Obliged to hate: The successful stigmatisation of anti-war voices in Azerbaijan

Abstract

In International Relations (IR) scholarship, there is a growing body of research on the connections between emotions, stigma, and norm violations. It is often presumed that for stigma imposition to be successful, norm violators should feel shame. We argue instead that the emotional dynamics that inform the management of stigma are more complex and involve overlooked emotions such as anxiety, sadness, and hopelessness. We substantiate this by analysing the successful stigmatisation of anti-war voices in Azerbaijan during the 2020 Karabakh war. While the vast majority of the Azerbaijani population supported the war, a small minority contested its legitimacy and the related emotional obligation to express hatred against Armenians. However, these anti-war voices became stigmatised as ‘traitors to the homeland’, and were ultimately pushed to self-silence. We contribute to the growing IR scholarship on emotions and stigma in two ways. First, we show how successful stigmatisation of norm violators may involve emotional dynamics that go beyond shame. Second, we discuss the power of emotion norms of hatred, which, especially in times of war, can push ‘ordinary people’ to pro-actively and vehemently stigmatise norm-violators. In conclusion, we elaborate on the potential future implications of stigma on peacebuilding activities between Armenia and Azerbaijan.

Video Abstract

Introduction

On the 27^th of September 2020 Azerbaijan launched a military offensive to retake control of Karabakh, de jure Azerbaijani land that had been de facto controlled by Armenian separatist forces since the First Karabakh war of 1988–1994. This sparked the beginning of the Second Karabakh War. It became evident that the vast majority of the Azerbaijani population passionately supported the war.¹ The loss of Karabakh in the early 90’s, with the consequent human suffering, was a traumatic event for Azerbaijani society as a whole.² Capitalising on this traumatic and humiliating loss, the government(s) in Azerbaijan over the years institutionalised an ethno-nationalist view of the conflict³ based on hatred against the Armenian enemy.⁴ Hence, societal support for the 2020 war became tightly tied to expressions of nationalist pride and hatred for Armenians.⁵ To be a ‘proper’ Azerbaijani citizen implied supporting the country’s war effort.

Nonetheless, alternative anti-war narratives, that contested the war and rejected hatred against Armenians, were raised by a minority of individuals and political activists. These anti-war voices faced daunting societal backlash, becoming stigmatised as ‘traitors to the homeland’, ‘Armenian-lovers’ or just ‘Armenians’.⁶ All these labels denied the Azerbaijani identity of the anti-war voices, languishing the legitimacy of their claims and objectifying them as outcasts in society. While state institutions—including the State Security Service—did occasionally interrogate and exercise psychological pressure against anti-war voices, it was mostly society en masse that stigmatised the activists.

By the end of the Second Karabakh conflict on the 10^th of November 2020, which resulted in an Azerbaijani victory and a Russian-brokered ceasefire agreement, anti-war voices had already been successfully marginalised: they self-silenced or cancelled themselves from social media. The stigma, being libelled ‘traitors to the homeland’, proved effective also in the long-run. Indeed, most activists stopped expressing their anti-war views in the years that followed, including when the conflict between Azerbaijan and Karabakh Armenians was revamped in 2023.

The case of the stigmatisation of anti-war voices in Azerbaijan during the 2020 Karabakh War is interesting for Area Studies scholarship on the South Caucasus, but also for the International Relations (IR) scholarship on emotions, stigma and norms. These three concepts relate to each other in at least two fundamental manners. First, norms can refer not only to international legal provisions, such as the norm of territorial integrity, but also to emotion norms that regulate the emotions that are appropriate to feel and display in specific situations.⁷ People who adhere to the same emotion norms form emotional communities ⁸ and can stigmatise, discipline, or shame both in- and out-group norm violators. Empirically, stigmatisation and shaming have been analysed as compliance mechanisms against states that violate international norms,⁹ states that potentially adopt weapons of mass destruction¹⁰, as well as country representatives and individuals that transgress in-group emotion norms.¹¹ Second, emotions are relevant to the stigmatised, as they are pushed to feel shame and need to engage in stigma management, for example, by accepting the stigma, rejecting it, or counter stigmatising.¹²

Yet, in IR scholarship on emotions, stigma, and norms, the emotional dynamics that play out during stigma management are still to be comprehensively understood. Scholars have called to ‘explore more the conceptual linkages and connections’ between emotions and stigma.¹³ While there is growing research on how stigmatised groups or states manage stigma,¹⁴ the emotional reactions that influence stigma management remain understudied.¹⁵ In particular, the emotional dynamics that make stigma successful are ‘under-researched’.¹⁶

This gap in the literature on the emotional dynamics underlying the effectiveness of stigma is partly due to methodological reasons. IR scholars tend to analyse emotional expressions by studying existing statements shared by relevant political-diplomatic actors. For example, Adrian Rogstad (2022a) analysed Russia’s stigma management, linked to its annexation of Crimea in 2014, through a study of statements shared by Russian president Vladimir Putin and the Russian Ministry of Foreign Affairs.¹⁷ Although this methodology is valid, we underscore that relying on existing official statements by relevant political actors, instead of conducting ex-novo interviews with them, can lead to the under-detection of emotions which are carefully concealed in official statements.

However, more importantly, part of the gap in the literature also emanates from the assumption that stigmatisation is effective only insofar as it generates shame in the stigmatised. Shame is by no means the only emotion considered by IR scholars to capture the effectiveness of stigma.¹⁸ However, there is always a presumption that shame is triggered for stigmatisation to be effective and to push norm violators back to compliance. Just as an example, Claudia Junghyun Kim (2024a) mentions how stigmatisation causes ‘shame and humiliation’, so that counter-stigmatizing entails states not treating stigma as a ‘shameful marker of deviance’.¹⁹

We argue that this presumption of successful stigma requiring the stigmatised to feel shame can constrain our understanding of the wider range of emotional dynamics occurring during stigma management. In particular, we may overlook which emotions, beyond shame, are most relevant for successful stigmatisation during internationally disruptive events such as wars and violations of international norms. We argue that understanding these emotional dynamics in light of the effectiveness of stigma is fundamentally important for IR scholarship, as respect for and enforcement of internationally relevant norms depend on these emotional dynamics. Groups or states violating these (emotion) norms will fall back into compliance only insofar as shaming and stigma have emotional tolls on them.

By analysing the societal stigmatisation of anti-war voices in Azerbaijan during the 2020 Karabakh war, we contribute to the IR debate on the connection between stigma, emotions, and norms in two ways. First, through in-depth interviews with 43 Azerbaijani anti-war voices, we pay greater attention to the emotional dynamics of the stigmatised, which ultimately informed their stigma management. In particular, we investigated which emotions were more conducive to self-silencing and marginalisation. Beyond shame, we explored whether often-overlooked emotions, such as anxiety, sadness, and hopelessness, are relevant in explaining the successfulness of stigma. Second, we discuss the emotional obligation of hatred as a specific emotion norm, which, during conflicts, warrants the stigmatisation of voices opposed to hatred and war. In particular, we discuss how hatred norms can push ordinary people to stigmatise norm violators vehemently. The case of Azerbaijan is exemplificative of this, as the national emotional obligation of hatred against Armenians was enforced during the 2020 Karabakh war through the stigmatisation of anti-war voices as ‘traitors to the homeland’.

The article’s structure is as follows. We first discuss the scholarship on emotions, stigma, and norms, and their relevant interconnections. We then provide the context of Armenian-Azerbaijani conflict and enmity, which also exemplifies the existence of an emotional obligation of hatred. This is continued by introducing the research design and methods used for the empirical analysis based on interviews with 43 anti-war Azerbaijanis. Afterwards, we discuss the stigmatisation process against anti-war voices during the 2020 Karabakh War, as well as the emotional dynamics, beyond shame, at the heart of their stigma management. Finally, we discuss how the successfulness of stigma in Azerbaijan may have implications for future peacebuilding activities between Armenia and Azerbaijan.

Theoretical background

Emotions, norms & hatred

The state-of-the-art literature on IR presupposes that emotions are socially constituted, constructed, and shared.²⁰ This implies that individuals ignore, negotiate, and validate emotional meaning in a social context.²¹ Emotions also serve a normative role. In particular, emotion norms, the shared feeling rules that underpin collective group identities, determine which emotions should be expressed by in-group members in a certain situation and which emotions are to be valued.²² Therefore, IR norms are not solely international legal provisions, such as the norm of territorial integrity, but also rules on how to feel in particular situations. To use the distinction made by Gürkan (2021), norms can be ‘emotional’, in the sense that their violation triggers emotions, while ‘emotion norms’ regulate the appropriateness and value of emotions in a social in-group.²³

This connection between emotion(al) norms and in-group membership shows how emotions and identity play a mutually constitutive role as they define who we are and what we feel.²⁴ We care emotionally about things because they are connected not only to our personal values, concerns, narratives, and stories, but also to our identity. Feeling certain emotions can become a shared experience among people belonging to collective identities. In this sense, groups can become emotional communities, a type of community with shared emotions, values, and identities.²⁵ Adherence to emotion(al) norms determines who is ‘us’ and who is ‘them’, who is ‘in’ and who is ‘out’ of the community. Individuals who care about an emotional community’s identity are expected to respect the community’s (emotion) norms. In particular, Karl Gustafsson and Todd Hall (2021) discuss how community members may be required to follow emotional obligations, a specific emotion norm concerning obligations to (not) feel certain emotions in a given situation.²⁶ For example, Naomi Head (2020) analyses how Israeli public and state institutions normatively ‘outlawed’ the emotion of empathy towards Palestinians during Operation Protective Edge in 2014.²⁷ Mustafa Gokcan Kosen and Melike Akkaraca Kose (2025) discuss how pro-government media in Turkey constructed an emotion norm of hostility towards Greece during tense relations between the two countries in 2022.²⁸

In the context of our research, we focused on the emotional obligation of hatred. Specifically, the emotional obligation of (intergroup) hatred against Armenians, which became particularly heightened during the 2020 Karabakh War. Hatred is an interesting emotion to study during times of conflict, as it merges with inter-group dynamics that lead to essentialised and Manichean attitudes. Intergroup hatred is an extreme emotion directed at a particular group which ‘fundamentally and all-inclusively’ vilifies the group and its members.²⁹ In the Manichean fashion, the hating group appraises the hated group as all-inclusively and inherently diabolical. Malevolent nature and malicious intent are considered fixed and innate characteristics of the group, which will never change: the enemy group does not just behave badly; it is inherently bad.³⁰ In terms of action tendencies, the hating group can desire to harm, humiliate or even kill the hated group for the sake of causing harm. Therefore, harm is not necessarily done instrumentally; it is harm for harm’s sake.³¹ Given that the group is appraised as inherently diabolic, hatred can even be pleasurable, as it is morally justified to harm, if not eliminate, an inherently evil enemy.³² By postulating that the enemy group is diabolic and taking action against it, the hating group adopts a narrative of righteousness which increases its self-perceived positive value.³³

Community members can reinforce the emotional obligation of hatred through disciplinary mechanisms including shaming and stigma. This is particularly true during wartime, when the national public is expected to unite and comply with the community’s emotion norms. Given the extreme nature of intergroup hatred, violation of this emotion norm by in-group members may lead to the vehement stigmatisation of the violators. The emotional dynamics of the stigmatised—whether they feel shame, fear, pride, etc.—influence how they react to this disciplining effort. To further discuss this process, we turn to a theoretical discussion of stigma, with a particular focus on stigma management and shame.

Stigma & shame

In state-of-art literature, the term stigma was initially conceptualised by Erving Goffman as a deeply discrediting attribute imposed on those in-group members who contest ‘natural’, ‘normal’ or ‘commonsensical behaviours’.³⁴ In this sense, stigma is presumed to be a normative social act since it is undertaken against the transgressors of the social norm.³⁵ When this social norm is violated, stigma will usually include elements like labelling and stereotyping the transgressors, which results in social separation between the ‘normals’ and the transgressors, status loss and discrimination.³⁶ This can be exemplified when transgressive individuals and groups are seen as traitors to the community during times of war and therefore deserve censure, punishment, and even banishment.³⁷

The stigmatised group must then engage in stigma management and decide how to react to it. In the original elaboration of stigma management, that of Adler-Nissen (2014), the stigmatised may accept the stigma, and thus correct and/or apologise for one’s transgressive conduct; reject the stigma, and thus accept the categories of deviance, but deny being different from the norm abiders; or counter-stigmatise, and thus reverse stigma as a source of pride by affirmatively self-identifying with the stigmatised group.³⁸ Additionally, according to Adrian Rogstad (2022a), the stigmatised group may also engage in stigma evasion, a situation where the stigmatised group accepts that they have done something socially discreditable but tries to evade stigma by contesting its public understanding or minimising the seriousness of the socially discreditable act.³⁹ Aniruddha Saha (2022) added that norm violators may engage in stigma redaction, by which they only occasionally engage in corrective conduct to prevent their identities from being cemented as deviants.⁴⁰ Finally, Claudia Junghyun Kim (2024a) discusses the possibility of stigma shifting, by which the stigmatised engage in distinction-seeking behaviour vis-à-vis hierarchically ‘lesser’ groups, thus also allowing themselves to take on the role of stigmatisers.⁴¹

Emotions are essential for stigma management. According to mental health practitioners and sociologists Bruce Link, Lawrence Yang, Jo Phelan, and Pamela Collins (2004), stigmatised groups usually experience negative emotions such as embarrassment, shame, fear, alienation, or anger.⁴² Shame, in particular, is sometimes argued to be a central emotion that distinguishes feelings of stigma from other similar concepts, such as marginalisation or discrimination.⁴³

Indeed, IR scholarship has mostly focused on the emotion of shame to explain stigmatising reactions. It is presumed that stigmatisers try to evoke shame in the stigmatised, which ultimately leads to acceptance of the stigma and corrective behaviour. For example, Rebecca Adler Nissan (2014) writes how stigmatisation helps clarify the ‘boundaries of acceptable behaviour and identity and the consequences of nonconformity, that is, shame, exclusion, or other forms of punishment’.⁴⁴ Simon Koschut (2022) mentions how in IR scholarship there tends to be a ‘general understanding’ that ‘logically links shaming to stigmatisation’.⁴⁵ Finally, Claudia Junghyun Kim (2024b) discusses Japan’s reaction to international shaming and stigma, triggered by the country’s violation of norms regarding its colonial and wartime past. In particular, she argues that ‘some aspects of shame’ have led to Japan’s willingness to apologise, while other aspects of shame, perceived as threatening to its morality, have led to Japan’s defensive behaviour.⁴⁶ Overall, in IR scholarship, the successfulness of stigma seems to be linked to whether shame is indeed solicited in the stigmatised. There may be different nuances of shame in stigma management based on its effects on identity and status; however, shame is nonetheless considered central to the success of stigma.

It is not difficult to understand why shame has received relatively more attention than other emotions in scholarships on stigma and shaming. Shame indicates an ‘internalization of the society’s judgement’ for violating a certain norm.⁴⁷ It is caused by public exposure to wrongdoing⁴⁸ and therefore implies a fundamental failure of the self to live up to values and norms that are shared with others.⁴⁹ Shame is typically associated with being hierarchically lower ranking, and thus with feelings of powerlessness.⁵⁰ Intuitively, stigmatisation is an exemplary case of such ‘societal judgement’ against norm violators. When successful, it can lead the stigmatised, particularly those of lower ranks, to feelings of shame and powerlessness. Ultimately, this brings about marginalisation and compliance.

We now briefly summarise the connections between emotions, stigma, and norms during conflict. The emotional obligation of hatred can be particularly relevant during times of war, when society is expected to unite and emotionally support the state’s war effort. The national public, in the form of an emotional community, can be required to express feelings of hatred against the enemy and support wars aimed at destroying it. Community members who violate this norm of hatred may trigger extreme reactions such as shaming, threats, and stigmatisation. The ultimate aim of societal stigmatisation is to push back into conformity, silence, or marginalise norm violators. In IR scholarship, it is generally presumed that successful stigmatisation involves feeling some form of shame. However, we argue that this presumption may overlook other emotions relevant to the stigma management process, such as anxiety, sadness, and hopelessness. The stigmatisation of anti-war voices in Azerbaijan during the 2020 Karabakh War serves to exemplify such overlooked emotional dynamics.

Context of the Azerbaijani-Armenian conflict over Karabakh

To understand this case, we first need to briefly discuss the modern origins of the Karabakh conflict at a time when Armenia and Azerbaijan were still Soviet Republics within the Soviet Union and Nagorno-Karabakh (NK) was an Armenian-majority Autonomous Oblast (AO) within Azerbaijan. The conflict unfolded in the late 80’s, with rising nationalist demands in Armenia and the Armenian population of NK itself, who demanded Moscow re-assign the AO to the Armenian SSR. These irredentist dynamics gradually led to localised inter-communal violence, and ultimately to a full-scale war between the two newly independent republics of Armenia and Azerbaijan. The result of the first Karabakh war (1988–1994) was the traumatic loss of a huge swathe of territory for Azerbaijanis, which extended not only to former NK but also to seven adjacent Azerbaijani territories. The new status quo saw Armenia de facto controlling approximately 16% of Azerbaijan’s de jure territory. The human toll is difficult to estimate, but on the Azerbaijani side, approximately 16,000 dead and 750,000–850,000 displaced people.⁵¹

Capitalising on this traumatic and humiliating defeat, the Azerbaijani governments institutionalised an idea of national identity along the lines of victimhood, suffering at the hands of aggressor Armenians, loss of territory, and a deep-rooted yearning to restore territorial integrity.⁵² Armenians become depicted as the national ‘enemy’,⁵³ the demonic ‘other’ of a ethno-nationalist interpretation of the conflict.⁵⁴ These national narratives were incorporated into and reproduced by the state media, school curricula, cultural productions, national commemorations, and official government discourses.⁵⁵ For example, Azerbaijani President Ilham Aliyev consistently utilised hateful and dehumanising rhetoric against Armenians. In his speeches Armenians are depicted as provocative liars, as well as threatening ‘savages’, ‘jackals’ and ‘dogs to be chased’.⁵⁶ History school textbooks convey to students images of Armenians as ‘eternal enemies’ capable of ‘unreasonable bloodthirstiness’,⁵⁷ thus assuring the perpetuation of inter-ethnic hatred in the newer generations.⁵⁸ As an example, a grade-5 history school textbooks, when discussing the first Karabakh war and massacres committed by Armenians, describes ‘Armenianness’ as standing ‘above all bestiality and cruelty’.⁵⁹

These narratives were successfully spread and enforced among the population, partly thanks to the country’s illiberal and authoritarian politics,⁶⁰ which systematically repressed alternative narratives about Armenian-Azerbaijani good neighbourly relations.⁶¹ The incapacity, if not the reluctance, of the Armenian and Azerbaijani political establishments to solve territorial disputes peacefully and diplomatically worsened these enemy images domestically.⁶² Ultimately, an emotional obligation of hatred against Armenians was developed and institutionalised by the state.⁶³ The Azerbaijani nation, qua emotional community, became bound by this emotion norm, thus tying the Azerbaijani identity to hatred against Armenians. Community members violating this norm deserved repression, stigmatisation, and marginalisation.

These emotional dynamics became particularly heightened in September 2020, when Azerbaijan launched a military offensive to retake possession of Karabakh. Azerbaijani society passionately supported the war effort, and began to fiercely stigmatise as ‘traitors to the homeland’ those few anti-war voices that contested the obligation of hatred against Armenians and the legitimacy of war.

Data & methodology

Methodologically, we drew upon 43 semi-structured, in-depth interviews with Azerbaijani anti-war voices. The interviewees were recruited through snowball sampling. Given that the most vocal of activists expressed their anti-war positions on social media such as Twitter and Facebook, the first interviewees were contacted through social media. To avoid selecting only the most vocal anti-war voices, the first interviewees were asked to confidentially suggest to us other individuals with anti-war positions within their social circles. Through this snowball sampling technique, we also interviewed people with anti-war positions who did not express their views online and would therefore have gone undetected through recruitment based solely on social media posts. Thus, our sample reflects both openly vocal and concealed anti-war positions.

The interviews were conducted between November 2021 and March 2024; however, most interviews were conducted in 2022. While almost all interviews were conducted in private settings, two were conducted in group settings at the discretion of the participants. In terms of language, 13 interviews were conducted in Azerbaijani and 30 in English. Moreover, 5 interviewees were interviewed twice to enhance our comprehension of certain emotional facets of their experiences. The average duration of the interviews was approximately 45 min. In terms of location, 12 interviews were conducted online and 31 were conducted in person in Azerbaijan. Almost all the in-person interviews in Azerbaijan were conducted in alleged ‘safe spaces’ for activists and civil society. Indeed, many interviewees belonged to Azerbaijani civil society and were thus politically active. Our interviewees politically and ideologically categorised themselves as classic liberals, leftists, and feminists. These three political groupings had further internal divisions, but stood out as meaningful political/ideological self-categorisations. Others expressed no particular political or ideological affiliation. Overall, we balanced our sample to have at least 10 anti-war interviewees from each of these political/ideological groupings, as well as 10 politically non-affiliated anti-war voices.

In terms of gender, the sample is composed of 14 females, 27 males and 2 non-binary persons. The age of the interviewees varied from 18 to 60 years, with the majority of interviewees in the 20–35 age group. In terms of education, many interviewees qualified as educated, holding at least a bachelor’s degree. However, this varied with age and not all interviewees had higher education.

The anonymised interviewee table is displayed in the Appendix. However, the latter was purposefully kept essential to protect the identity of our interviewees. This is important in authoritarian settings, such as Azerbaijan, where state repression always looms in the background.

The first author designed and conducted the semi-structured interviews. The second author contributed with knowledge of Azerbaijani’s contemporary history, as well as contextual and analytical insights from his lived experience and insiders’ perspectives in the region. Owing to confidentiality, the second author had access only to anonymised and selected extracts from the interviews.

Given that many of our interviewees estimated the total number of anti-war voices to be 300, our sample would have credible grounds to claim representativeness of Azerbaijan’s anti-war positions. This figure of 300 is a rough estimate of the interviewees. However, this cannot be confirmed with certainty. As a caveat, this estimated figure of 300 includes only radical anti-war positions—those who opposed the war based on humanitarian and political/ideological reasons. For example, an Armenophobe mother who is against the war because she fears her son(s) being drafted, would not be included in this estimated 300 figure.

Potential caveats to the research may be the interviewees’ social desirability bias on the one hand, and a male heteronormative proclivity in hiding certain emotions—like shame and fear—on the other hand. Finally, interviews were conducted one to three years after the 2020 Karabakh War, so emotions had to be recollected. The language in which the interviews were conducted may have also had an impact; however, overall, we noticed no substantial differences between the interviews conducted in Azerbaijani and those conducted in English.

Empirics

War, counter-narratives, and societal stigmatisation

During the 2020 Karabakh War, a small minority of individuals articulated alternative anti-war and reconciliatory narratives, especially on social media. The earliest dissenting voices arose from a group of 17 Azerbaijani leftists condemning the war, with an anti-war manifesto signed and circulated at its onset. This was followed by another anti-war stance in early October drafted and signed by a politically diverse group of Armenian and Azerbaijani civil society individuals. However, anti-war activism (mostly online due to Covid restrictions) was by no means limited to these two statements. Taken together, these loosely aligned and heterogeneous dissenting voices were tagged as the ‘No-War Movement’.⁶⁴ Even if the anti-war voices were ideologically very heterogeneous, encompassing classic liberal, leftist, and feminist stances, they all advanced counternarratives based on two common points: first, the rejection of the emotional obligation of hatred against Armenians; and second, the rejection of violence and war as an action tendency to solve the Karabakh territorial dispute.⁶⁵ To illustrate this, we can read a relevant extract of the leftist No-War manifesto, which captures the two common points of the whole ideologically heterogeneous No-War movement.

“Our enemy though is not a random Armenian, whom we have never met in our lives and possibly never will… We strongly condemn every move taken to prolong the conflict and deepen hatred between the two peoples. We want to look back and take the steps necessary to rebuild the trust between our societies and the youth. We reject every nationalist and state-of-war narratives that exclude any possibility of us living together again on this soil. We call for peacebuilding and solidarity initiatives. We believe that there is an alternative way out of this stalemate through mutual respect, peaceful attitude, and cooperation”.⁶⁶

Logo of the leftist anti-war manifesto, which became the de facto logo of the whole (heterogeneous) No-War movement

As a consequence of breaking the community’s norms of hatred, anti-war voices faced societal stigma, insults, and threats. This came not so much from state institutions but from society en masse. According to our interviewees, it was ‘ordinary people’ who by far engaged most in stigmatisation. The heaviest stigma labels were said to be traitors to the homeland (vətən xaini), lovers of Armenians (ermənipərəst) or simply Armenians (erməni). These labels either denied their identity qua Azerbaijanis (being ‘Armenians’), or posed them as internal enemies of the nation (‘traitors to the homeland’). They were pushed into feeling shame for violating the community’s emotion norms, thereby becoming national outcasts.

Stigmatisation was accompanied by insults and threats, including death threats. According to our interviewees, women and LGBTQ+ individuals bear the brunt of this situation. This is likely connected to the harsh prevalence of patriarchal and homophobic norms in Azerbaijani society.⁶⁷ For example, one particularly heavy gendered threat received on social media read: ‘Hey bastard, I heard that your name is among those who signed the peace statement. I fuck the honour of you bastards made of Armenian material. May the [war] martyrs fuck your dead relatives in the afterlife’. ⁶⁸

Ordinary people also launched a campaign of naming and shaming on social media, including apps such as Telegram and WhatsApp, by sharing the names of those who signed the September and October anti-war statements. This campaign also included photos of signatories to make them more publicly recognisable. The circulating screenshots and photos were normally accompanied by stigma—like ‘traitors to the homeland’. Some of our interviewees also recounted how even friends and acquaintances, after seeing the screenshots, asked them if their names were indeed on the list. For example, in the words of one interviewee: ‘Some people texted me on Messenger, and cursed my mother and they called me a “traitor to the homeland”. A friend of mine even shared on Facebook one of those screenshots and tagged me: “Hey Malik [pseudonym], is this you?” I replied to him “Yeah, it’s me”. He didn’t curse me … but told me I should feel ashamed’.⁶⁹

Another example is a message, shared with us by our interviewees, that listed the names of signatories of the anti-war statements and called for their death: ‘[Listed names of signatories, omitted], the dishonourable people in this list say that our fight is with our government, not with the Armenians. Such signatories during war time, such dishonourable people, must be killed through shooting as punishment’. In the latter example, we have a case of naming and shaming accompanied by stigma (‘dishonourables’), which in war times warrants extreme punishment.

Moreover, the signatories of the October anti-war statement were also targeted because the list of signatories contained a certain ‘Anna Hakobyan’, which was (intentionally) associated with the homonymous wife of Armenian prime minister Nikol Pashinyan. Thus, they were accused of siding with the Armenian government and against Azerbaijan, an act of unforgivable treason. Although the September No-War manifesto is still available online, the October anti-war statement was retrieved after an excessive amount of societal backlash and stigma.

Interestingly, according to our interviewees, repressive state institutions played a minor role. However, a few interviewees faced state pressure, particularly from secret security services (DTX). For example, one interviewee was called DTX and was given an appointment at their headquarters. He recalled how, once he entered the room, the DTX agents printed pictures of his anti-war statements on Facebook. The agents told him to be ashamed of his behaviour, that he was an ‘Armenian traitor’, and that it was disrespectful towards the martyrs who died at war to write such anti-war statements. The interviewee was also asked to cancel these Facebook posts; otherwise, he would face unspecified consequences.⁷⁰

However, aside from a few threats from secret security services, almost none of the interviewees received pressure or repression directly from the state. It was ordinary people who stigmatised the articulators of these alternative narratives. By calling them ‘traitors to the homeland’, ‘Armenians’, ‘dishonourables’ etc., they aimed at shaming and delegitimising both their alternative narratives and them personally as bearers of those alternative narratives. Society itself became a proactive enforcer of the violated emotional obligation of hatred.

Emotional dynamics of the stigmatised

Beyond the stigmatisation process per se, we can also discuss the emotional dynamics of the stigmatised. Which emotions play the biggest role in ultimately pushing alternative voices toward self-silence and self-censorship? For example, is shame, as the IR theory on stigma suggests, the most relevant emotion, or should other emotions, such as anxiety, sadness and hopelessness, also be considered? In our case, confronted with societal stigma, our interviewees underwent two divergent experiences of stigma management: a minority openly rejected the stigma, while most were emotionally overwhelmed and downheartedly accepted it.

Starting with those who rejected stigma, they expressed a sense of determination in their devotion to principled alternative narratives. Some even expressed a sense of pleasure and pride in challenging the nationalist hateful norms and the ‘mass’ of stigmatisers. Fear was welcomed as a challenge. This is best exemplified in the words of one interviewee: ‘I take pride and pleasure in changing people, both because they change and because it is me who changes them. I do not care about peer pressure, because I think they are idiots … I do not fear social exclusion because I am already a social outcast. I love social media because it allows a loser [myself] to become famous’. ⁷¹ In this sense, the interviewee already felt like a social outcast, so perhaps the exclusionary power of the stigma was less emotionally burdensome. Likewise, two other interviewees claimed that they felt no real psychological pressure or anxiety due to societal backlash. They dismissed the stigmatisers as idiots, uneducated people, ‘zombies’ and ‘trolls’. Thus, it was not worthwhile feeling shame or anxiety because of such valueless people. Indeed, some interviewees openly mentioned how they took pleasure arguing with other people, as this allowed them to win the debates. One interviewee also mixed his rejection of stigma with gendered language. In his own words ‘I actually enjoy fighting with them online (the stigmatisers). I feminise their hysteria’. ⁷²

However, most interviewees accepted the stigma and expressed various forms of emotional discomfort akin to shame. Many identified the expression ‘traitor to the homeland’ (vətən xaini) as the most psychologically heavy label. The issue was not just being libelled a traitor, but the fact that the entire Azerbaijani nation was mobilised against them and called them traitors. They came to perceive themselves as a marginalised micro-minority against society as a whole. Some outright expressed feeling ‘excluded by the nation’. A couple of interviewees mentioned that even their parents called them ‘traitors to the homeland’, which was very psychologically straining for them. The stigma was so overwhelming that a few interviewees admitted to questioning their values and principled positions. For example, in the words of one interviewee: ‘If you are alone, you may start questioning your own ideas… Sometimes I did feel like a criminal or indeed a traitor’.⁷³ This cautioned many interviewees to either refrain from engaging online in the Azerbaijani language by posting only in English, or to simply deactivate their social media accounts to disappear from public attention. Exemplary of this are the words of one interviewee: ‘I did not express my opinion openly to the Azerbaijani audience, but on Twitter I was openly making statements in English against the war. I did not write in Azerbaijani because there was too much hatred, but I support and admire the courage and patience of the people writing in Azerbaijani. It was important to create an alternative discourse. It was not so much about “fear” for me, I simply don’t like to be marginalised’.⁷⁴ Interestingly, in both the Azerbaijani and English interviews, interviewees tended to avoid using the word shame and preferred describing their emotional state with similar terms, such as embarrassment, discomfort, or a feeling of marginalisation.⁷⁵ We do not want to dismiss the relevance of shame in stigma management but suggest instead that in practice, interviewees may describe emotions akin to shame with a variety of expressions. Indeed, interviewees expressing embarrassment, discomfort, or marginalisation also expressed a feeling of powerlessness, which, according to theory, is normally associated with shame.⁷⁶

The emotion of anxiety was also expressed by many of our interviewees and played an important role in stigma management. Women and LGBTQ+ interviewees tended to admit more openly and discuss the intensity of psychological strain caused by societal stigma. This contrast with male interviewees arguably stems from the absence of heteronormative proclivity for woman to conceal emotions in Azerbaijani culture. Notwithstanding these differences between male and female/LBGTQ + interviewees, most interviewees admitted some form of anxiety caused by societal stigmatisation. Anxiety was said to be ‘paralysing’, as it hindered their capacity to concentrate and to react to the backlash. This also led some interviewees to feel ‘lonely’ since they were left alone in the fight against society and its emotion norms. For instance, in the words of one female interviewee: ‘I started getting backlash, and this created a lot of anxiety in me. I started getting mocked and stigmatised. I was super anxious, and couldn’t concentrate on anything. I was always thinking about the mocking and stigma… I felt lonely, sad. I wasn’t happy, I thought there was no purpose in life’.⁷⁷ Similarly, another interviewee expressed how societal stigma created great stress and anxiety in him, which was worsened by the feeling of being powerless: ‘I cried. I was so anxious, it was overwhelming. Online I was called a “traitor to the homeland”. I had bad nightmares, I was shaking. But I could do nothing’. ⁷⁸ One of the few ‘cures’ to this anxiety, and related feeling of loneliness, became mutual psychological support among the stigmatised anti-war voices. Indeed, many interviewees reported that private social media group chats played a very important role in alleviating their sense of anxiety, as they felt that they were not completely alone in this fight after all.

Most interviewees expressed some form of fear. In particular, many mentioned fear of consequences for family members. For example, if they had relatives working in some state-connected institution or firm, they feared that because of them their relatives would lose their job. Therefore, they worried that the state might punish them indirectly by having their relatives face consequences. This, in turn, puts pressure on them from family members. In addition, women and LGBTQ+ interviewees tended to fear and take threats more seriously than heterosexual men. In particular, the threats of being physically attacked on the streets or raped. These gendered threats were psychologically heavy and induced fear. Exemplary are the words of one female interviewee: ‘The worst part of all of this was the gender related insults and threats. This has a tense psychological effect on me, it created fear, I was afraid that if I went to the streets I would be attacked. I was lynched on Facebook’. ⁷⁹ Overall, also the emotion of fear played a role in limiting or silencing the articulation of alternative anti-war, conciliatory narratives. While not as paralysing as anxiety, fear pushed some interviewees to self-silence and self-censor, or at least to be cautious about what they expressed online.

However, the emotions described as by far the most paralysing, often discussed in combination, were hopelessness and sadness. The backlash and stigma pushed many interviewees to self-silence because the sheer number of attacks made them lose faith in the people and the possibility of positive change. Thus, they plunged into a state of depression and despair. A few interviewees even mentioned having gone to therapy after the war because of depression and hopelessness. A couple of interviewees also mentioned ‘apathy’ towards Azerbaijan and the people as a consequence of the backlash. They saw no point in ‘fighting’ anymore. In many cases, the sense of despair persisted after the war. Indeed, sadness and hopelessness were associated with long-term self-silencing and self-censorship. If shame, anxiety, and fear were surely relevant to explain the paralysing effect of stigma in the short term—namely, during the 2020 Karabakh War—sadness and hopelessness played the most relevant role both in the short and long term. These emotional states can well be captured through the words of one interviewee: ‘I am tired of being harassed, stigmatised, insulted, and losing friends at such a young age… They share personal information of yours, like photos, videos and messages… Nobody gives a fuck about us. We are left to die. If my family was not here, I would leave immediately and go abroad. I am just tired of fighting. We only live once, and I don’t want to become a sacrificial animal’.⁸⁰ Although not all interviewees used these extreme words, most of them nonetheless expressed some form of hopelessness or sadness. For example, one interviewee explicitly tied these emotional states to his decrease in political activism: ‘I am less “activist” than I used to be. Because I feel frustrated and disappointed. Because I have been an activist for 5 years, but nothing changed. Things actually got worse. So at a certain point I started asking myself “what’s the point of this?” If everything goes worse anyway, what’s the meaning of this?’.⁸¹ Indeed, many of the interviewees expressed a general sense of ‘defeat’. They had no more energy to fight, because they simply ‘had lost’ against the authoritarian state and the nationalist norms of society. They had tried to propose alternative reconciliatory narratives, to oppose the norm of hatred, but ‘ethno-nationalism’ had triumphed. There was no hope. By now, the people had been completely ‘zombified’, and the authoritarian government would have become even more authoritarian. We can summarise this widely shared sense of defeat, sadness and hopelessness through the words of one interviewee: ‘So that’s why I’m even more pessimistic now: we lost to nationalism. Feminists, liberals, socialists, we all lost to nationalism’.⁸²

Conclusion

In conclusion, our analysis on the successful marginalisation of anti-war narratives during the 2020 Karabakh war reveals complex emotional stigma dynamics that go beyond just the emotion of shame. Society en masse stigmatised as ‘traitors to the homeland’ the anti-war voices who violated the national community’s emotional obligation of hatred against Armenians. In terms of stigma management, most of our anti-war interviewees accepted the stigma, ultimately leading to self-silence and self-censorship. Interestingly, only a handful of interviewees mentioned feeling pride for their principled position, leading them to reject the stigma and call their offenders ‘idiots’ or ‘trolls’. However, this successful stigmatisation cannot be explained solely by the presence of shame or a lack of pride. While most IR scholarship on stigma focuses on shame to understand the effectiveness (or lack thereof) of stigma,⁸³ our case reveals that more complex emotional dynamics may occur. Indeed, our interviewees expressed a greater variety of emotions, along with being stigmatised: shame (or akin to it), anxiety, fear, hopelessness and sadness. The emotions of fear and shame undeniably played an important role in pushing some interviewees toward self-silence and self-censorship. We also noticed gender differences in the emotional dynamics of stigmatised individuals. In particular, fear seems to have been especially psychologically heavy for women and LGBTQ+ interviewees. This seems to be related to the combination of stigma and gendered threats, from which men were spared.

However, the most daunting and paralysing emotions have been said to be anxiety, hopelessness and sadness. In particular, hopelessness and sadness seem to have been the emotions that were most impairing in the long-term. They were also associated with self-censorship in the years following the 2020 war. The quasi-totality of interviewees mentioned, in some form or the other, having lost hope, the desire to fight, and purpose. Some described this feeling as ‘apathy’ or ‘defeat’. Although not everyone expressed emotions akin to shame, almost everyone expressed hopelessness and sadness. Interestingly, this was also true for interviewees who (allegedly) rejected the stigma.

Thus, the successfulness of stigmatisation should not be solely attributed to the (non) triggering of shame in stigmatised individuals. We argue, instead, that it is important for the growing IR scholarship on stigma and emotions also to consider the possibility of a more complex variety of emotional dynamics behind stigma management. In line with the call to explore the ‘under-researched’ topic of what makes stigma successful,⁸⁴ we believe that paying attention to the complex emotional dynamics of the stigmatised can help scholars better understand when stigma imposition is effective. In particular, the paralysing long-term effects of hopelessness and sadness. We do not downplay the centrality of shame in stigma management. However, the successful marginalisation of anti-war voices in Azerbaijan shows that, at least in certain instances, the emotional dynamics of stigma go beyond shame. Indeed, contrary to what is often theoretically assumed, not all our interviewees expressed emotions akin to shame, but almost all mentioned sadness and hopelessness as a result of stigma. Exploring these complex emotional dynamics was possible only through semi-structured interviews with the stigmatised, which sometimes required a second interview. Therefore, we invite researchers to consider interviews a valid method to explore stigma imposition and management during internationally relevant events.

These results may also have implications for conflict-transformation efforts in other conflict cases, especially for track-two diplomacy between civil societies. Breaking national enmity or hatred norms, as well as proposing alternative reconciliatory narratives, may prove to be a particularly daunting task for peacebuilders facing societal stigma. Aside from feeling shame, fear, and anxiety in the short-term, the emotions of hopelessness and sadness may prove to be the most paralysing in the long run for peacebuilders who are unable to reject stigma.

Discussion

The power of hatred norms

Through the incorporation and reproduction of ethno-nationalist narratives by state media, school curricula, cultural productions, national commemorations, and official government discourse, an emotional obligation of hatred against Armenians was made hegemonic in Azerbaijan. This prompted society—ordinary people—to passionately support the 2020 war but also vehemently stigmatise anti-war voices who violated this emotion norm.

Although beyond the purpose of this research, we speculate that the reasons for such en masse and vehement stigmatisation could be linked to the emotional power of hatred norms. As discussed in the theory, when there is intergroup hatred, the enemy is typically appraised as inherently diabolic. Malicious intent is a fixed and innate characteristic; therefore, there is no possibility of change or redemption.⁸⁵ This makes it intolerable and shameful for community members to violate the hatred norm and express reconciliatory emotions, such as compassion or empathy, towards the enemy. Thus, while hatred is directed against the enemy out-group, unspeakable anger may be directed against in-group violators of the hatred norm. It can be argued that the stigmatising public not only gains a sense of righteous pleasure from hating the diabolical enemy⁸⁶ but also from stigmatising the community members who violate the norm of hatred. This feeling of righteous pleasure may push ordinary people to pro-actively stigmatise and enforce the community’s emotional norms.

We believe that in the presence of a norm of hatred in intergroup conflicts, it could be interesting for future research to pay attention to this emotionally pleasurable aspect of stigmatisation. This may help understand other cases where ordinary people passionately engage in the stigmatisation of norm violators.

The future of anti-war voices in Azerbaijan

As for Azerbaijan, alternative conflict narratives remain marginalised and repressed.⁸⁷ After the war, also the state began adopting the ‘traitor to homeland’ stigma label to silence the remaining anti-war dissenters. This became prominent during the ‘Know the traitors!’ campaign, launched in 2022 by the ruling party and media, targeting pro-peace voices.⁸⁸ Thus, while stigmatisation was performed mainly by society itself during the 2020 Karabakh war, the regime began a stigma campaign against anti-war voices in the years that followed.⁸⁹ In particular, during the summers of 2023–2024, state propaganda targeted anti-war activists by calling them ‘traitors to the homeland’ who were ‘working for Armenians’.⁹⁰ Moreover, the government began arresting people engaged in peacebuilding activities or just in independent dialogues with the Armenians.⁹¹ This signalled an escalation from stigmatising the few remaining vocal anti-war voices to an upfront coercive repression. Ideologically, feminist groups are among the only remaining critical voices in Azerbaijan that propose anti-war and reconciliatory narratives about Azerbaijani-Armenian relations.⁹² Most other anti-war voices have either been marginalised by successful societal stigmatisation or, later, by sheer government repression.