaijani discourse selectively focuses on particular events and builds on the memory narrative in a way that fits its self-victimization framework. The institutionalization and widespread

remembrance of the Khojaly massacre further solidifies this aspect. This paper further argues that the government instrumentalizes victimized memory abroad as a political currency and domestically to divert attention from its own wrongdoings, such as lack of freedoms. Finally, it points out the importance of memory in conflict resolution and that a young society

becoming more and more disaffected by protracted conflict can lead to disengagement, instead of the active participation of civil society needed to work towards a stable and lasting peace.

Contents

Introduction ... 4

Literature review ... 9

Research Methodology ... 24

Strategies Of Victimization In The Formation Of National Identity ... 37

The Collective Remembrance Of The Sumgait Pogroms And The Khojaly Massacre In Azerbaijan ... 58

The violence in context: Historical background of the Sumgait pogroms and the Khojaly massacre... 61

The government’s tale ... 64

The peoples’ tale ... 70

Conclusion ... 83

Bibliography ... 89

Appendix ... 94

Introduction

This section establishes the knowledge necessary to understand the context and significance of the study. It sets out the antecedents of the Nagorno-Karabakh conflict, placing emphasis on its longevity and the intricate forces at play. By examining the historical and socio-political environment, I intend to provide a solid basis for arguing that there is a need for further assessment of the Azerbaijani stance on the conflict. I will also outline the precise research objectives of this study. My main objective is to study the memory

transmission and political implications of victimhood identity in Azerbaijan. I intend to demonstrate how in-group and out-group self-victimization tactics are transmitted through collective memory using ethnographic interviews and data from official government agencies.

This section will lay the groundwork for the later analysis and conclusions.

On the morning of September 27, 2020, Azerbaijan launched an offensive in Nagorno- Karabakh in order to reclaim the southern territories. Nagorno-Karabakh is an ethnic

Armenian enclave within the territory of Azerbaijan, to which it belongs de jure, but which is governed by the unrecognized Republic of Artsakh since the First Karabakh War that ended in 1994. This latest attack led to six weeks of war in which many people died from both sides.

Azerbaijan took over the territories that Armenia had previously occupied, but also held a significant part of Nagorno-Karabakh itself. After the capture of Shushi, a city located inside the disputed territory and of great strategic value due to its proximity with the enclave’s capital Stepanakert, Russia mediated a ceasefire that was signed by both Azerbaijani

President Ilham Aliyev and Armenian Prime Minister Nikol Pashinyan. The 2020 War, also called 44-day War, was the second full-scale war of the Nagorno-Karabakh conflict that started during the collapse of the Soviet Union, and from then until now hostilities have not ceased. The conflict in Nagorno-Karabakh between Armenians and Azerbaijanis, often

referred to as intractable, has caused bewilderment to many scholars, who have turned to numerous disciplines to try to find a solution to this seemingly hopeless “frozen conflict”.

My intention is to go beyond the characterization of “frozen conflict” to understand what contributes to the conflict remaining intractable and why reconciliation has not been possible. It is important that more studies be conducted on the inhibitors of peace between Armenians and Azerbaijanis from standpoints other that geopolitics or international relations, since, despite numerous studies and attempts to negotiate, tension still exists and the situation is extremely volatile. Peace in Nagorno-Karabakh is not only imperative for the Armenian residents of the territory to stop living under the threat of ethnic cleansing, but it is also necessary for the Armenian diaspora, who also conceive this conflict situation as a matter of survival of a community that already experienced genocide in 1915 under the Ottoman Empire during the First World War. Peace is also an urgent matter for many young

Azerbaijanis who have inherited a “problem” that is not theirs and of which they are growing increasingly jaded. The dangers of a young society becoming more and more disaffected by protracted conflict can lead to disengagement instead of the active participation of civil society needed to work towards a stable and lasting peace.

My study is located at the nexus of collective memory, collective victimhood and national identity, demonstrating that the Azerbaijani discourse on the Nagorno-Karabakh conflict, shaped by collective victimhood, contributes to the understanding of the dynamics of the conflict and poses an obstacle to the resolution of the conflict and peace negotiation between the two sides. I argue that Azerbaijan has articulated its narrative on the Nagorno- Karabakh conflict, which underpins the basis of its national identity, using strategies of victimization that are transmitted through collective memory. These strategies function in two directions: selective and biased information-processing, justification and rationalization of negative group behavior, self-pity, and denial are directed at the in-group; whereas attribution

of blame, moral superiority and paternalism, and moral disengagement are addressed to the out-group. Using the Sumgait pogroms and the Khojaly massacre as cases, I show how Azerbaijanis articulate the memory narrative to fit their self-victimization framework. In Azerbaijani society, the conflict is understood and framed in a manner that makes its own group the victimized. This way of presenting the conflict significantly affects how

Azerbaijanis justify and rationalize in-group behavior towards the out-group, how they select and process information about the events that occurred and how they construct the narrative on the conflict.

Collective victimhood in Azerbaijan primarily serves two functions. The government instrumentalizes strategies of victimhood abroad to carry out its foreign policy, especially with respect to the military conflict with the Armenians, framing it as the number one national security threat. At the same time, the government also employs inward-facing victim

narratives domestically, as they serve to divert attention from the government's wrong-doings, such as corruption or lack of freedoms in the country, and thus assert political authority and maintain Ilham Aliyev in power. In addition, it helps Azerbaijanis understand their existence in the post-colonial period by creating a narrative that transcends Russia's involvement.

I consider my work as a contribution to the field of conflict and memory studies that adds a new layer to the intractability of the conflict, namely the self-perceived victimization of Azerbaijanis. While Nagorno-Karabakh remains an important topic in social and political studies, collective memory and victimized identity, as well as their social and political

consequences, are still overlooked issues. In that sense, I aim to offer insight into Azerbaijani victimhood-based identity and its political implications as representing a further obstacle to conflict resolution efforts., a perspective that has been neglected in the body of knowledge of discourses on Nagorno-Karabakh. My research seeks to bridge this gap by providing evidence on the treatment of the memory of the Khojaly massacre and the Sumgait pogroms. While

both events have been extensively studied, there is insufficient attention to how these

elements of the past contribute to the understanding of conflict dynamics today. In that sense, my research is unique in that it brings together and uses the various existing literatures on the conflict and in the field of collective memory and victimhood, as well as national identity, and evidences how Azerbaijanis shape collective memory using strategies of victimization

directed at the out-group and the in-group, which has a significant impact on their behavior towards the conflict. Hence, my study adds a new perspective to the scholarly conversation on Nagorno-Karabakh and expands the literature on memory and conflict, further underscoring the importance of dealing with the past in peace negotiations.

Furthermore, this research aims to contribute to Eurasian studies by demonstrating that in Azerbaijan the Nagorno-Karabakh conflict opens a space for post-colonial discourse. The systematic blaming of Russia for the Nagorno-Karabakh conflict by the Azerbaijanis only points to the colonial failures of the Soviet Union. That Nagorno-Karabakh has shaped Azerbaijani national identity and not so much its Soviet past hints at a break with the “post- Soviet” categorization attributed to it in search of an identity no longer dominated by Russia.

The thesis is structured as follows: I start by reviewing the academic literature on Nagorno-Karabakh from different disciplines and also delve into the literature on collective victimhood, stating why it is significant to study the conflict through this approach that has been neglected. In Research Methodology I describe the methodology I have used for my research, all the steps I have taken in detail to analyze the data and the logic behind all the decisions, as well as a brief reflection on the limitations of my study and address some ethical considerations I took into account during my fieldwork. This is followed by two empirical chapters: Strategies Of Victimization In The Formation Of National Identity and The

Collective Remembrance Of The Sumgait Pogroms And The Khojaly Massacre In Azerbaijan.

In the first one I examine the strategies of victimization used by Azerbaijanis and how collective victimhood shapes the discourse that underpins national identity. I identified seven mechanisms that work in two directions: selective and biased information-processing,

justification and rationalization of negative group behavior, self-pity, and denial are directed at the in-group; whereas attribution of blame, moral superiority and paternalism, and moral disengagement are addressed to the out-group. In the second substantive chapter, I reconstruct and analyze the disparate treatment of the memory of the Sumgait pogroms and the Khojaly massacre, demonstrating that the Azerbaijani discourse selectively focuses on particular events and builds on the memory narrative in a way that fits its self-victimization framework using in-group and out-group strategies. The government’s instrumentalization of the Khojaly massacre further solidifies this aspect.

Finally, I end with a concluding chapter in which I address the importance of

examining the Nagorno-Karabakh conflict through collective victimhood in order to deeper understand the underlying psychological dynamics that influence Azerbaijan’s attitudes towards the conflict and highlight the significance of reconciliation of the past between two communities for the resolution of the conflict.

Literature review

In this chapter I will elaborate a critical review of the literature on collective victimhood and also of the range of angles from which the Nagorno-Karabakh conflict has been studied. In establishing the scholarly conversation on this topic, my intention is to offer a deeper understanding of the existing knowledge on the conflict and to find the gap that I intend to address in order to advance the body of scholarship in the field of conflict and memory studies. More precisely, my main objective is to identify the limitations in

understanding the role of collective victimhood and memory in the Azerbaijani experience of the conflict.

Many different academic disciplines have discussed and attempted to better

comprehend the Nagorno-Karabakh conflict, including international relations and political science, security, peace and conflict studies, as well as interdisciplinary studies, such as sociology or memory studies. The former prioritizes analyzing, especially through case

studies, the conflict historically and geopolitically, looking at the factors that have brought the conflict to the current situation, what is its role in the security of the region or why peace negotiations have not been fruitful. Interdisciplinary studies have focused on exploring the socio-cultural dimensions of the conflict. None of these studies have explored the national identity of Azerbaijanis through collective victimhood. While scholars have focused on addressing collective memory and historical narratives within the framework of the ethnic conflict in Nagorno-Karabakh, the only studies conducted on collective victimhood focus on Armenia, mostly related to Turkey and the Armenian Genocide. While it is true that Armenia has achieved the status of historical victim through its campaign to recognize the atrocities they were subjected to as genocide, Azerbaijanis have also built their national identity based on self-perceived victimization. This phenomenon, however, has been overlooked in the academic literature.

In the following, I will evaluate the literature necessary to unfold my argument, which focuses on Nagorno Karabakh through the prism of ethnic conflict and historical narratives, and also collective victimhood and memory. This academic literature helps me to identify the possible cause of the violence, as well as the narrative patterns to which Armenians and Azerbaijanis have resorted since the beginning of the conflict and what is the role of

collective memory. These approaches are important because regardless of the events that have occurred, it is the perception of events that determines the behavior of societies in relation to the conflict and the warring party.

One of the most complex issues in territorial ethnic conflicts is the historical claims and competing narratives each party develop about the territory. The main Armenian

perspective draws on national identity and sees the conflict as a fight for survival justified by a common history of genocide, whereas the need of upholding international norms of

territorial integrity is emphasized in the major Azerbaijani stance on the conflict (O’Lear &

Whiting, 2008; Toal, 2019; Broers, 2019).

Although Azerbaijanis frame their territorial claim to Nagorno-Karabakh mainly in terms of international law, they also use certain historical claims to legitimize their presence in the territory, asserting that Armenians were brought by the Russian Empire from Persia.

Avdoyan (1993) attempted to shed light on this issue by drawing on ancient and mediaeval Armenian primary source materials, that proof that Armenian occupation in Nagorno-

Karabakh extends far back in time. According to the author, “the Armenians have been in the area now called Nagorno-Karabakh since c. 370 A.D., if not before” (citation needed). While the Russian government introduced Armenians from Persia into Nagorno-Karabakh after the Second Persian War of 1828, Armenians were already living in the area of Artsakh (the tenth province of the former Armenian kingdom) in the Upper Karabakh (“Nagorno” means high, while “Karabakh” is a word of Turkish origin meaning Black Garden) since before the 5th

century at the earliest, when the Armenian culture spread in Artsakh after the creation of the Armenian language by Mesrop Mashtotz. This territory, Upper Karabakh was inhabited by Armenians but also by Caucasian Albanians. The Lower Karabakh area, on the other hand, has been populated by the Azeris, descendants of nomadic Turkic tribes who migrated from Central Asia, since the 11th century (Rasizade, 2011).

Both Armenia and Azerbaijan have always been subject to the rule of other empires, including the Soviet Union. Many authors (Cornell, 2005; Suny, 1993; Altstadt, 1992; De Waal, 2013) have tried to shed light on the origins of the conflict and attribute the Nagorno Karabakh “problem”, like the rest of the conflicts in the Caucasus, to the political cartography of the early Soviet Union. The origins and nature of the conflict as well as its interpretation is also a matter of disagreement not only between the conflicting parties, but also among many scholars. What most do agree on is that these conflicts were not the cause of the collapse of the USSR, but rather a by-product. For Zürcher (2007), ethnic conflicts in the Caucasus were inevitable because of the ethno-federalist system built by the Soviets. Ethnic tensions were growing because of the different groups fighting for power and privileges granted by the autonomous republics within the union republics. When the Soviet Union collapsed, the fragmentation became more manifest. Hughes and Sasse (2001) point out that these conflicts arose from the collapse of communism and are, in part, an echo of past struggles. They add that the USSR regime cynically manipulated nationalism using “quasi-federal institutional devices”, and blame the Soviet border-making for the 1920 Karabakh conflict, as it failed to connect the Armenians with their homeland, which triggered the nationalist irredentist

sentiment in Artsakh (p. 29). Such is also the opinion of Kolossov (1999), according to whom the collapse of the USSR rendered evident the disjunction between the political organization of the territory and various ethnic identities.

After the October Revolution of 1917, the Bolsheviks seized power and inherited the territories of the Russian Empire. Members of the Caucasus section of the Communist Party met in 1920 to determine the status of Karabakh and assign it to Soviet Armenia, but soon after the decision was reversed and Karabakh was transferred to Azerbaijan under the pretext of its economic dependence with this republic. Karabakh kept its status as an autonomous oblast within the territory of Soviet Azerbaijan. However, Karabakh Armenians lost their link with Armenia, as they did not enjoy their national symbols or education and means of

communication in their native language. These unfavorable conditions resulted in the emigration of Armenians from the territory, while Azerbaijanis immigrated.

Eventually, after the Stalin era, Karabakh Armenians began to claim to be transferred to Armenia and finally the first ethno-political conflict of the Soviet Union emerged (Zürcher, 2007). Babayev (2020), however, notes that Karabakh Armenians were allowed to keep their cultural autonomy and adds that the reason for granting autonomy to Karabakh lay in the need to maintain peace between Muslims and Armenians as well as economic ties between Upper and Lower Karabakh. Following the decision of the Armenian population of Nagorno- Karabakh to hold a referendum and finally declare itself an independent state in 1992, war broke out between Armenia and Azerbaijan which lasted until 1994 and resulted in hundreds of deaths. The secession of the hitherto Azerbaijani-held territory was successful thanks to Russian military assistance and Armenian intervention. As a result of the first war, not only many people died, as Azerbaijan had also to accommodate around 610,000 internally displaced persons (IDPs) (Kuburas, 2011).

Determining the cause of the conflict is not such a simple task. Attention should be paid to factors that influence ethnic relations and can ignite violence, such as the historical past, cultural or religious differences, but also direct causes of aggravation of relations, such as national, territorial, linguistic or socio-economic factors. In economic terms, Azerbaijan

enjoyed lower living standards than Armenia during Soviet times, while at the socio-cultural level, Karabakh Armenians were in an unequal position because their cultural and educational ties with Armenia were severed. The ethnic make-up of the region also played an important role, with Armenians accounting for the majority of the population (Yamskov, 1991). The religious factor, however, does not seem to have been important in the conflict, and early research (Yamskov, 1991; Cornell, 1998; Vaserman & Ginat, 1994) rules out this element as influential in the case of Nagorno-Karabakh. While it is true that religion is an intrinsic component of Armenian identity, Cornell (1998) argues that, although Armenian and Azeri religious leaders supported the causes of their respective sides, they were never at the forefront of the movements and religion was never appealed to.

Geukjian (2016) agrees that the conflict is rooted in the legacy of the Soviet Union and its nationalities policy, but considers that other factors have contributed to the conflict, such as the geopolitical interests of regional powers, economic competition and historical

grievances, as well as the interference of external powers such as Russia, Turkey or Iran. The role of external factors, especially Russia, is also emphasized by Melander (2001), who notes that the violence flared up because of nationalism. Thus, although ethnic conflicts usually have a territorial dimension (Milanova, 2003), part of it must be understood in terms of national identity, which is based on historical ideas and collective myths.

Despite being “scattered”, Armenia is a strong nation whose culture and identity is well defined in historical and linguistic terms and also possesses a cohesive and productive diaspora; Azerbaijan, however, is rather “weak” in terms of national identity (O'Lear &

Whiting, 2008). O'Loughlin and Kolosov (2017) also pointed out Armenian’s heightened ethnic nationalism by examining Nagorno-Karabakh’s pantheon of national heroes, which in addition evidences their collective trauma. The Armenian nation, though dispersed, forms an

exclusive community that uniquely fights for its survival in the face of possible external threats, such as hostility from Turkey or Azerbaijan.

Both Armenians and Azerbaijanis have created different national discourses to claim their exclusivity over the territory and, in the case of Armenians from Karabakh, also the right to self-determination. In many cases, the different interpretations are conflicting and

opposing, which in many occasions has fueled violence. The “Karabakh problem” is deeply rooted in the past for both parties. The historical claims date back to three distinct periods: the nationalist movements in Armenia and Azerbaijan in the nineteenth century, the consequences of territorial decisions in the Soviet era, and the rise of nationalist sentiment again in the 1980s (Milanova, 2003). Since the Soviet era, in order to justify their aims, ideologues of the conflicting antagonist parties have decisively pushed back in time the question of when the conflict started, which has been drifting into the question of “we were here first”. Although this question generally concerns the legitimacy of the territorial boundaries determined by the colonial powers, the question of “who came first” has been raised by both sides as a

fundamental claim to ownership of the ethnically defined territory (Zürcher, 2007).

Tensions between Azeris and Armenians are further based on mutual distrust and rooted in hostile historical memories that have ended up forming an “us against them”

rhetoric that permeates institutions and media, to the point where they blame each other for any present tragedy. There are some differences in the nature of the claims of the warring parties. O'Loughlin and Kolosov (2017) argue that for Armenians the narrative on Nagorno- Karabakh focuses on national identity, while Azerbaijanis stress the importance of respecting international standards of territorial integrity. It seems more necessary for de facto states such as Nagorno-Karabakh to construct and “over-enhance” identities, surely as a reaction to the lack of legitimacy at the international level. This differentiation is also emphasized by

Gamaghelyan (2010), according to whom for Armenians the conflict is motivated by national

identity, because along with language, culture and history, Nagorno-Karabakh is part of their historical territory over which they consider to have the right to self-determination. Losing Nagorno-Karabakh would mean losing part of their identity. Moreover, it represents a national symbol that was liberated from the “Turks”, who have oppressed Armenians for centuries. Armenians associate Azeris with Turks and link them with the Genocide of 1915, so in their collective imagination Azeris are part of the “genocidal Turkish nation”. For Azerbaijanis, that Armenians occupy Nagorno-Karabakh is a violation of national heritage, and the war was a traumatic event in which they not only lost territory, but ended with a large number of victims and IDPs (Gamaghelyan, 2010).

The narrative about the genocide should not be overlooked, as it is not possible to understand the conflict between Armenia and Azerbaijan without integrating the discourse of

“genocide denial” that was produced in Turkey and later adapted by Azerbaijan (Cheterian, 2018). Moreover, Azerbaijan has developed its own state discourse on the genocide, categorically denying that the genocide took place and presenting Azerbaijan itself as the victim (Rauf Garagozov, 2012). Armenia, Azerbaijan and Turkey have referred to the events of 1915 to formulate their policies towards the conflict and the trauma of the past, although repressed, was present in the mass psychology of both conflicting sides (Cheterian, 2018).

Political elites have legitimized these narratives to implement their political agenda (Voronkova, 2013), but so have intellectual elites through different strategies of war and myth-making. Azerbaijan's victory in the Second Karabakh War in 2020 and the way it presented its victory as the unequivocal success of Ilham Aliyev's authoritarian regime have compromised the position of peacebuilders and democracy advocates, which was already fragile in itself. The discourse makers discursively transformed this contemporary conflict into a historical one with roots in antiquity, giving primacy to the primordialist ethno- nationalist vocabulary. Both Nikol Pashinyan and Ilham Aliyev adopted an historical

narrative of the conflict from radicalized positions, increased animosity and invested in war, thus becoming two of the most militarized societies in the world and portraying the other as the mortal enemy to be destroyed instead of learning to coexist with it and truncating opportunities for democratization and regional integration (Gambaghelyan & Rumyantsev, 2021). Indeed, the leaders are largely responsible for perpetuating the discourses, but they are not the only ones to blame. For example, anti-Armenian propaganda in Azerbaijan has been instigated by journalists as well as by commentators, historians or high-level officials, so that the negative image of Armenians has become institutionalized. The state uses the “enemy image” through mechanisms that embed it in society and help divert attention from the government's own failures (Novikova, 2012).

While not diminishing the responsibility of the authorities, it is necessary to mention how societies contribute to the perpetuation of the conflict because of the collective

remembrance of events. Even if the new generations have not experienced certain events, such as the Karabakh war or in the case of Azerbaijan the Khojaly massacre, it remains a painful issue for them because the collective memory is oriented towards social norms and social representations (Garagozov, 2016). Without some changes to historical narratives, it is impossible to effectively address a contentious past, especially in societies that are divided and increasingly alienated from each other (Cheterian, 2018). Collective memory can be the inhibitor of peace, but it can potentially become the opposite. The narration reconfiguration and modification of the collective remembrance of a conflict can contribute to peacemaking (Paez & Lui, 2015).

Since collective victimhood is a powerful psychological construct that can shape group identity and behaviors, it is a relevant approach that can help us unravel the complexity of the processes that influence the perceptions of individuals and groups in a conflict-ridden society, as well as their demands and policies. Moreover, because collective victimhood

intersects with collective memory, this approach offers insights into the selective construction of the memory narrative.

The relationship between memory and the identity of a society is not a new phenomenon. Already in the 1930s, the art historian Aby Warburg and the sociologist Maurice Halbwachs, moving away from theories that framed the collective memory in biological terms, elaborated their theories on “cultural memory” which conceives of behavior and experience within a social framework and which is transmitted from generation to

generation as well as through social practices (Assman, 1995). Memory has the capacity to transform individual memories into collective memories. Collective memories, in turn, are depictions of the past that shape collective identity, and for this to be the case, the memory must be shared and must define a broad group in society (Brown, Kouri & Hirst, 2012).

Likewise, members of society maintain collective identity by identifying with the past through collective memories (Gongaware, 2011).

While it is true that societies look to the past as a source for identification and unity and in order to make sense of the present, it is not so much the past itself, but the way in which it is built into the psyche of the nation that says the most about what a society is like today and why things are the way they are. Past experiences are often interpreted, crafted and spread through discourses that conform the memory if a nation. Traumatic past events are one of the most difficult to process and the way they are handled has serious implications in the present and the future. This is the case of the Nagorno-Karabakh conflict between Armenians and Azerbaijanis, who after nearly 35 years of conflict and two full-scale wars have not found a way to cease hostilities.

The sense of collective self-perceived victimhood is a crucial part of the narrative about shared discourse in conflict societies. This collective phenomenon, again, has dangerous consequences in conflict, as it can act as an inhibitor of peacemaking (Bar-Tal,

Chernyak-Hai, Schori, & Gundar, 2009). Azerbaijanis have come to deny and belittle

Armenian’s experiences even though they have suffered them too, or rather, precisely because they have suffered them and Armenians were the perpetrators. The victim status that Armenia has managed to secure through its campaign to classify Turkish crimes as genocide has not only given them some moral cover in the region when they in turn victimize their neighbors but also poses a risk to Azerbaijan. Because Armenians made them suffer, in the Azerbaijani psyche, they no longer have the right to be victims, as the political currency of victimization has been fundamental for regional politics.

Sometimes in conflicts both parties believe that they are the victims, and this self- perception is an inseparable part of the narrative that societies construct to make sense of the conflict and forms part of the collective memory. However, to better understand the conflict ethos and consider potential strategies to mitigate hostilities, it is necessary to identify what tools society uses to create victimization discourses and why they work the way they do.

Because collective victimhood is the result of collective violence towards a targeted group, the consequences of such harm shape not only the group identity, but also the interaction with other groups (Noor et al., 2017).

Groups experience a sense of victimhood that is often the result of an event that has affected a significant portion of that group. Just as an individual may perceive a sense of victimhood indirectly, people who belong to the group against which harm has occurred may consider themselves victims because of their membership in that group. This implies that the members of the collective share a set of beliefs that unite them and reaffirm their collective sense of victimhood or may even be the basis of their identification with the group, since sharing beliefs is one of the basic elements of group formation. The sense of victimhood consists not only of shared beliefs, but of a whole system of attitudes and emotions that behave with respect to the type of harm the collective has suffered, such as oppression,

humiliation or loss. Because, emotionally, the sense of victimhood is associated with feelings such as fear or anger, it leads to the urge to want revenge for the harm suffered. The sense of collective self-perceived victimhood is based on and reflects shared beliefs that serve as a foundation for the creation of a common reality and culture. Within this reality are located past traumatic events, which end up becoming intrinsic narratives of that social identity.

The sense of collective self-perceived victimhood is based on and reflects shared beliefs that serve as a foundation for the creation of a common reality and identity. Within this reality are located past traumatic events, which end up becoming intrinsic narratives of social identity (Bar-Tal, Chernyak-Hai, Schori, & Gundar, 2009). Shared beliefs often serve to educate people about the past and present through moral lessons, as in the case of the Hutus (Malkki, 1995), but also to identify enemies and have a frame of reference through which to interpret the conflict, which can lead to a kind of siege mentality in which the group believes that the world is against them (Bilali & Ross, 2012). Victimhood has become for some nations a desired status, especially in the context of conflict, to demonstrate that the ingroup has suffered more than the outgroup.

Noor et al. (2017) define this phenomenon as competitive victimhood and argue that its motivation is underpinned by the intergroup relations, which are inherently competitive.

Competitive victimhood is prevalent in historical asymmetric conflicts in which one party has inflicted considerable harm on the other, and sometimes the perpetrators are precisely those who claim victim status (Bilali & Dasgupta, 2012), as the sense of victimhood is necessarily related to the power of collectives, but in fact collectives that are more powerful militarily, economically or politically also perceive themselves as victims in the conflict (Bar-Tal, Chernyak-Hai, Schori, & Gundar, 2009).

In their study on anti-Semitism and competitive victimhood, Bilewicz and Stefaniak (2013) suggest that there is a positive correlation between ingroup victimhood and the

tendency to justify negative ingroup actions. The authors demonstrate a positive relationship between victimization (which they call “absolute”) and anti-Semitic feelings, thus suggesting that the notion of victimhood within the group impairs the relationship with other victimized groups (a historical victim does not feel responsible for other victims) as well as the attitude towards other historically perpetrator groups. The effect that Polish identification has on anti- Semitic beliefs is due to the narrative of victimhood. Additionally, competitive victimhood encompasses another form of moral dimension, more specifically regarding the moral significance of suffering and the perceived legitimacy of who qualifies as a victim. Even when a group experiences objectively more pain, the opposing group may still contest the legitimacy of each other's experiences and question whether the victim group itself was to blame for the suffering (Noor et al., 2017).

As I have already mentioned, societies elaborate discourses about past experiences to make sense of the reality around them but also, among other things, to reinforce the positive image of the group to which they belong (Paez & Liu, 2015). Thus, elements of the violent past that are crucial for the elaboration of outgroup victimization are “forgotten,” but other parts that are crucial for ingroup victimization “survive” (Green et al. 2017). But forgetting and the strategic selection of elements from the past-and on many occasions, also from the present, is only one of the mechanisms used to elaborate victimization discourses. Victims, because of their status as victims, often feel moral superiority and their right to receive the sympathy of others, they believe they are exempt from criticism, and this leads them to justify the actions committed by their group. This is common in societies that have experienced past conflict and also in societies that find themselves dealing with intractable conflict. Memories of group victimization help members to legitimize their actions but also to feel less sympathy for the suffering of the outgroup (Bilali & Ross, 2012).

The attribution of responsibility is usually towards the outgroup, although the

researchers Bilali and Dasgupta (2012) speak of three targets of responsibility, which include the ingroup, the outgroup, but also external factors. Blaming external agents is another mechanism for denying self-responsibility and it is usually the individuals who most identify with the group who strive to maintain the positive image of the group. As the conflict drags on, the harm becomes more intense and is perceived as more unjust and undeserved, so the self-perception of victimhood becomes more ingrained in the collective and cycles of violence continue through acts of revenge and preemption (Bar-Tal, Chernyak-Hai, Schori, & Gundar, 2009) to protect themselves from the outgroup they see as a threat (Bilali & Ross, 2012). The sense of victimhood serves several functions in the conflict-ridden society. On the one hand, it provides a frame of reference through which to understand the conflict, as it identifies foes and allies and provides information and explanations. It also serves as a moral justification and as a form of differentiation and superiority. The ingroup frames the outgroup in delegitimizing terms that illustrate the immoral and unjust acts it has perpetrated, and considers the conflict irrational. The sense of collective victimhood is also a good patriotic mobilizing agent (Bar-Tal, Chernyak-Hai, Schori, & Gundar, 2009).

The sense of victimhood serves several functions in the conflict-ridden society. On the one hand, it provides a frame of reference through which to understand the conflict, as it identifies foes and allies and provides information and explanations. It also serves as a moral justification and as a form of differentiation and superiority. The in-group frames the out- group in delegitimizing terms that illustrate the immoral and unjust acts it has perpetrated, and considers the conflict irrational. As mentioned above, the sense of collective victimhood can mobilize the population. Edward Said (2000) points out that the Palestinian self-determination project has not worked because of the lack of a powerful historical narrative, as opposed to the narrative presented by the Zionist movement in Israel. It is therefore relevant to mention

here also the campaign by Armenians for the recognition of the genocide committed by Turkey, which demonstrates the strength of national narratives.

Although perhaps more fragile as a state, having to rely on Russia for many affairs, Armenia exhibits the characteristics of a strong nation if we take into account its powerful historical narrative and its cohesive international diaspora. Azerbaijan, on the other hand, has failed for example to find an international status for the Khojaly massacre which they qualify as genocide, and also generally to position itself as the victims of the conflict, Armenia being the one who has received this status in the eyes of the international community. Although the narratives about Khojaly and the loss and occupation of territories in Karabakh may be strong collective memories, perhaps the difficulty in turning such narratives into powerful

mobilizing agents that achieve international status is due to the vulnerability of the nation, which may be the result of a lack of a cohesive diaspora, multi-ethnicity, multiple languages and alphabet changes, and, in general, the lack of a shared identity (O’Lear & Whiting, 2008).

As we can see, collective memory is a topic integrated in the literature on Nagorno- Karabakh, however, it focuses more on Armenia and it makes little to no reference to the role it plays in transmitting specific behaviors. Although the conversation on disparate historical narratives between Armenians and Azerbaijanis, as well as the importance of the

reconfiguration of the past in order to build a more peaceful present, are still relevant topics, it is necessary to improve our understanding of what attitudes are conveyed through memory, how is the memory narrative constructed, and what are the implications. A critical review of the literature on collective victimhood suggests that this focus is relevant to the study of conflicts, as it can reveal attitudes, behaviors and patterns that can help us uncover barriers to conflict resolution.

Because it is Armenia who gained its status as a victim, this perspective from the Azerbaijani side has been neglected and scholars have paid little attention to how collective

victimhood affects their discourse and, therefore, their identity and attitudes towards the conflict. Groups that consider themselves as victims may become morally entitled, seek retaliation, and will be reluctant to compromise. Yet, the literature has failed to explain how collective victimhood in Azerbaijan contributes to the longevity and intractability of the conflict. I fill that gap in the subsequent chapters by delving into the Azerbaijani societal discourse in order to better comprehend the mechanisms of self-victimization that people use to build the memory narrative and that ultimately hinder peace process.

Research Methodology

In this chapter I explain in detail the methodology followed in my research. As I mentioned in the previous chapter, my research aim was to find out what civil society discourses in Azerbaijan could tell us about the intractability of the conflict in Nagorno- Karabakh. I explored societal narratives through qualitative interviews I conducted in Baku in the summer of 2022. My data revealed that Azerbaijanis have articulated their narrative on the conflict, which shapes their national identity, using in-group and out-group strategies of self- victimization that are transmitted through collective memory. Using the memory of the Khojaly tragedy and the Sumgait pogroms as case studies, the data also proved that the Azerbaijani discourse selectively focuses on specific events and constructs the memory narrative in a way that fits its self-victimization framework, which is supported by the government’s intervention. My data analysis focused on axial coding, grouping related concepts together into categories and ultimately into themes that provided an explanatory framework for how Azerbaijanis understand their experience of conflict. I used state

narratives sourced from government agencies to contrast official and unofficial discourses and to understand how and to what extent victimized collective memory has been transmitted.

I will first explain in detail the research design, which is based mainly on ethnographic interviews but also on content analysis of newspaper articles, and I will also explain the key design choices, such as how I selected the newspaper articles I analyzed, why I chose broad and open-ended questions for the interviews, and the age range of the participants from 18 to 40 years old. Next I will describe the data collection process, in which I used NVIVO, a qualitative analysis software, to transcribe the interviews that I had previously recorded with my cell phone. I then expound the process of data analysis, in which I also used NVIVO to perform axial coding, extracting subcategories, categories and themes. I will finally reflect on the treatment of my data, as well as the methodological shortcomings and limitations, such as

the lack of time or trust with my participants, and how I have tried to mitigate the impacts of these. I conclude by discussing some ethical considerations I took into consideration to ensure that my participants were comfortable taking part in my study.

I conducted qualitative research for my thesis, focusing on exploring the intractability of the conflict in Nagorno-Karabakh and the potential explanations provided by the discourses of civic society. Azerbaijan being an authoritarian country, I also wanted to take into account the state discourses to know to what extent they resonate with that of the population and how they converge and interact with societal discourses. To achieve my research goals, I

conducted ethnographic interviews to gather data and relied on them for the bulk of my research. I conducted the interviews in Baku, mainly in English but also in Azerbaijani, and in this case another participant would help be by acting as a translator. Before conducting the interviews, I conducted a preliminary study in order to formulate the right questions. For this study, I performed discourse analysis on news articles in English sourced from the State News Agency (AZERTAC). I chose AZERTAC for this analysis because it is the only official source of the official government news, so in addition to releasing government statements, it publishes news articles, contains official documents such as orders and letters from the president and it is possible to find a wide range of information, including commemorative days or special projects such as “Armenian vandalism in photos” concerning the destruction of cultural heritage. Moreover, all this information is available in English. I considered that this website alone housed enough information for my purpose.

I performed a preliminary analysis of the texts in order to make an assessment of the content and to elaborate a series of categories about the persistent narratives in the discourse. I considered the narratives persistent when I saw them framed in several articles about various topics. For example, “Armenian aggression” is used on multiple occasions to talk about border skirmishes, an “ethnic cleansing operation” or a possible occupation; “native lands”

also constitutes a narrative because it is frequently used in different contexts, such as when talking about massacres, IDPs and refugees, or a military counterattack. I ascribed these categories to different fields of academic literature, which allowed me to give the narratives a formal framework through which to account for them.

The AZERTAC interface enables the selection of a date range to search for articles, but it is necessary to add a keyword to perform the search. I selected ranges of dates close to

hostilities of major and minor severity, in 2010, 2016 and 2020 and used the term “Karabakh”

as keyword for the search. The headlines guided me to see if the content of the article was going to be relevant, i.e. had something to do with the conflict. Once I selected the articles, I skimmed through them one by one, as they are generally short. I was interested in analyzing the official Azerbaijani discourse, therefore, articles about Azeri officials or institutions or official statements by the president were mainly what I looked for. If reading them I realized that the article was reproducing the opinion of, for instance, a non-Azerbaijani minister on the Nagorno-Karabakh conflict, I would discard the article in preference to others that reproduce the opinion of Azerbaijani officials. Articles about foreign political personalities and their position on the conflict were also relevant because they serve to legitimize the official discourse, but are not themselves part of the official discourse I wanted to take into consideration for my analysis in that moment.

Once I selected the articles I was going to analyze, I read them more carefully and looked for fragments that told me what the position of the discourse on the conflict is. Knowing the historical and socio-cultural context allowed me to know that in an article, what I would emphasize are expressions like “the return of the displaced Azerbaijanis” because I know that for Azerbaijanis, Karabakh has always been theirs and it is a place to “return”. In this sense the methodology had both a deductive and an inductive approach, since I knew where to look first but I was open to what else the discourse was saying or, sometimes, to what it was not

saying. For example, on several occasions the Khojaly massacre against Azerbaijanis was mentioned, but at no time is there was any reference to the Sumgait pogrom committed by Azerbaijanis against Armenians. Sometimes, what the discourse does not mention is also part of the narrative. In those cases where I noticed a lack of information, I took it into account to contrast it with the popular discourse.

I developed a number of categories that constitute different narratives. I elaborated these categories once I read a considerable sample of texts and I saw that several elements were repeated quite often. Although the analysis was preliminary, this assessment already shed light on the most recurrent narratives in the discourse. This research was extremely valuable because it helped me to explore the official discourse in a more superficial way to know what to expect and to have a general idea. But more importantly, it allowed me to elaborate the questions for my interviews. Choosing several articles about Karabakh on different dates when the conflict escalated was enough to have a big picture of the most important narratives about the conflict.

As mentioned above, to constitute a narrative it had to be persistent, not an isolated element. For example, I elaborated a narrative of “appeal to norms of sovereignty” because there are multiple references to territory. In this category would fit “territorial integrity”,

“occupied regions of Azerbaijan”, or “internationally recognized as part of Azerbaijan”. The issue of IDPs by refugees was also constant, and therefore I elaborated a category of

“migration and displacement” where I placed, for example, “the return of the displaced Azerbaijanis”. Other important expressions have to do with the status of Karabakh, thus

“territorial integrity” and “native lands” were also important. Much of the conflict in Karabakh has to do with the relationship Azerbaijanis have with Armenians. The way the discourse portrays them is also relevant. For example, “occupiers” or “invaders”, again, tells us that according to them Armenians are alien to that territory. In discourse analysis the data

is not only composed of words, but more broadly, of expressions or fragments, because it depends on the context within the text. Thus, the data collected could be simply words like

“IDPs” or “provocation”, but also longer phrases like “They (Armenians) threaten Azerbaijan with a new occupation”.

Different narratives can belong to several categories, for instance, “occupation” was often mentioned in the sense that a foreign people has occupied a territory that belonged to them before. In that case, “occupation” would be part of the category of “first occupancy”. On the other hand, “threaten Azerbaijan with a new occupation” would be part of the category

“security dilemma”, since, although occupation is mentioned, it has a different meaning:

while in the first case it refers to historical claims, in the second case the possible occupation of Armenia is used as a pretext for defense.

I elaborated eleven broad and open-ended questions based on the discourse analysis (see Appendix). I chose to approach this study through interviews because I considered that

people's narratives, beyond the literal meaning we can draw from them, can say much more if we take into account the way people talk about a certain topic or what resources they use to explain a situation. My intention was that the interviews, in a way, would speak for

themselves. Thus, I asked rather general questions of the type “Do you think Armenians and Azerbaijanis are enemies?” or “What can you tell me about the Shusha pogroms and the Khojaly massacre?”. This approach has advantages and disadvantages. The disadvantage is that, being broad questions sometimes the participants got a bit off topic and I had to resort to several follow-up questions to redirect them. At the same time, the upside is that broad questions allow participants to describe their experiences in a less restricted way. The more specific and concrete the question or the way it is phrased, the more the interviewee is being forced to make certain choices. In my case, I wanted to be open to everything the interviews could tell me.

For the interviews, I selected a sample of participants aged between 18 and 40. The only requirements to participate in the sample were to be of legal age and not older than 40 because I was more interested in knowing the narratives of those who have grown up after the first Karabakh war and who had not participated in those events. This was one of the most

important points, as people who did not take part or witnessed the beginning of the conflict or were too little to realize what was going on, only have the discourse to rely on. Their narrative about the conflict is based on the experiences and memories their relatives and acquaintances passed on to them, on history textbooks or official state discourses. Interviewing people this age, I could assess how the discourse and identity of Azerbaijanis has been shaped with the beginning of the conflict as a starting point. The sampling strategy for the study involved the snowball method: I contacted potential interviewees, and they recommended other

participants who might agree to participate in the study. I conducted 15 interviews with participants of various ages and backgrounds. The fieldwork took place in Baku, Azerbaijan, where I also conducted participant observation, which was crucial to better understand my work.

I transcribed the interviews with the assistance of NVIVO, a qualitative research software.

Once transcribed them, I used axial coding as a technique to analyze them, identifying concepts in subcategories, categories and ultimately in themes. Axial coding is a technique closely linked to grounded theory that allows data to be related to each other, as it groups and combines the concepts into more comprehensive categories (Merriam, 2009). I chose to analyze the interviews using this technique rather than using, for example, discourse analysis, because I was not as interested only in the lexicon, keywords or occurrences of certain words or expressions, but rather how participants understood their experience and what resources they used to convey it in words. According to Juliet Corbin and Anselm Strauss (2015), the purpose of axial coding is to identify and define the central categories and concepts that

emerge from the data and to establish connections between them. In grounded theory, categories refer to a phenomenon, which can be, for example, a problem or question, a specific event or occurrence significant to the participants. Subcategories are also categories, but they provide greater explanatory value. As coding proceeds, the concepts that make up the categories and subcategories become clearer. When analyzing data axially, we seek answers to questions such as why, how, when or where and what are the consequences. By answering these questions, the categories reveal relationships between each other and provide greater context and understanding of the phenomenon under investigation, as well as a better understanding of the dynamic and evolving nature of events (Corbin, J. & Strauss, 2015).

I developed the coding process in two phases, which in turn consisted of almost the same steps. In the first analysis, or the first phase, I identified “labels” line by line according to what the interviewee said, which constitute the subcategories. For example, if the interview said “We were friends with them and we shared our food”, I would label it as “friends” and

“sharing”. If the interview said “We lived in peace, we married each other and have babies” I would label it “peace” and “intermarriage.” Labels were not just single words, they were also expressions: I would label phrases like “The international discourse is that Armenians are the victims and the Azerbaijani past is not represented” as “victimhood” but also “lack of

international recognition.” After labeling all the interviews, I grouped the labels, i.e., the subcategories, into categories, which could well coincide with the subcategories. These categories were, for example, “peaceful methods”, “friendship in the past”, “victimhood”,

“Karabakh should be ours”.

Once I elaborated the categories, I performed axial coding to identify the most

representative and recurring themes. These themes are what ultimately constitute the common beliefs about the conflict, a whole system through which Azerbaijanis understand their

present, past and future. These themes that emerged from my data became representative of

common beliefs among Azerbaijanis about the conflict because they were consistently

repeated in interviews in one form or another and constitute common narratives. For instance, participants used different resources, opinions, and explanations that represented the common belief that “Nagorno-Karabakh Armenians should be Azerbaijanis.”, through opinions such as

“[Armenians] can stay [in NK] with an Azerbaijani passport and abide by the law” or

considering that “If they want to live in our lands they have to show us that they are worthy of it” and that “if they assimilate and become part of our society, [coexisting] will not be a problem for us.” Some other examples are: “We were friends”, which explains the narrative of living together in the past, in which Azerbaijanis do not understand why violence suddenly broke out if Armenians were treated well and were even friends; “We tried peacefully” is the belief that on the Azerbaijani side every attempt has been made to address the conflict peacefully; “Khojaly” sheds light on the deepest wound of the conflict for Azerbaijanis; and

“Russia is to blame” is the belief that, as small countries they are, Armenia and Azerbaijan are victims of big geopolitics.

In this first phase I identified the belief system through which Azerbaijanis understand the conflict, but I had to narrow it down further. In a second analysis I realized that the common thread of all the beliefs was victimhood, therefore I re-coded the interviews trying to identify how participants used victimhood to explain their existence. The coding process was similar as in the same as in the first phase, identifying categories through labels and grouping them into themes, that is, the strategies of victimization most used by the respondents. In this case, since I took into account the victimization strategies that I extracted from the literature, the categories were less flexible because I stuck to the literature. Thus, for instance, “They shouldn't feel afraid of Azerbaijanis” I would label as “lack of understanding”; “From now it depends on the youth of Armenia” as “detachment” and subsequently placed both under the category of “moral disengagement”. Finally, I divided the strategies according to their

directionality, that is, whether they are directed to the out-group or the in-group. This helped me assess better Azerbaijanis behavior and discourse and therefore their stance on the conflict.

Since collective memory plays a crucial role in victimhood, I also focused on recomposing, using the interviews, the remembrance of Sumgait and Khojaly from the Azerbaijani perspective. I analyzed how the participants used in-group and out-group mechanisms of self-victimization in their memory narratives, which revealed a difference in the treatment of events and confirmed that Azerbaijanis construct their understanding of the conflict in a way that fits their self-victimization framework. This further highlighted the importance of victimization strategies in the construction of collective memory.

Although the fieldwork and data collection were satisfactory, I can only reflect on the methodological limitations and shortcomings of my study. First of all, I would like to reflect about the language barrier. Many young Azerbaijanis speak English and are replacing Russian as their foreign language, and I was able to communicate easily with them, but this also could mean that I was having access to a select group that had specific education. However, the English-speaking respondents had various backgrounds and did not share a specific profile, as among them there were a photograph, a programmer, a PhD student or an artist. I do not think this affected the interviews, but interviews in Azerbaijani with the help of a participant- translator might have been affected. I conducted four interviews in Azerbaijani, three with veterans from the 2020 Karabakh War and one with a historian. The same participant helped me with all of them but had a better relationship with the veterans. I think the interview with the historian might have been more precise, while the interviews with the veterans might not have been as accurate. My believe is that maybe the translator used a different, more familiar language with his friends when translating my questions and this might have affected the way in which the participants answered. Nonetheless, since the three interviews were completely

different in terms of responses and approach, I do not believe the translation was biased or the interviews totally compromised.

I would have liked to have had a more gender-balanced sample, but for various reasons, including my personal safety, I had to stop looking for more respondents and settle for fifteen interviews, an adequate number for a qualitative study. Nevertheless, I am aware that fifteen interviews might not be representative. In total I interviewed 4 women and 11 men. The women showed, although having different opinions, much more thoughtfulness in expressing themselves and were more retrospective, adding to their experiences also their feelings.

Perhaps having more women in the sample would have changed the perspective of the analysis. Another point to keep in mind is what the interviews can and cannot say. One of the limitations, due to lack of time, was building trust with the participants, something I was not able to do well. This ensures that the interviews are more genuine and interviewees are less concerned about what I think about their opinion or position in the conflict, as well as paying less attention to language because they feel more relaxed when they are in trust.

On several occasions I noticed that the participants did not quite understand what my role was and I had to clarify it numerous times. Because of the limited relationship with them, I think many of the participants were self-censoring so that I would not have a bad image of them and Azerbaijanis in general. In other words, they saw these interviews as a way to give them a voice and show their truth. The opinion or the way they talked about Armenians did not coincide, in some cases, with what they had often read on social media, for example. The participants did not use discriminatory language beyond considering them enemies or

occupants. I expected them to be more visceral and more explicitly show hatred or rancor, something inherent to ethnic conflict, but it did not happen in my interviews, except for one specific instance. It is possible that my participants were simply not that type of person, but we cannot rule out the possibility that, again, they were restraining themselves.

Adopting a broad, general questioning approach was a good decision to avoid bias.

Sometimes, the participants themselves are very careful about their lexicon and what they do or do not want to say and do not pay attention to the resources they use to tell their

experiences. My interpretation of the interviews did not take into account gender or lexicon as such, but ultimately focused on what the interviews could say, which revealed that

Azerbaijanis have constructed their identity using the Nagorno-Karabakh conflict as a reference through strategies of victimization.

I believe it is also relevant to give an overview of my research progress and process.

The first step of my research was to prepare for the interviews, which included building a questionnaire and find enough participants. The first step in my research was to develop the right questions, for which I used the Azerbaijani State News Agency (AZERTAC) to conduct discourse analysis. This process took place in March and April. I had to perform this task in order to present my research to the ethics committee that approved my fieldwork. In the summer semester, from June through August, I focused on interviews. I spent from mid-June to mid-July recruiting participants and doing the paperwork with my school for my three- week field trip to Azerbaijan. I traveled to Baku on July 22 and conducted most of the interviews the first week. The second week I visited the Military Trophies Park, a museum dedicated to displaying tanks, mines, uniforms and other Armenian military equipment

“captured” in the Second Karabakh War. I conducted more interviews that week and the last interview took place on August 8, one day before the end of my fieldwork trip. All interviews were face-to-face at locations chosen by the participants and recorded on my cell phone with the consent of the interviewees. Since the data analysis process took place in several steps and parallel to the writing of the thesis, I was working on it from about September until January.

Since January I focused mainly on the written work.

Finally, I believe that I should address certain ethical considerations that I took into account when conducting this research. The Nagorno-Karabakh conflict is a sensitive issue for both Armenians and Azerbaijanis. It is also a topic that can cause bias for researchers, who are no strangers to the atrocities that are committed in a war. However, I did my best to not let my opinion on the conflict affect either the process or the results of my research. Regardless of my position on the conflict, my task as a researcher was to avoid any bias in order to obtain the most genuine responses from the participants. My participants came from different

backgrounds, and whatever their opinion on this conflict, they have had to see their country fighting a war. Some of them would even tell me how their relatives or acquaintances had suffered the consequences.

The most important thing was to take into account the safety and confidentiality of my participants. For this reason, in order to guarantee their anonymity, I did not reveal any of their names or any other identifying information. To ensure that participants were making an informed decision to participate in my study, prior to the interviews I obtained their oral or written consent, where I explained to them the possible risks or benefits of their participation.

In addition, to be completely transparent with them, I explained what their involvement in the study entailed and how their data would be treated. None of the participants had any problems with this.

I also wanted to make sure that the interviewees felt comfortable and could speak freely, so I asked everyone to choose a location that suited them best to conduct the interviews. To my surprise, all the participants chose crowded places, such as cafes or the office where they worked, and did not mind that other people could hear what they were saying, no matter how dissenting their opinion might be. This made me reflect on the role of popular discourse on Nagorno-Karabakh in Azerbaijan and how important it is to the people,

so much so that there is room for dissenting opinions, as long as they do not compromise the government's plans.

Strategies Of Victimization In The Formation Of National Identity

In this chapter I attempt to shed light on the victimization strategies Azerbaijanis use to make sense of their experience. I use national identity and collective victimhood as a theoretical framework to explore how they converge, that is, to find out why and in what ways victimhood shapes or can become the basis of national identity, especially in tumultuous moments of history. This chapter establishes my main thesis statement, that Azerbaijanis use in-group and out-group victimization strategies to elaborate their discourse on Nagorno-Karabakh and to frame their memory narrative. Taking into account my typology of inward and outward functions, I identify seven victimization strategies that I try to

illustrate with empirical evidence from my data, which reveals instances and patterns of self- victimization.

As I discussed in the first chapter, victimization is often attributed to Armenians, especially related to the Armenian Genocide, and collective victimhood in Azerbaijan has been overlooked. Overall, this chapter contributes to a deeper understanding of collective victimhood in Azerbaijan and its importance in the formation of national identity, how it is portrayed in discourse and its political implications. This chapter also serves as a foundation for the following analysis of the cases of the Khojaly massacre and the Sumgait pogroms that I will elaborate in the next chapter, where I illustrate how in-group and out-group strategies of self-victimization are used to elaborate the memory narrative.

The fall of the Soviet Union and the subsequent regime change and state-building were significantly conflictive processes. Between 1988 and 2005, forced to decide their future political status, borders and identity, some Soviet states became embroiled in wars that have not yet found a resolution. During the turbulent years preceding up to 1991, when events in distant Moscow determined that Azerbaijan would break away from the Soviet