Accepting Alien Rule? State-Building Nationalism in Georgia’s Azeri Borderland

Abstract

How did ethnic Azeris in the Marneuli, Bolnisi and Dmanisi districts, located inside Georgia but bordering Azerbaijan, react to the reorganisation of political space along national lines after the Soviet Union’s dissolution? ‘Beached’ in foreign states bent on nationalising their domains, minorities throughout Eurasia sometimes rejected and sometimes accepted their alien rulers. This essay examines reactions to this predicament among Georgia’s Azeris. Drawing on elite interviews and data from a matched-guise experiment, it concludes that locals have come to accept their host state after its state-building nationalism took an inclusive turn and the distinction between aliens and natives faded.

Even before Georgia declared its independence from the Soviet Union in 1991, ethnic Azeris in Marneuli, Bolnisi and Dmanisi had begun to feel—and resent—its rule. Residents in these districts, situated across the border from Azerbaijan, had in recent months seen their villages rechristened with Georgian-sounding place names. Jafarli turned into Samtredo, Fakhrali became Talaveri, and so on, in what locals perceived as an attempt to sever their bond to their birthplace (MRMG 2011a). Stranded inside a foreign state and under the thumb of alien rule, local Azeris had to contemplate their response.

Minorities across Eurasia faced a similar dilemma. As the USSR broke up, ‘titular’ nationalities proclaimed independence in the Soviet socialist republics bearing their names and then embarked on ‘nationalising’ policies (Brubaker 1996; 2011, p. 1786). Ethnic differences had been institutionalised under Soviet rule, since titulars benefited from affirmative action policies at the expense of others in their republic (Slezkine 1994; Beissinger 2002; Broers 2004, p. 116). After the sudden border demarcations that occurred in 1991, minorities found themselves ‘beached’ inside states under the control of titulars pursuing policies promoting their own language, culture or demographic dominance (Laitin 1998). Interethnic tensions flared between titulars, beached minorities and their kin states. All parties shared the belief that nations and states were ‘destined for each other; that either without the other is incomplete’ (Gellner 2006, p. 6), but their preferred solutions differed.

Titulars favoured turning minorities into co-nationals or expelling them from the state, a process described as ‘nationalising’, ‘nation-building’ or ‘state-building nationalism’ (Hechter 2000, p. 15; Kuzio 2001). Minorities who found integration or emigration intolerable mobilised in defence of their rights, at times with support from kin states set on redrawing borders to subsume their co-ethnics. Minorities beached in ‘borderlands’, separated from their respective kin state, became the target of competing nationalist projects. Among these groups, the most notable were the 25 million ethnic Russians, and a still larger number of Russian speakers, stranded in the ‘near abroad’. In the words of Russian President Vladimir Putin, ‘millions of people went to bed in one country and awoke in different ones, overnight becoming ethnic minorities in former Union republics’ (2014). Defiant Russophones soon carved out their own unrecognised statelet in Moldova’s Transnistria region, and ethnic Russians in Ukraine’s Crimean Peninsula later backed its outright annexation into their kin state. Yet in Estonia’s Narva region and other Russian-populated borderlands, locals submitted to life under alien rule.

Beached minorities also faced nationalising states in the South Caucasus. In Azerbaijan’s Nagorno-Karabakh region and Georgia’s South Ossetia, locals took up arms in irredentist struggles. In Georgia’s Abkhazia region, minorities mobilised for independence. Other groups, including the ethnic Azeris stranded in Georgia’s Marneuli, Bolnisi and Dmanisi districts, acquiesced to the imposition of alien rule. This essay aims to trace their reactions to this predicament and probe the conditions under which beached minorities shun conflict and come to accept alien rule. Even though alien rule is often rejected as unequivocally illegitimate, social actors at times prefer alien over native rule, not least when the former offers superior governance (Hechter 2013, pp. 1–45).¹ Since Azeris in Marneuli, Bolnisi and Dmanisi have been subjected to state-building nationalism of different shades, their reactions offer a rare window into the mechanisms that do—or do not—render alien rule acceptable in other borderlands harbouring beached minorities in post-Soviet Eurasia.

With this aim in mind, this essay tackles the question: how did ethnic Azeris, residing in these Georgian districts on the Azerbaijani border, react to the reorganisation of political space along national lines following the collapse of the Soviet Union? Through a close reading of public reports, analysis of original data from 15 semi-structured elite interviews and a matched-guise experiment administered to 483 adolescents, I retrace locals’ reactions to the imposition of alien rule over their borderland. In accordance with Hechter’s (2013) predictions, I conclude that Azeris have come to accept their Georgian host state as it—over time—turned towards a more inclusive kind of state-building nationalism. At first, locals beached in the borderland mobilised around ethnic markers to protest against Georgian discrimination. In order to placate them, officials co-opted Azeris into state structures, but the corrupt state still failed to offer public goods. After the November 2003 Rose Revolution, Georgian authorities instituted more direct rule over the borderland, but also began delivering public goods in an equitable fashion and acculturating minorities into their host state. As Georgian rule became more capable and less alien, Azeris came to accept the prospect of morphing into co-nationals.

This manuscript sheds novel light on the potential for conflict among beached minorities in post-Soviet Eurasia. Earlier research has focused on ethnic Russians in the ‘near abroad’ (Laitin 1998; Poppe & Hagendoorn 2001; Commercio 2004). Other minorities have attracted attention when rebelling against their host states (Kaufman 2001; King 2001; Cornell 2002a; Toft 2003; Kolstø & Blakkisrud 2008; George 2009). Despite representing Georgia’s most populous ethnic enclave (Geostat 2002; CSEM 2014), the Azeri-inhabited Marneuli, Bolnisi and Dmanisi districts have received little academic notice. Nor has this borderland been conceptualised as a case among the larger population of beached minorities submitting to ‘alien rule’. I address this omission and make a twofold contribution. I present original empirical data that reveal Azeri reactions to life under Georgian rule. As the investigation focuses on factors thought to affect native attitudes to alien rule (Hechter 2000, 2013), I also probe the theoretical conditions under which beached minorities reject or accept their host states.

The argument unfolds over six sections. I first make conceptual clarifications, introduce some predictions about the legitimation of alien rule, and discuss the challenge of measuring minorities’ reactions thereto. I then reconstruct Georgia’s initial attempts to impose direct rule over the borderland and its ensuing reversion to indirect rule in the period leading up to the Rose Revolution. Next, I detail Georgia’s turn towards inclusive state-building nationalism and explore the reactions this shift elicited among elites and adolescents in the borderland. In the conclusion, I pinpoint the mechanisms that led Azeris to accept their host state and stress the importance of overcoming the distinction between aliens and natives if beached minorities are to refrain from challenging the borders of the post-Soviet order.

Alien rule and its discontents

If members of one group promulgate and enforce rules inside a governance unit that also, in some corner of its domain, contains members of a different group, then the latter often see the former as ‘alien rulers’ (Hechter 2013, p. 2). This problem is commonplace in international politics. Norms of national self-determination, codified in Article One of the United Nations Charter, hold that a people should control its ‘own’ governance unit. However, this prerogative is for the most part reserved for territorial administrations rather than ethnic groups within them (Roeder 2007). Thus, when the Union of Soviet Socialist Republics and the Socialist Federal Republic of Yugoslavia fell apart in the 1990s, the right to self-determination devolved to their constituent republics and not to minorities stranded inside them—a frustrating restriction for ethnic groups across Eurasia and the Balkans.

However, the puzzle remains: under what conditions do beached minorities reject or accept the imposition of alien rule? What drove Russophones in Transnistria to rebuff the Moldovan state but ethnic Russians in Narva to submit to the Estonian state? What led Armenians in Nagorno-Karabakh to rebel against Azeri rule but Armenians in Akhalkalaki to assent to Georgian rule? No single explanation can fully account for this variation, but Michael Hechter’s (1975, 2000, 2013) writings offer a useful starting point. In short, Hechter (2013, p. 43) argues that alien rule gains local acceptance to the extent that reforms make it, first, more native and, second, more capable.² Alien rule becomes more ‘native’ the more inclusive it is. Inclusiveness can arise from co-opting selected locals into the state (indirect rule) or from acculturating all locals to its culture (integration or assimilation). Alien rule becomes more ‘capable’ the more public goods it delivers without regard to ethnic markers. Let us take a closer look at the logic pushing locals to accept alien rule under these conditions.

To begin with, resistance softens if selected natives are co-opted and charged with enforcing order in the borderland (Hechter 2000, pp. 35–55). Empires often use this tactic of indirect rule since compliance is easier to monitor with the help of local mediators, who also make central control look less ‘alien’ in the peripheries. Soviet nativisation policies are a prime example (Martin 2001). According to Stalin (1920), Bolshevik rule could be made ‘near and dear’ to minorities by staffing state organs in border regions with ‘local people acquainted with the manner of life, habits, customs, and language of the native population’. Russia’s policies towards the peoples of the North Caucasus are still planned along these lines (Siroky et al. 2013; Hall 2015). However, local mediators can also turn against the centre, so some alien rulers prefer to enforce direct control over their borderlands. In order to prevent natives from calling for national self-determination, it then becomes essential to integrate or assimilate them (Hechter 2000, pp. 56–69). For this state-building nationalism to succeed, alien rulers must deliver public goods without (ethnic) inequities in the allocation process.

Four possible scenarios result from these tactics. First, if alien rulers fail to co-opt natives, and fail to deliver public goods in an equitable fashion, then beached minorities face strong incentives to reject their host state. Second, if alien rulers do co-opt natives, but fail to deliver public goods in an equitable fashion, then beached minorities face weak incentives to reject their host state. Third, if alien rulers fail to acculturate natives, but deliver public goods in an equitable fashion, then beached minorities face weak incentives to accept their host state. Fourth, if alien rulers acculturate natives and deliver public goods in an equitable fashion, then beached minorities face strong incentives to accept their host state.

In line with these predictions, this essay focuses on the cultural and instrumental policies that alien rulers use to garner acceptance from beached minorities. To be sure, some alien rulers are indifferent to these minorities and prefer to engage in predation to extract disproportionate rents for their in-group, thus creating a situation similar to internal colonialism. This form of abusive rule often arises when interethnic interactions are rare, as social controls then fail to restrain the alien rulers. However, if it dawns on them that in the long run their own fortunes are entwined with those of their subjects, then the aliens often include natives. Alien rulers can either share their spoils through selective co-optation and thus stabilise the borderland, or offer all natives equitable access to public goods and thus entice them to acculturate into the host state (Hechter 2000, p. 62; 2010, p. 422; 2013, p. 44).

Discerning whether stranded minorities ‘reject’ or ‘accept’ the imposition of alien rule is, however, no simple task. Rejection manifests itself in overt forms, such as protests or other expressions of discontent against the host state’s policies. Nevertheless, acceptance cannot be inferred from the absence of demonstrations, since compliance in the face of sanctions suggests subservience (Hechter 2013, p. 19). We thus face the challenge of differentiating enforced subservience from genuine acceptance. To disentangle them, I use a mixed-methods approach. I first use interviews with Azeri elites to map patterns of protests and compliance over time, and then a matched-guise experiment to probe Azeri attitudes to Georgia’s efforts to integrate/assimilate the borderland. Measuring Azeri perceptions of the social distance between Azeris and Georgians is crucial, since ‘alien rule’, by Hechter’s (2013, p. 28) own admission,³ cannot earn strong acceptance unless the latent distinction itself loses significance. With this model in mind, let us examine the initial reactions of Georgia’s stranded Azeris.

Beached in Borchali

The Marneuli, Bolnisi and Dmanisi districts have been part of the Georgian region of Kvemo Kartli—meaning ‘Lower Kartli’, in Georgian—since the 1990s. However, Azeri residents of these lands have lived under alien rule before. Centuries ago, their ancestors in the Turkic Borchalu tribe settled in the area, then part of the Persian Empire. It later became part of the Borchali uezd (district) of the Tiflis gubernia (governorate) in the Russian Empire. The 1897 census classified 29.4% of this larger district’s population as ‘Tatars’, as Azeris then were called (Komakhia 2004). After the tsar was ousted in 1917, the peoples of the South Caucasus first set up a Transcaucasian Federation. Soon, however, it dissolved into separate Georgian, Armenian and Azerbaijani republics. As a result, Borchali’s Turkic inhabitants ended up enclosed in the Democratic Republic of Georgia in 1918. Its leaders considered the Georgian language an essential tool to strengthen the nation and integrate minorities (Nodia 2009, p. 89; Berglund 2016a, pp. 40–55).⁴ ‘Our duty is to create a whole firm national body … minorities must acknowledge that our republic is their homeland too’, the head of the Constitutional Commission argued (Amirejibi 2011, p. 210). In 1918, when Georgian officials sent troops to Borchali to defend the borders of the republic, some Tatars fled; others addressed petitions to Istanbul and Baku entreating their kin states to rescue them from Georgia’s grip (Yilmaz 2009, p. 42).

After the Sovietisation of the South Caucasus in 1921, the area fell to the Georgian SSR, and in the 1940s towns in the region were given Georgian names: Borchali became Marneuli and Bashkicheti became Dmanisi.⁵ In the meantime, Soviet officials began referring to the Turkic residents of the region, first as ‘Turks’ and then as ‘Azeris’ (Yilmaz 2013). Their life in the Georgian SSR developed in social isolation from the titulars. Ethnic Georgians administered the republic in their mother tongue, but Azeris in the borderland districts, who seldom spoke Georgian, instead fell back on Russian as the language of interethnic communication (Broers 2004, pp. 89, 258).

In 1989, 179,000 Azeris resided in the Marneuli, Bolnisi and Dmanisi districts.⁶ As titulars in the Georgian SSR moved towards independence, these locals were—once again—beached in Borchali, separated from their kin state, and left to face a harsh campaign of internal colonisation. Zviad Gamsakhurdia and other Georgian dissidents calling for Georgian ‘ownership’ of their republic contested the presence of non-Georgians in Georgian lands and unveiled plans to revive the Georgian Orthodox Church and promote the titular language (Cornell 2002b, pp. 159, 212; Hewitt 2013, pp. 57–80). After coaxing the communist authorities to ban regional parties and in so doing disenfranchising minorities, Gamsakhurdia’s Round Table coalition won the 1990 parliamentary elections. In 1991, when the Republic of Georgia declared independence, Gamsakhurdia became president. His rule soon became associated with the mantra ‘Georgia for the Georgians’ (Gamsakhurdia 1990). Regarding minorities as ‘newcomers’, ‘brought here by the Kremlin, by Russia, by the Empire’, he even swore to ‘save’ Georgia from ‘absorption by other nationalities’ (Shane 1991).

In accordance with Hechter’s first scenario, this campaign to exclude and discriminate against minorities created strong incentives for stranded Azeris to reject the imposition of alien rule. Most backed the communists in the 1990 elections. After the Round Table coalition won, Gamsakhurdia dismissed Azeri officials from their posts in the borderland, threatened to withhold citizenship and the right to own land from minorities opposing Georgian independence, and appointed an ethnic Georgian prefect to govern the area. This attempt to enforce direct rule triggered protests on such a scale that martial law had to be declared (Slider 1997, pp. 170–76). Feeling targeted on the basis of their origin, locals founded the Geyrat (Honour) movement.⁷ Some of its members wanted an autonomous ‘Borchali’, but this demand failed to gain traction.⁸ Nonetheless, locals formed self-defence units in response to nationalist Georgian gangs intent on expelling them from the area (Kaufman 2001, p. 110; Cornell 2002b, p. 159; Wheatley 2005, p. 13). In Marneuli, self-defence units succeeded in shielding Azeris from harassment, but Georgian bands forced Azeris to flee Bolnisi town, and adjacent villages were assigned Georgian place names. In parallel, Georgian eco-migrants settled in Bolnisi and Dmanisi—sometimes in the homes of Azeris the state had failed to protect (Trier & Turashvili 2007, p. 33).

By late 1991, Gamsakhurdia had alienated not just minorities but Georgians too through his polarising policies and dictatorial persona (Kaufman 2001, p. 126). An alliance of democratic intellectuals, displaced communists and other opposition politicians ousted him and invited Eduard Shevardnadze, former leader of the Georgian SSR, to take his place in what they assumed would merely be a figurehead role. At first, Georgia fell into ‘a quasi-medieval condition, with separate fiefdoms ruled by different warlords’, but Shevardnadze soon outfoxed his rivals, restored a degree of order and was elected president (Nodia 1995, p. 111). To strike a balance between Georgian nationalists and estranged minorities, he decided to ‘let sleeping dogs lie’ by, as far as possible, avoiding ethnic controversies (Nodia 2002, p. 6). One the one hand, Shevardnadze paid tribute to multiculturalism in his rhetoric: Georgia’s 1995 constitution declared all citizens equal, and the ruling Citizens’ Union of Georgia (sakartvelos mokalaketa k’avshiri) abolished ethnicity markers in passports. On the other hand, he flirted with Orthodox nationalists and did little to either protect or integrate minorities (Jones 2006, p. 260; Sabanadze 2010, p. 104).

Shevardnadze’s reforms had appeasing effects. He reined in criminals and co-opted notables in the region, but in the process he also farmed out state institutions to corrupt figures (Hale 2014, p. 152). Levan Mamaladze became his liaison in the Azeri borderland. After merging Marneuli, Bolnisi and Dmanisi with three adjacent Georgian-populated districts into the Kvemo Kartli mkhare (region), Shevardnadze appointed Mamaladze, an ethnic Georgian, as its governor. He ran the borderland through patronage and indirect rule, wielding influence over the nomination of majoritarian MPs from Kvemo Kartli as well as the appointment of district administrators. Even local businessmen were organised into a confederation of entrepreneurs under his oversight (Wheatley 2005, p. 15). Meanwhile, to facilitate monitoring and make his control look less ‘alien’, Mamaladze co-opted influential Azeris. In exchange for their allegiance, Geyrat activists took minor posts in the borderland and at times stepped in to mediate disputes, in effect serving as colonial subalterns.

In line with Hechter’s second scenario, this lessened locals’ incentives to reject alien rule but hardly endeared it to them. Corruption rendered the host state incapable of delivering public goods, let alone doing so in an equitable fashion (Nodia 2002, pp. 13–8; Komakhia 2004; Wheatley 2005). Without speaking the state language, Azeris struggled to find jobs and protect their legal rights. Most civil servants were ethnic Georgians and demands for bribes were frequent, so interactions with officialdom often gave rise to frustration. To boot, well-connected Georgians acquired much of the arable soil in the region, leaving some Azeris no choice but to lease land at steep prices. Though Mamaladze’s patrons delivered local votes to Shevardnadze at election time, Azeris in the borderland felt alienated from their host state and lost faith in Geyrat as a defender of public interests.⁹ Locals became resigned to Shevardnadze’s rule and eked out a living selling illegal imports or agricultural produce (Berglund 2016a, p. 74).

Onset of inclusive state-building nationalism

After the 2003 Rose Revolution, the Georgian authorities adopted a different approach to the Azeris beached in Kvemo Kartli. Whereas President Shevardnadze had been content with controlling them, his successor, Mikheil Saakashvili, set out to gain their acceptance through an inclusive state-building nationalism (Hechter 2000, p. 62). Incoming officials from Saakashvili’s United National Movement (ertiani nacionaluri modzraoba—UNM) were Western-educated, Anglophone and liberal (Berglund 2018). Several had non-Georgian roots or adhered to religions other than the Georgian Orthodox Church,¹⁰ causing some opponents to disparage their rule as ‘anti-national’. Saakashvili therefore considered it to be both in his personal self-interest, and in the long-term interest of the Georgian state, to foster a more inclusive national conception: ‘we need to know our history. And our history teaches us that tolerance is the basis for sovereignty in our region. It is not only a moral duty: it is an issue of national security’ (Saakashvili 2013).

Mindful of the difficulties in garnering acceptance from locals as long as corrupt officials kept diverting public goods into their own hands (Hechter 2000, p. 56; 2013, p. 40), Saakashvili’s first order of business became to craft a state capable of enforcing its policies throughout its realm. He pushed through constitutional amendments that increased his leverage over parliament and courts and entitled him to choose the interior minister and chief prosecutor. Utilising these levers, he reformed core institutions. The traffic police, who had made their living extorting motorists,¹¹ was disbanded, and a new force was set up from scratch. State agencies were downsized and opportunities for extracting bribes dried up. Crooked officials and crime bosses were arrested and paid huge fines into state coffers to avoid conviction. Saakashvili then raised civil servants’ salaries so that they could get by without asking for kickbacks, thus creating a pliant state apparatus (World Bank 2012; Bolkvadze 2016). Civil servants began enforcing decrees and delivering public goods, even in the peripheries, resulting in lower crime rates, less smuggling, growing tax revenues and much-needed repairs to infrastructure.

As Azeris became more dependent on the state, it became more important to signal that public goods were to be shared among citizens through an equitable allocation process (Hechter 2013, p. 53). To emphasise his commitment to fighting discrimination, Saakashvili altered the state flag and national anthem and adopted an emblem sporting the motto: ‘Strength is in unity’. He spoke about moving ‘Forward to David the Builder’, evoking the feats of a medieval king who united the divided Georgian lands, and sprinkled speeches with references to minorities as part and parcel of the national fabric (Berglund 2016b). ‘Every citizen who considers Georgia his homeland … is our greatest wealth and treasure’, Saakashvili said at his presidential inauguration (Saakashvili 2004). ‘It is our responsibility’, he reiterated on another occasion, ‘to maintain the multi-ethnic and multi-confessional Georgia, which has been left to us by our ancestors’.¹² In this vein, he declared the Azeri Nowruz celebrations a ‘Georgian National Holiday’, telling locals in Marneuli: ‘you are our flesh and blood, you constitute a significant part of Georgia. We will fight together for a better future and we will celebrate together’.¹³

In light of the need to acculturate Azeris to their host state and discourage locals from protesting against their enclosure inside the Georgian national realm (Hechter 2000, pp. 63, 71), Saakashvili strove both to entice them to learn the state language and to protect them from discrimination (CITC 2008).¹⁴ He ratified the Framework Convention for the Protection of National Minorities, recruited an adviser on civil integration to chair a Civil Integration and Tolerance Council, and appointed a state minister for reintegration. Despite coordination troubles, these actors crafted a ‘National Concept and Action Plan for Tolerance and Civic Integration’, instructing state agencies to turn minorities into co-nationals (Government of Georgia 2009a).

To ensure that Azeris had equitable social and professional opportunities inside Georgia, Saakashvili initiated reforms promoting the state language. He began enforcing laws obliging civil servants to perform their duties in Georgian. Officials, from doctors to school directors, had to pass language tests. But, since mastering a foreign tongue is a time-consuming process, the tests ended up endangering the career prospects of ethnic Azeris, and made officials less inclined to accommodate locals unable to speak Georgian (CIPDD 2006, p. 7; ICG 2006, p. 23). Later on, the authorities established an affirmative action scheme to recruit more ethnic Azeris into the local police departments (Government of Georgia 2012, p. 112). Moreover, ‘Language Houses’ offering Georgian classes to adults opened in Bolnisi and Dmanisi in 2009 and Marneuli in 2011 (Berglund 2016a, p. 112). In parallel, Saakashvili tried to ‘win over’ the next generation of Azeris.¹⁵ Georgian-language teachers received updated textbooks and were allotted more hours, schools in the region piloted multilingual classes, and ‘befriending programmes’ enabled exchanges between Georgian and Azeri pupils (SMR 2014). However, given the lack of qualified teachers, locals failed to learn enough Georgian to pass the university entrance exams.¹⁶ From 2009, Azeris were permitted to take the exam in their native language and had access to another affirmative action scheme that enabled them to spend two semesters learning Georgian before starting their degree programmes.

Understanding that groups mobilise around markers that affect their welfare (Hechter 2000, p. 98), Saakashvili paired these nationalising policies with safeguards against discrimination. He added provisions to the 2004 Law on Higher Education (Art. 3), the 2005 Law on General Education (Art. 13) and the 2006 Labour Code (Art. 2), and strengthened antidiscrimination clauses in the criminal code. The public defender’s office opened a branch in Marneuli. Public radio and public television aired news segments in Azeri and produced programmes like Our Georgia and Our Yard, showcasing Azeri traditions. The Azeri-language newspaper Gurjistan obtained state funding, as did museums, theatres and festivals popularising Azeri culture (Government of Georgia 2009b, 2012). Ethnocentric Georgians opposed this inclusiveness, and unidentified Christian zealots even erected crosses outside Azeri settlements. While Saakashvili sought to placate the Georgian Orthodox Church with offers of lavish transfers from the budget and a measure of influence in state affairs (US Department of State 2011),¹⁷ its protests did not prevent his government from granting religious minorities equal legal status.¹⁸

Compliance among Azeri elites

How did Azeri elites react to this turn towards inclusive state-building nationalism after the Rose Revolution? The reforms resulted in a host state more capable of delivering public goods in an equitable fashion (scenario three) and taking ever more concerted steps to acculturate minorities (scenario four) so Hechter (2013, p. 43) would predict increased acceptance for the alien rulers. However, the transition triggered an acrimonious struggle among elites. Azeris co-opted under Shevardnadze’s administration lost their mediating role and tried to mobilise the populace against the central authorities, but most locals no longer considered ethnic discrimination to be the source of their hardships and did not heed this call. While Hechter (2000, p. 70; 2013, p. 50) recognises that direct rule can provoke protests, these reactions suggest that even the inclusive kind of state-building nationalism can be a hard sell.

Saakashvili’s first challenge came from the region’s governor: Levan Mamaladze. Referring to an incident in Tbilisi prior to the 2002 local elections, he rebuked the incoming president as an anti-Azeri politician (Komakhia 2003, p. 22).¹⁹ In return, Saakashvili labelled Mamaladze a ‘criminal and bandit’ on account of his past misdeeds.²⁰ Fearing arrest, Mamaladze fled abroad, and Saakashvili appointed a new governor of Kvemo Kartli who began his tenure by observing that ‘many local residents do not even understand what country they live in’ (Fuller & Giragosian 2004). In order to avoid Mamaladze’s fate, elected officials in the region soon resigned, and allies of the ruling UNM filled their posts. Even local voters fell in line behind Saakashvili after the 2004 repeat elections to parliament (Lobzhanidze 2004). Nonetheless, the centre’s tightening grip over the borderland generated distrust.

In the spring of 2004, hundreds of locals gathered in Marneuli to demonstrate against ongoing personnel changes in the district administration, occurring well before the end of their electoral mandates and under pressure from the centre. Azeris ‘are forced out of their jobs and new people, who belong to the ruling clan [the United National Movement] are appointed’, Geyrat’s leader Alibala Askerov complained (Ismailzade 2004a; ICG 2006, pp. 6, 11). Customs officers at crossing points to Azerbaijan were among those replaced with Georgians. Officials countered that these ‘outsiders’, who lacked local allegiances, were more eager to cooperate with efforts to shut down cross-border smuggling routes.²¹ After periodic clashes with contraband traders, illicit markets disappeared; affected Azeris experienced the crackdown as a blow to their livelihood (Peuch 2004a, 2004b; Alieva 2005; George 2008). Concern about inequitable access to arable land also featured high on the agenda. In late 2004, an Azeri died after a scuffle on a Georgian estate of more than a thousand hectares, where locals had to rent land at inflated prices. In subsequent protests, Azeris resorted to road blockades to press for land reform (Bukia 2004; Alieva 2006).

Some activists interpreted these socio-economic gripes through a more sinister lens. The requirement that civil servants speak Georgian, the closure of contraband markets, and the unfair distribution of land made it so difficult to earn a living that Azeris might leave of their own accord, resulting in a ‘cultured and soft’ eviction (Tolerance Public Association for Human Rights Protection 2008, p. 28). The use of entrance exams that required students to understand Georgian likewise impeded Azeris’ access to universities, precipitating an exodus of students to Azerbaijan. Other aspects of Saakashvili’s education reforms fed unease too. Askerov lamented that ‘Schools are losing their role as an incubator for our culture’ as Azeri pupils ‘read textbooks [from Tbilisi] claiming that we migrated here just 200 years ago’.²² ‘Veiled discrimination’, critics charged, also stemmed from Saakashvili’s failure to restore historic place names and from his reluctance to grant equal legal status to religious minorities (MRMG 2011a, 2011b). In short, some Azeri activists in the borderland tried to challenge ‘the incursions of the central state’ by ‘playing the alien card’ to mobilise locals (Hechter 2013, p. 50).

Aware of this danger, Saakashvili courted his counterpart in Baku. He visited the president of Azerbaijan, Ilham Aliyev, in March 2004 and received him four months later, when the parties reaffirmed their strategic partnership.²³ ‘You are ethnic Azerbaijanis, but you are also citizens of Georgia. You need to try to further integrate into Georgian society’, President Aliyev told Marneuli’s residents in June 2004 (Ismailzade 2004b). Despite Baku’s cooperation, recurring Russian subversion efforts against Georgia led Saakashvili to keep Azeri activists under close watch.²⁴ Dashgin Gulmamedov of the National Assembly of Azeris in Georgia (Gürcüstan Azərbaycanlılarının Milli Assambleyası) experienced this first-hand. From his base in Baku, he accused Saakashvili of criminal misdeeds, called for the creation of a Georgian confederation of constituent regions, and circulated an inflammatory publication—Hümmət—in the borderland. It had funding from a Georgian–Azeri émigré close to the Kremlin who in August 2006 was accused of disseminating bellicose leaflets in Marneuli (ICG 2006, pp. 14, 20; Ismailzade 2006; Wheatley 2009, pp. 19, 34).²⁵ To control the situation, Georgian officials pressured independent candidates to drop out of the local elections that autumn, resulting in landslide victories for the ruling UNM (ODIHR 2006, p. 9).

In response, Gulmamedov stepped in as campaign manager for Fazil Aliyev, another harsh critic of the authorities, who was running against Saakashvili in the 2008 presidential elections. However, Georgian police arrested Gulmamedov upon his return, accusing him of inciting ethnic hatred.²⁶ Aliyev dropped out of the presidential race but then tried to make a comeback with the help of the Christian-Democratic Alliance (qristian-demokratiuli aliansi),²⁷ which nominated him as its candidate in Kvemo Kartli’s Gardabani district for the legislative elections later that spring. Aliyev came up short with a meagre 1.5% of the vote, but his erstwhile partner, Gulmamedov, won a high profile in the Russian media. He condemned the ‘liquidation of Azeri schools in Borchali’ and other acts of ‘chauvinism’, rejected NATO accession, and advocated making Azeri an official language, predicting that ‘the Georgian state will collapse’ unless it heeded his advice.²⁸ Not long after venting these opinions, the authorities in Azerbaijan—perhaps at the request of their counterparts in Georgia—arrested Gulmamedov on charges of fraud.²⁹

In parallel, the more moderate Congress of Georgian Azeris (Gürcüstan Azərbaycanlıları Konqresinin) ran afoul of the authorities in Tbilisi (Tolerance Public Association for Human Rights Protection 2008, p. 32). Its founder, Ali Babaev, did not question the need for integration as much as the policies adopted to this end.³⁰ His run for parliament on behalf of the Republican Party (respublikuri partia) in the Gardabani district in 2008 mustered a mere 4% of the vote. After accusing officials of treating Azeris as ‘second-class citizens’, Babaev’s congress became the target of a hostile takeover, whereby Kvemo Kartli’s deputy governor, Hussein Yusubov, usurped the chairmanship.³¹ Yusubov, alongside other Azeri officials from the ranks of the UNM, also set up the All-Georgia Muslim Administration (sruliad sakartvelos muslimta sammartvelo) in mid-2011. Its founders argued that it offered Georgian Muslims their own administration, independent of both the Baku-based Caucasus Board of Muslims and the Iran-linked Ahl ul-Bayt Foundation. Sceptics considered it as another instrument of state control (Prasad 2012).

‘Trust in Allah, but tie your camel’, the Georgian authorities seemed to reason, according to one activist remarking on their hesitance to entrust Azeris with official posts.³² Nonetheless, most campaigners in Kvemo Kartli considered knowledge of Georgian as essential to escaping their isolation and gaining representation in state structures: ‘you have to speak Georgian to be a functioning citizen of Georgia’.³³ Echoing this sentiment, another respondent asserted: ‘we are Georgian citizens, and want to be accepted as such’.³⁴ Support for the idea of granting the Azeri language official status had little appeal, as it could ‘undercut incentives to learn Georgian’ and ‘leave us more segregated’, Marneuli’s MP Azer Suleimanov argued.³⁵ Even though fringe activists and those displaced from their roles as mediators after the Rose Revolution rejected the policies of their host state,³⁶ most influential Azeri power brokers in the borderland were embracing Georgia’s turn towards inclusive state-building nationalism.

It is, of course, possible that Saakashvili’s controlling tactics led activists to defer to his policies without accepting them in earnest. It is also possible, however, that ethnic entrepreneurs had little success, since the demand for self-determination decreases ‘to the degree that the centre provides [beached minorities] with protection, and economic and social welfare’ (Hechter 2000, p. 113). Other data support the second interpretation. According to a 2011 poll, 96% of minorities in Kvemo Kartli reported a desire to learn Georgian, 97% saw Georgia as their ‘motherland’, and a mere 3% considered Tbilisi’s attitude to them as a ‘barrier’ (CIDA 2011). Local Azeris were also staking their children’s future, and their own, on morphing into Georgian-speaking co-nationals, as shown by the increasing numbers of parents enrolling their children in Georgian-language schools.³⁷ The share of ethnic Azeri pupils in Marneuli town’s Georgian school reached 70% (Sultanova 2012). Once the affirmative action scheme entered into force, students from the borderland also flocked to universities in Georgia. In 2010, 226 Azeris from Marneuli, Bolnisi and Dmanisi took the entrance exam; 130 passed. In 2013, 807 took the exam and 493 passed.³⁸

Acceptance among Azeri adolescents

In light of Hechter’s (2000, 2013) prediction that minorities are more inclined to accept their host state if tempted to acculturate, this section maps Azeri perceptions of the social distance separating them from Georgians after the Rose Revolution. However, extracting the ‘reliable subjective evidence’ needed to reveal whether Azeris look down on or up to Georgians poses a formidable methodological challenge (Hechter 2013, pp. 19–21). Since the anticipation of sanctions from the state or one’s social milieu can taint answers to direct questions posed in interviews or in polls, I administered an experiment among 483 Azeri adolescents to disclose their implicit attitudes towards the prospect of becoming Georgian co-nationals, building on an established sociolinguistic technique known as a matched-guise test (Lambert et al. 1960; Romaine 1995, p. 289). Political scientists (Laitin 1998; Driscoll et al. 2016) have used the technique. Because inclusion can entail different degrees of amalgamation into the titular nation, I adapted the technique and designed the experiment to measure attitudes both towards linguistic integration and cultural assimilation.

I recorded a standard Georgian speaker and a standard Azeri speaker, both females in their early 20s with similar aural characteristics, each reading the same text aloud in her native language. These recordings were later played back to respondents who were asked to assess the voices in terms of 17 traits using a Likert scale (for example, ‘Is she reliable?’ ranging from 1, ‘very little’ to 6, ‘very much’). I told respondents that I was ‘interested in how people form impressions about others by hearing their voices’, but I did not inform them in advance that both recordings recurred twice: using name tags characteristic of different ethnicities, the speaker in each recording was presented once as a monocultural person and once as someone posing as a member of the out-group (see Table 1). After listening to each recording, respondents jotted down their reactions on a questionnaire. This enables us to compare their rating of these recordings, holding extraneous variables constant, and infer whether the same person is seen in a different light depending on which language she speaks (vertical comparisons) or her ethnic origin (horizontal comparisons). Using a small amount of deception, we can thus ‘match’ reactions to these ‘guises’ and calculate the social costs or gains associated with different degrees of amalgamation.

TABLE 1 The Logic of the Experiment as Seen from the Perspective of Azeri Respondents

	Georgian name tag	Azeri name tag
Standard Georgian speaker recording	Tamar Maisuradze speaking Georgian ‘One of them’	Afa Mamedova speaking Georgian ‘One of us posing as one of them’
Standard Azeri speaker recording	Tamar Maisuradze speaking Azeri ‘One of them posing as one of us’	Afa Mamedova speaking Azeri ‘One of us’

First, we can compare the lower right box (‘one of us’) to the upper right box (‘one of us posing as one of them’) to see if Azeris punish peers for linguistic integration, that is, for speaking Georgian rather than Azeri. In the second step, we can compare the upper right box (‘one of us posing as one of them’) to the upper left box (‘one of them’) to reveal whether Azeris punish cultural assimilation, that is, fusing into the titular nation rather than merely speaking its language. The premise of the experiment is that an individual’s name and language serve as a ‘convenient cognitive shorthand for rapidly inferring a wide range of information about a person’ (Hale 2008, p. 243; 2017). For it to work as intended, though, respondents had to pick up on these cues without noticing the ruses involved; namely, that the same recording is presented under different names or that different speakers are voicing the same ‘persona’. I therefore arranged focus groups and a pilot round to test the resonance of the speakers and the name tags.³⁹ I also inserted dummy voices between the recordings that pertained to the experiment, keeping those recordings as far apart from one other as possible, and made sure that respondents could not go back in the questionnaire to compare their ratings for (what they thought was) the same person. Thanks to these ruses, almost no respondents showed signs of ‘seeing through’ the experiment.

Aware of state officials’ focus on integrating the first generation of Azeris born into independent Georgia after the Soviet collapse, I chose to administer this experiment in high schools. In December 2011, I obtained consent to access pupils in grades 10–12 in 20 public schools located throughout the borderland. A total of 483 students from these schools participated in the experiment and listed their nationality as ‘Azeri’ at the end of the questionnaire. They studied in Marneuli (270 students), Bolnisi (159) and Dmanisi (54). Males made up 53% of respondents and 47% were female. The majority were born in the borderland and spoke Azeri at home (96%) and with friends (90%).⁴⁰ Research assistants fluent in English, Georgian, Azeri and Russian accompanied me to each school. We introduced the experiment to students in their regular medium of instruction and had them fill in the questionnaire in Azeri, Georgian or Russian; almost all chose Azeri. To attain as reliable data as possible, completed questionnaires were registered into a computer twice, then screened for discrepancies and outstanding errors corrected.

Did Azeri adolescents look down on, or up to, Georgian-speaking Azeris? Table 2 summarises the 483 respondents’ attitudes towards linguistic integration. It details their reactions to hearing recordings presented under the same Azeri name tag. The left column lists the average scores given to Afa Mamedova as she spoke Azeri (representing ‘one of us’ in Table 1), and the right column lists the average scores for the same Afa Mamedova as she instead spoke Georgian (representing ‘one of us posing as one of them’ in Table 1). In the rows, we see how the Azeri respondents rated these ‘guises’ in terms of questions A–Q: ‘Is she [relevant attribute]?’. The last row presents the average rating across all traits. Higher scores represent more positive evaluations on a scale from 1 (‘very little’) to 6 (‘very much’). If respondents assess the Georgian-speaking Azeri (right column) as less likable, it amounts to a rejection of integration into their host state.

TABLE 2 Implicit Attitudes Towards Linguistic Integration

Attribute		‘One of us’	‘One of us posing as one of them’	p-value
A	Educated	4.86	4.81	0.613
B	Cultured	4.89	4.65	0.000
C	Reliable	4.56	4.35	0.001
D	Witty	4.48	4.77	0.000
E	Open	4.12	4.25	0.029
F	Self-confident	4.25	4.40	0.050
G	Magnanimous	4.15	4.18	0.555
H	Pleasant	4.47	4.46	0.674
I	Attractive	4.24	4.30	0.677
J	Spiritual	4.33	4.25	0.231
K	Intellectual	4.57	4.64	0.485
L	Proud	4.23	4.25	0.669
M	Leader	4.04	4.34	0.000
N	Hardworking	4.42	4.59	0.013
O	Amusing	3.96	4.18	0.002
P	Wealthy	3.91	4.24	0.000
Q	Patriotic	4.87	4.85	0.972
Average		4 . 38	4.42	0.229

To facilitate interpretation of these results, the ‘one of us’ and ‘one of us posing as one of them’ columns are shaded in three different hues: dark for lower scores, white for higher ratings and light grey when the difference between the means is too small to be significant according to the result of a two-tailed t-test (p > 0.05). What stands out, through the lens of this interpretive grid, is that Azeris do not differentiate much between ‘one of us’ and ‘one of us posing as one of them’. Regardless of whether Afa Mamedova speaks Azeri or Georgian, she is evaluated in about the same fashion on eight of the 17 character traits (A, G, H, I, J, K, L, Q). Respondents considered her Georgian-speaking guise less cultured (B) and less reliable (C), but they rated it as more likeable across seven other character traits (D, E, F, M, N, O, P). Though the overall difference between the columns is insignificant, Table 2 permits the conclusion that Azeri adolescents do not subject Georgian-speaking peers to in-group scorn. Their attitudes towards linguistic integration are permissive or even positive.

Did Azeri adolescents look down on, or up to, Georgians? Table 3 turns to this question, detailing the 483 respondents’ reactions after having listened to the same Georgian-speaking recording. The left column contains the average scores given to the Georgian speaker introduced as Afa Mamedova (representing ‘one of us posing as one of them’); the right column displays the average scores for the same recording, presented as that of Tamar Maisuradze (representing ‘one of them’). The rows show how respondents evaluated the identical recordings, introduced under disparate name tags, in terms of items A–Q and, last but not least, in terms of the average ratings across all attributes. As before, the scale runs from 1 (‘very little’) to 6 (‘very much’). If respondents assess the Georgian (right column) as less likable it amounts to a rejection of assimilation into their host state.

TABLE 3 Implicit Attitudes Towards Cultural Assimilation

Attribute		‘One of us posing as one of them’	‘One of them’	p-value
A	Educated	4.81	4.67	0.026
B	Cultured	4.65	4.73	0.396
C	Reliable	4.35	4.10	0.002
D	Witty	4.77	4.62	0.061
E	Open	4.25	3.98	0.000
F	Self-confident	4.40	4.28	0.047
G	Magnanimous	4.18	3.91	0.000
H	Pleasant	4.46	4.26	0.010
I	Attractive	4.30	4.02	0.000
J	Spiritual	4.25	4.17	0.105
K	Intellectual	4.64	4.68	0.375
L	Proud	4.25	4.09	0.021
M	Leader	4.34	4.14	0.018
N	Hardworking	4.59	4.56	0.444
O	Amusing	4.18	3.65	0.000
P	Wealthy	4.24	4.06	0.011
Q	Patriotic	4.85	4.77	0.321
Average		4 . 42	4.26	0.000

Using the same interpretive grid as before, we can infer that the 483 Azeri adolescents assess the same Georgian speaker as less likable when presented as Tamar Maisuradze (‘one of them’) rather than as Afa Mamedova (‘one of us posing as one of them’). To be sure, respondents did not differentiate between the guises in terms of some characteristics (B, D, J, K, N, Q). Not in a single case, though, did the Georgian strike them as more likable than their ethnic kin. By contrast, Afa Mamedova received higher scores in terms of 11 attributes (A, C, E, F, G, H, I, L, M, O, P), and the difference is significant in all these cases. Table 3 supports the conclusion that Azeri adolescents prefer Azeris posing as Georgians to regular Georgians, and thus subject Georgian-speaking peers who attempt to assimilate to ‘in-group scorn’, as Laitin (1998, p. 56) dubs it. Cultural assimilation, or becoming Georgian, is looked down on.

Taken together, these results suggest that the 483 adolescent Azeris accept linguistic integration but reject cultural assimilation. Locals can learn the state language, or switch from ‘one of us’ to ‘one of us posing as one of them’, without incurring social costs inside the borderland. The matched-guise experiment elicited implicit reactions untainted by self-censorship, implying that Azeris’ readiness to integrate was not born of fear of external sanctions but a result of earnest choice. Still, one cannot assimilate, or switch from ‘one of us posing as one of them’ to ‘one of them’, without losing in-group status in the borderland. Thus, the upcoming generation of Azeris beached in Kvemo Kartli hold tolerant attitudes towards their host state’s policies of acculturation. Morphing into a co-national is acceptable if this entails learning the titular language while holding on to one’s cultural origins.

Moving past alien rule

How did ethnic Azeris in Marneuli, Bolnisi and Dmanisi—Georgian districts bordering Azerbaijan—react to the reorganisation of political space along national lines following the Soviet Union’s demise? The disintegration of the USSR left countless minorities ‘beached’ inside foreign countries, not after having moved there, but as a result of the sudden border demarcations that occurred in 1991. These groups became subjects of host states pursuing policies promoting the dominance of their titular nationalities. Minorities stranded in borderlands, separated from their kin state, experienced this as a traumatic turn of events and often mobilised in defence of their rights, sometimes with support from the external homeland. Ethnic conflicts flared across Eurasia as beached minorities struggled to come to grips with their ‘alien rulers’. Some rose up against their host state; others submitted to it. Russophones in Transnistria rejected Moldovan control, but Russians in Narva acquiesced to life in Estonia. Armenians in Nagorno-Karabakh lost no time in rebelling against Azeri rule, but Armenians in Akhalkalaki assented to Georgian rule. Against the backdrop of these conflicts and near-conflicts in post-Soviet Eurasia, this essay set out to investigate reactions to the imposition of Georgian rule among ethnic Azeris beached in three districts across the border from their kin state.

To retrace their reactions to this specific predicament and probe the general conditions under which beached minorities accept their host state, I started with Hechter’s (2000, 2013) claim that alien rulers earn acceptance through cultural and instrumental policies. Minorities are predicted to accept their host state if included, either through indirect co-optation or direct acculturation, and offered access to public goods through an allocation process free of ethnic inequities. Azeris in Marneuli, Bolnisi and Dmanisi have been subjected to different policies under different Georgian presidents, so their responses offer us a rare chance to explore whether this recipe succeeds in generating acceptance from beached minorities. Thus, based on public reports and original data from elite interviews (n = 15) and a matched-guise experiment (n = 483), this essay has investigated whether the stranded Azeris, from Georgia’s rebirth up until 2013, reacted along lines recognisable to Hechter.

After a brief look at locals’ past experience of alien rule in the Democratic Republic of Georgia and the USSR, I mapped their reactions to President Gamsakhurdia’s rule. He made it clear that Azeris had no place in independent Georgia and introduced policies that discriminated against them, culminating in efforts to uproot them from the borderland. In line with Hechter’s predictions, locals felt compelled to mobilise around ethnic markers, creating the Geyrat movement and forming self-defence units to protect their rights. Policies changed under President Shevardnadze. He merged the borderland districts into the larger Kvemo Kartli region and appointed as its governor an ethnic Georgian, who asserted his control through patronage and co-optation. In return for minor posts in the borderland, influential Azeris helped the governor mediate disputes and monitor compliance. As Hechter would expect, these policies appeased locals by making Georgian rule look less ‘alien’ but it did not endear itself to Azeris either because the corrupt host state failed to offer even the most basic public goods.

In the wake of the Rose Revolution, Georgia’s policies changed once again as President Saakashvili and his team, some with non-Georgian roots, embarked on a campaign of inclusive state-building nationalism. Saakashvili’s anti-corruption reforms established a state capable of enforcing policies throughout its realm. As Azeris became more dependent on their alien ruler, he undertook to dispense public goods in a more equitable fashion and encouraged locals to acculturate into the domain of the host state. In exchange for protecting them against discrimination, he exhorted Azeris in Kvemo Kartli to learn the state language (Georgian). Hechter predicts that such policies foster acceptance, and for the most part, the ensuing examination of local reactions proves him right.

Co-opted Azeris displaced from their roles as mediators and some fringe activists sought to mobilise locals, but Saakashvili limited their scope for action. Most activists regarded the Georgian language as a tool for escaping their segregation, in effect backing his acculturation policies. Moreover, data suggest that this compliance arose not from fear of state sanctions, but from genuine acceptance. Increasing numbers of Azeris were enrolling in Georgian-language schools and universities, a choice predicated on the assumption that their command of the state language would improve the opportunities open to them, regardless of their Azeri origin. To further probe their interest in acculturating, I fielded a matched-guise test eliciting implicit attitudes towards linguistic integration and cultural assimilation. It indicated that local Azeri adolescents accept the former but reject the latter, and thus harbour attitudes that align with their host state’s policies for turning them into co-nationals—and for moving past the predicament of alien rule.⁴¹

Azeris beached in Kvemo Kartli, across the border from their kin state, had come to accept their host state as the distinction between ‘aliens’ and ‘natives’ faded after Georgia’s state-building nationalism took an inclusive turn. This finding is an empirical contribution in its own right. Past research on stranded minorities in post-Soviet Eurasia has neglected Georgia’s Azeris and focused on ethnic Russians in the ‘near abroad’ or groups rebelling against their host states. This essay also makes a theoretical contribution, lending credence to Hechter’s (2000, 2013) predictions concerning the mechanisms for legitimating alien rule. Georgia’s Azeris rejected their host state as it first excluded them and denied them access to public goods. However, once its policies began to include them and, at last, offer them access to public goods, minorities stranded in this borderland came to accept the prospect of morphing into Georgian co-nationals. Thus, host states in post-Soviet Eurasia that include their beached minorities (through co-optation or acculturation) and offer them equitable access to public goods appear more likely to garner acceptance from residents in restive borderlands.

Additional information

Notes on contributors

Christofer Berglund

Christofer Berglund, Assistant Professor, Department of Global Political Studies, Malmö University, Box 71, Malmö 20506, Sweden. Email: [email protected]

Notes

1 For example, even Russian businessmen close to the Kremlin have, to the latter’s chagrin, preferred to hold their capital in foreign (often Western) jurisdictions where their wealth is safe from domestic predation, thus revealing that instrumental concerns trump cultural ones (Jensen 2013; Movchan 2016).

2 I present Hechter’s ideas in a simplified form and at times use different terms to make the same point.

3 See Coakley (2010), Roeder (2010) and Roshwald (2010) for more on the limits of his arguments.

4 Georgian nationalists focused on the Georgian language as a cultural adhesive for two reasons. On the one hand, it united ethnic Georgian subgroups (Svans and Mingrelians) and even the Muslim Adjars with the dominant Kartvelians; on the other, it demarcated them all from Orthodox Russians.

5 Bolnisi once had the Turkic name ‘Choruk Gemerli’, but it became ‘Katharinenfeld’ after Germans settled there in 1819. It became Bolnisi in 1943, after the Germans were deported during World War II.

6 By 2002 their number had dropped to 166,000, and in 2014 it was only 134,000. Despite ongoing depopulation, the borderland has become more homogenous: the Azeri share of the population rose from 71% in 1989 to 75% in 2002 and 76% in 2014 (CSEM 2014; Berglund 2016a, p. 76).

7 Interview with Alibala Askerov, Geyrat’s leader, Marneuli, 3 November 2015.

8 As did the irredentist nationalism of Abulfaz Elchibey, Azerbaijan’s president between June 1992 and September 1993, who aspired to bring lost lands back into a ‘Whole Azerbaijan’.

9 See Mamuka Kuparadze’s 2006 film ‘The Alienation Syndrome’ (Tbilisi, Studio RE).

10 Zurab Zhvania, prime minister from 2004 to 2005, had Armenian origins. Ivane Merabishvili, a Catholic, served as interior minister from 2004 to 2012. Temur Yakobashvili, state minister for reintegration from 2008 to 2010, practised Judaism. Saakashvili himself had married a Dutch national.

11 The traffic police had accosted Azeris with particular fervour since their vehicles, when laden with agricultural products en route to markets, were easy to spot and a prime target for extracting ‘taxes’.

12 ‘Saakashvili Speaks of Unity in Independence Day Speech’, Civil Georgia, 26 May 2007, available at: https://civil.ge/archives/112446, accessed 3 March 2018.

13 ‘The President of Georgia Mikheil Saakashvili Declared Nowruz-Bairam as the National Holiday’, 2010, available at: http://www.saakashviliarchive.info/en/PressOffice/News/Releases?p=4873&i=1, accessed 3 March 2018.

14 Interview with Temur Yakobashvili, state minister for reintegration, Tbilisi, 29 July 2009.

15 Interview with Tamar Kintsurashvili, president’s adviser on civil integration, Tbilisi, 4 August 2010.

16 Interview with Bela Tsipuria, former deputy minister of education, Tbilisi, 6 August 2010.

17 ‘Georgia: Radical Orthodox Groups Stops Mosque Construction; Local Authorities Support Azeri Community’, Wikileaks, 9TBILISI1757, 23 September 2009, available at: http://www.wikileaks.org/plusd/cables/09TBILISI1757_a.html, accessed 3 March 2018; ‘Georgia: Church and State, Maintaining a Delicate Balance’, Wikileaks, 09TBILISI2106, 3 December 2009, available at: http://www.wikileaks.org/cable/2009/12/09TBILISI2106.html, accessed 3 March 2018.

18 ‘Thousands Protest Law on Religious Minorities Legal Status’, Civil Georgia, 10 July 2011, available at: https://old.civil.ge/eng/article.php?id=23728, accessed 3 March 2018.

19 In order to fulfil his role as Shevardnadze’s patron, Mamaladze had bussed in protestors from Kvemo Kartli to counter pro-Saakashvili demonstrators in the capital. Saakashvili denounced the former, many of whom were Azeris, as ‘pitiful people’, which Mamaladze then presented as an ethnic slur.

20 ‘Saakashvili Called Shevardnadze for Resignation’, Civil Georgia, 14 February 2003, available at: https://civil.ge/archives/102883, accessed 3 March 2018.

21 Interview with respondent at the Ministry of Internal Affairs, Tbilisi, 3 August 2010.

22 Interview with Alibala Askerov, Geyrat leader, Marneuli, 3 November 2015.

23 Their rapprochement could not be taken for granted. Saakashvili came to power through a revolution, whereas Ilham Aliyev succeeded his father, Heydar, in an orchestrated election. Heydar Aliyev had been a steadfast supporter of Eduard Shevardnadze. Thus, the 2004 state visits were essential to putting Georgian–Azerbaijani relations on a solid footing (interview with Shota Utiashvili, former Head of Information-Analytical Department at the Ministry of Internal Affairs, Tbilisi, 27 October 2015).

24 ‘Russian Active Measures in Georgia’, Wikileaks, 20 July 2007, 07TBILISI1732, available at: https://wikileaks.org/plusd/cables/07TBILISI1732_a.html, accessed 3 March 2018.

25 Sakit Allahverdiyev, the sponsor of Hümmət, headed both the All-Russian Azerbaijani Congress and a Volgograd business that build sonars for the Russian navy. He had been accused of spreading leaflets reading: ‘Hey Tatars! These are our lands. We warned you in 1991. Georgia is for Georgians!’ (ICG 2006, p. 20).

26 ‘Leader of Azerbaijani Diaspora in Georgia Arrested for Two Months: 1 December’, Central Asia-Caucasus Analyst News Digest, 9, 24, 12 December 2007, available at: http://www.cacianalyst.org/resources/pdf/issues/20071212Analyst.pdf, accessed 3 March 2018.

27 Because of the ban on regional parties, adopted back in 1990 and still in force, local activists often strike up tactical alliances with nationwide parties during election campaigns (Zollinger & Bochsler 2012, p. 620).

28 ‘Fakty diskriminatsii azerbaidzhantsev Gruzii uchastilis’: NAAG’, Regnum, 19 January 2009, available at: https://regnum.ru/news/1112073.html, accessed 3 March 2018; ‘Ethnic Azeris Demand Change in Georgia’s Political System’, Interfax, 28 January 2009; ‘Gruziya dolzhna byt’ konfederatsiei: NAAG’, Regnum, 9 February 2009, available at: http://regnum.ru/news/1122283.html, accessed 3 March 2018.

29 ‘Head of National Assembly of Azerbaijanis of Georgia Arrested in Baku’, Today.az, 11 March 2009, available at: http://www.today.az/print/news/society/51003.html, accessed 12 September 2019.

30 ‘Founding Meeting of Georgian Azerbaijanis’ Congress to Take Place on 17 March—Co-founder’, Trend News Agency, 15 March 2008, available at: https://en.trend.az/azerbaijan/society/1157388.html, accessed 3 March 2018; ‘Ali Babaev: Situation with Azerbaijanis in Georgia is Grave’, Caucasian Knot, 18 April 2008, available at: http://www.eng.kavkaz-uzel.eu/articles/7436/, accessed 3 March 2018.

31 ‘Georgia’s Azeri Minority Treated As “Second-Class Citizens”’, Radio Free Europe/Radio Liberty, 4 May 2009, available at: https://www.rferl.org/a/Georgias_Azeri_Minority_Treated_As_SecondClass_Citizens/1621243.html, accessed 3 March 2018; ‘CAG Plans to Convene Extraordinary Congress’, Trend News Agency, 16 October 2010, available at: https://en.trend.az/azerbaijan/society/1767468.html, accessed 3 March 2018; interview with a high-ranking Azeri official from the region, Marneuli, 2 November 2015.

32 Interview with Zaur Khalilov, Civic Integration Foundation (samokalako integratsiis pondi), Tbilisi, 28 October 2015.

33 Interview with Nargiz Aliyeva, Azerbaijanis’ Cultural Centre (Azərbaycanlıların Mədəniyyət Mərkəzi), Marneuli, 29 October 2015.

34 Interview with Leila Suleimanova, Union of Azerbaijani Women of Georgia (Gürcüstan Azərbaycanlı Qadınları Birliyi), Marneuli, 29 October 2015.

35 Interview with Azer Suleimanov, Member of Parliament for Marneuli, Marneuli, 3 November 2015.

36 Interview with Mansur Iuzbashov, former representative at the Public Defender’s Office in Kvemo Kartli, Marneuli, 3 November 2015.

37 Interview with Emin Akhmedov, Marneuli Youth Centre (Marneuli Gənclər Mərkəzi), 29 October 2015; Ruslan Hajiev, Director of Marneuli’s Educational Resource Centre, 3 November 2015.

38 Personal communication with the National Assessment and Examinations Centre, 6 September 2010 and 23 March 2015.

39 The speakers sounded similar enough to be confused for one another in the pilot round. ‘Maisuradze’ and ‘Mamedova’ are common, characteristic surnames in Georgia and Azerbaijan respectively. ‘Georgian Names’, JumpStart Georgia, 12 February 2013, available at: https://www.jumpstart.ge/en/our-creations/visualizations/georgian-names, accessed 3 March 2018; ‘Kakie samye rasprostranennye familii v Azerbajdzhane?’, 1News, 12 January 2012, available at: http://1news.az/news/kakie-samye-rasprostranennye-familii-v-azerbaydzhane, accessed 3 March 2018.

40 Azeri and Russian were spoken at home by 2%, another 2% listed Azeri and Georgian; 1% spoke Azeri and Russian with friends, and 9% used Azeri and Georgian.

41 Still, ethnic conflict could return. If Georgian-speaking Azeris graduate from universities but fail to gain positions of influence, then locals might conclude that discrimination persists, thus creating fertile ground for ‘natives’ seeking to reject their ‘alien’ rulers. Though not studied in this essay, much depends on the policies of the Georgian Dream (k’art’uli oc’neba) coalition that has been in government since 2013.