The Minsk Group, led by the United States, France, and Russia, has brokered the Armenian-Azerbaijani conflict since the mid-1990s after Armenia-backed secessionists in the unrecognized Nagorno-Karabakh Republic won the first Karabakh war of 1992–94. That mediation embodied the ideals of the mid-1990s unipolar moment, which assumed that liberalized markets and democratic transitions would converge internally to resolve legacy conflicts in postsocialist states while bringing them into convergence externally with Euro-Atlantic nations. Those assumptions withered away over the next quarter-century. Neither Azerbaijan nor Armenia transitioned to liberal democracy. Backed by an increasingly assertive Turkey, Azerbaijan prevailed in a bloody war in 2020. This time, the regional authoritarian powers, Russia and Turkey, are overseeing what could be a test case for a new form of “illiberal peace.”

“Adrift and unstable, the conflict was ripe for poaching out of its multilateral stupor by entrepreneurs of a new world order based on multipolarity.”

In the wave of self-determination conflicts that became a hallmark of the Soviet dissolution in the Caucasus, the Armenian–Azerbaijani conflict over Nagorny Karabakh was always an outlier. It certainly shared formative attributes with subsequent conflicts in Abkhazia, South Ossetia, and Chechnya: contested Soviet border delimitations, nationalist movements sweeping decrepit communist power structures aside, and the weak, fractious statehood of the republics that succeeded Soviet rule in December 1991. The secessionist victory in the first Karabakh war of 1992–94 was also consistent with the wider post-Soviet trend. With Armenia’s support, the Armenians of Nagorny Karabakh broke away from Azerbaijan and established their own, unrecognized Nagorno-Karabakh Republic (NKR).

(Variant names and spellings reflect the territory’s layered belongings and the competing geopolitical traditions claiming it. Nagorny Karabakh, with variant spellings Nagorno and Karabagh, is a name drawing on Russian, Turkish, and Persian, meaning “mountainous black garden.” Armenian sources prefer Artsakh, a pre-Turkic Armenian name for the area in antiquity. Azerbaijani sources increasingly prefer just “Karabakh,” rejecting the Soviet-era delimitation of a separate Armenian-majority highland space.)

Yet in several of its dimensions, the Armenian–Azerbaijani conflict was not an archetypal Eurasian “frozen conflict” (if, indeed, such a thing exists). In its temporal dimensions, the conflict involved factors beyond the immediate dynamics of the Soviet collapse. Beginning in February 1988, it was the early riser in the tide of nationalist mobilization that brought down the Soviet Union. Its early appearance reflected a unique history of reciprocal communal violence and mass displacements, going back to the early twentieth century.

In 1918, short-lived independent Armenian and Azerbaijani republics emerged from the Russian imperial collapse and contested their intermingled demographics through practices we would now call ethnic cleansing. The disputed borders were eventually pacified by the Soviet state, yet processes of ethnic homogenization continued. Enabled by the Soviet mania for promoting the indigenousness of “titular” nations (the ethnic majorities for which various Soviet republics were named) as an ideological gloss for the USSR’s ostensibly non-imperial identity, cultural intelligentsias crafted discourses “othering” minorities as the illegitimate residue of historically recent migration.

In the late 1980s, when Mikhail Gorbachev introduced his perestroika policy to reform the USSR, the Armenians of Nagorny Karabakh saw an opportunity to raise grievances related to their status in Azerbaijan. They declared their desire to self-determine by unifying with Armenia. Communal violence rapidly ensued. For still-intermingled populations—Armenians in Azerbaijan and Azerbaijanis in Armenia were each Soviet republic’s largest minority—the impacts were devastating. More than half a million became refugees within two years, in a reciprocal demographic purge on a scale unseen elsewhere in the former Soviet Union.

In their spatial dimensions, the conflict’s outcomes far exceeded the self-determination script of post-Soviet secessionists. The Armenian version of this script had originally focused on the Nagorno-Karabakh Autonomous Oblast (NKAO). Founded in 1923 by the young Soviet regime, this regional administrative unit ostensibly compensated its local ethnic Armenian majority with territorial autonomy in exchange for their incorporation into Soviet Azerbaijan. But the enclave geography created a dilemma. The core Armenian demand of the late 1980s—unification with Armenia—could be imagined only within a continued Soviet framework, or one necessitating a revision of Azerbaijan’s borders.

This dilemma was resolved in the course of the first Karabakh war, in 1992–94. Armenian forces occupied a belt of territories around the NKAO. These gains doubled the size of the territory under the control of the NKR, but came at the cost of a new wave of forced displacement. Over half a million ethnic Azerbaijanis were forced out of their homes, and their towns and villages were looted.

Ethnic cleansing of Azerbaijanis was initially framed as a necessity to protect the Armenians of Karabakh from the very real threat of Azerbaijani ethnic cleansing. But over time, new meanings began to congeal around these lands. They became increasingly invested with an invented nostalgia depicting them as “liberated territories,” restored to rightful Armenian ownership. The submergence of the Karabakh Armenians’ self-determination claim within this expansive territorial vision entailed a growing moral ambiguity, enabling Azerbaijan to reframe the conflict as entirely irredentist in nature.

Azerbaijan has emerged as the keystone of an illiberal regional configuration.

That the conflict was indeed irredentist, rather than secessionist, appeared to be confirmed by its specific structural dimensions: it became embedded at the interstate level between Armenia and Azerbaijan. “Parent states” enduring secession typically see geopolitics rather than local grievance as the true driver of frozen conflicts, which play out among an unrecognized republic or “de facto state” seeking secession, the patron state supporting its revolt, and the parent state. But in the context of what had started out as a unification movement among co-ethnic communities, the line between patron state and de facto state was particularly difficult to define in the Armenian case. This structure also put the Karabakh conflict at the center of the nation-building processes of two sovereign states. From the geographic periphery, Nagorny Karabakh moved to the conceptual center of Armenian and Azerbaijani national narratives, generating pithy formulations of nested—and non-negotiable—territorial belonging: “There is no Armenia without Artsakh!” “Karabakh is Azerbaijan!”

This totalizing effect combined with a fourth, strategic dimension of the conflict. Whereas in Eurasia’s other frozen conflicts, it is the status quo power—Russia—that is militarily preponderant, in this case it was the challenger, Azerbaijan, that gradually attained preponderance. In 2006, Caspian oil began to flow from the Azerbaijani capital Baku, via Georgia, to the Turkish Mediterranean port of Ceyhan. In that year, and again in 2011, Azerbaijan’s military expenditures doubled. While Armenia’s arms procurements derived almost exclusively from its bilateral ties with Russia, Azerbaijan’s oil revenues permitted the purchase of state-of-the-art equipment from Israel, Turkey, and others.

This newfound preponderance rapidly transformed the dynamics along the roughly 200-kilometer Line of Contact separating Armenian and Azerbaijani forces. What had been until 2014 a sniper’s war, claiming 20–30 lives annually, became a low-intensity conflict characterized by regular skirmishing, experimentation with new technologies, the targeting of civilians, and occasional large-scale escalations, notably a “four-day war” in April 2016 that left more than 200 dead. The conflict was pervasive in every domain of Armenian–Azerbaijani relations, re-militarized through substantial rearmament, and punctuated by regular military crises. By the late 2010s, predictions of a major new war were commonplace.

Beyond its temporal, spatial, structural, and strategic specificities, the Armenian–Azerbaijani conflict was distinctive among Eurasia’s territorial disputes for what might be called its “unipolar memory.” It uniquely preserved a mediation structure and constellation of international actors that recalled the unipolar moment of the mid-1990s.

From the outset, international mediation efforts served a dual purpose: to resolve the conflict, and to institutionalize a new regional security architecture following the end of the Cold War. The mediation mission was taken up in March 1992 by the Conference (after 1994, Organization) for Security and Co-operation in Europe (CSCE/OSCE), founded on the basis of 1975’s Helsinki Final Act.

A few weeks earlier, on January 30, 1992, the Soviet successor states had all been accepted as members of the Conference (to “loud and prolonged applause,” as then-US ambassador to the CSCE John Maresca recalls in his memoir). The British delegate observed that a conflict was ongoing between two of the new members, Armenia and Azerbaijan, and it was the CSCE’s responsibility to act. After a show of hands for volunteers to form a group to prepare a peace conference, no fewer than 11 states signed up: Belarus, Finland, France, Germany, Italy, the Netherlands, Portugal, Russia, Sweden, Turkey, and the United States. The Belarusian representative offered his country’s capital as a venue, and the body that would come to be known as the “Minsk Group” was established.

The Minsk Group eventually stabilized in a slightly smaller configuration, minus the Netherlands and Portugal. From 1997, it was led by a permanent troika of France, Russia, and the United States. This structure emphatically embodied the ideals and assumptions of the mid-1990s unipolar moment. It inserted the ostensible victor of the Cold War directly into the mediation of a conflict that, in Washington’s corridors of power, was both distant and barely known. The Group also embodied the predominant post–Cold War approach to Russia among Euro-Atlantic powers, both recognizing and seeking to contain the realities of its presence on the ground by bringing it—however unwillingly—into a multilateral framework. The presence of France, meanwhile, affirmed what in the 1990s appeared to be an unassailable certainty: the European destiny of the Caucasus.

OSCE mediation embodied the values and assumptions of the unipolar moment in two other crucial ways. First, the OSCE took decisions by consensus, granting voice to small powers and assuming that responsible states would work collaboratively to resolve problems according to the principles of the liberal international order. Second, those principles translated into an array of conflict resolution models based on liberal peacebuilding and democratic politics. Self-determination conflicts would be resolved through liberal alternatives to secession: territorial autonomy, self-governance, and the upholding of human and minority rights within parent states.

OSCE mediation was thus grounded in the quintessential mid-1990s belief in a “double convergence”: democratic transitions, liberalized markets, and peacebuilding would converge internally to resolve legacy conflicts in post-socialist states, and these states would converge externally with Euro-Atlantic institutions and structures. Over the following quarter-century, these assumptions steadily withered away. The visions of a democratized and regionally consolidated South Caucasus both failed to materialize.

Neither Armenia nor Azerbaijan converged internally on liberal democratic and market transitions. Between the end of the first Karabakh war in 1994 and 2018, Armenia had a “competitive authoritarian” regime with a multiparty system. Competitive electoral campaigns were neutralized by the use of administrative resources and, on several occasions, coercion to suppress protest. In 2018, Armenia belatedly experienced a “Velvet Revolution,” an unexpected nonviolent uprising that ejected the stagnating regime and installed former journalist Nikol Pashinyan as prime minister. Over the following two years, reforms vied with populist inclinations as Pashinyan confronted the lingering influence of the former regime, for example in the judiciary.

In Azerbaijan, oil revenues enabled the development of a hegemonic authoritarian regime, built on the foundation of the former Soviet Union’s first, and so far only, dynastic succession. Ilham Aliyev succeeded his father Heydar as president in 2003. Much like in Russia, the autarky afforded by oil wealth enabled Azerbaijan to practice a postmodern form of authoritarianism more reliant on public spending, co-optation, and narrative hegemony than actual coercion. A few years into its oil boom, in 2013, Azerbaijan became a consolidated authoritarian regime, according to the scale of political rights and civil liberties established by the US monitoring group Freedom House.

Neither Armenia nor Azerbaijan embraced Euro-Atlantic integration over their first quarter-century of independence. Armenia’s potential for such integration was constrained by its security linkages with Russia. As a founding member of the Russian-led Collective Security Treaty Organization, Armenia also relied on a number of bilateral agreements with Moscow to compensate for the growing military asymmetry with Azerbaijan. Efforts to diversify its relations met with limited success: Yerevan was forced by Russian pressure to renounce a quietly negotiated association agreement with the European Union in 2015, and to join Russia’s Eurasian Economic Union instead.

In Azerbaijan’s case, autarky enabled a transactional approach to integration with the West. Baku pursued a thin form of integration on the basis of limited partnership agreements with the EU, balanced by participation of different kinds in non-Western groups such as the Non-Aligned Movement, the Organization of the Islamic Conference, and the Shanghai Cooperation Organization.

The Minsk Group, meanwhile, operated as an iterative process in which one of the three co-chairs would push an initiative or proposal, backed up by the others, which would be discarded and replaced by the next co-chair's effort. The last time that the United States invested serious effort in the process was at peace talks it hosted in Key West, Florida, in April 2001. The talks failed, and six months later 9/11 transformed US foreign policy priorities in ways that sidelined the Armenian–Azerbaijani conflict for the next two decades. For a time, French President Jacques Chirac took a personal interest in the conflict, culminating in another unsuccessful summit at Rambouillet in February 2006. After that, the United States and France increasingly deferred to Russian leadership: President Dmitry Medvedev led the Minsk Group’s last significant push, ending in yet another fruitless summit in Kazan in June 2011.

Nagorny Karabakh moved to the conceptual center of national narratives.

For the next nine years, the Minsk Group circled in a diplomatic wilderness punctuated occasionally by summits, but more often by military crises along the Armenian–Azerbaijani front lines. As these crises displaced diplomacy, the Minsk Group increasingly embodied only the memory of unipolarity, rather than the anticipated convergence with liberal practices and norms of conflict resolution.

Still, this memory insulated the Armenian–Azerbaijani conflict from resurgent Russian–Western bipolarity after the Ukrainian crisis erupted in 2014. Even as Eurasia’s geopolitics fissured anew, the Minsk Group’s institutionalized memory of the Russo-Euro-Atlantic consensus on resolving a conflict of more peripheral importance sustained a simulacrum of collaboration, though of a kind unable to adapt to consequential new currents and trends. Adrift and unstable, the conflict was ripe for poaching out of its multilateral stupor by entrepreneurs of a new world order based on multipolarity.

The enclosure of conflicts in Abkhazia, South Ossetia, Transdniester, and the Donbas within the paradigm of Russian–Western rivalry expressed the European aspirations of the parent state in each case—Georgia, Moldova, and Ukraine. Like these republics, Azerbaijan had also looked westward during its early independence. On the one hand, Baku had sought a transactional partnership to supply capital and know-how for exploiting its Caspian oil fields. On the other, it sought normative affirmation of its territorial integrity. While it found the former in abundance, the latter proved elusive.

This was largely because economic autarky afforded by oil rents translated into normative autarky. Whereas the Rose, Orange, and Euromaidan revolutions presumed “civilizational” conjunctures between Georgians, Ukrainians, and the “free world,” Euro-Atlantic commitments to Azerbaijan’s territorial integrity remained formal rather than substantive—as did Azerbaijani commitments to democratic norms and human rights. Azerbaijani petrodollars financed the extensive penetration of national and multilateral institutions in Europe and the United States through strategies of influence-buying, memorably dubbed “caviar diplomacy” by Gerald Knaus of the European Stability Initiative.

Normative autarky defined Azerbaijan’s unique position among Eurasia’s parent states as a skeptic of liberal multilateralism. Frustrated by what it perceived as tolerance of Armenian transgressions in Western capitals, Azerbaijan cultivated its own brand of multilateralism as a “middle power” and a variety of special relationships with other states similarly defined, notably Turkey, Israel, and Pakistan. At home, resentment of Western hypocrisy and orientalism featured strongly in an emotional culture that foregrounded national trauma and kept the humiliation of the 1990s alive for a new generation. Combined with lavish public spending, the strategy of demobilizing dissent by keeping society mobilized around the axis of the conflict was largely effective. Protests in Azerbaijan have been few and far between in recent years, and generally easily dispersed.

But in July 2020, the Azerbaijani state was confronted by a form of protest that it could not easily quell. In the aftermath of another border clash, spontaneous crowds outraged by the killing of a popular general gathered in central Baku, calling for war with Armenia and breaking into the parliament building. Unlike the massive crowds in Belarus that had been protesting against President Alexander Lukashenko since the end of May, this crowd could not be dismissed as the product of Western funding or George Soros’s putative influence. It reflected back at the regime its own narrative about the intolerable Armenian occupation, instilled in every Azerbaijani citizen from childhood.

Yet without a tested military capability, facing an entrenched enemy in a conflict in which the actions of great powers were likely to be both unpredictable and decisive, Baku needed an ally before starting a war. Here, Azerbaijan’s agenda finally converged, not with that of liberal multilateralism, but with that of a key entrepreneur of a post-Western multipolar world—Turkey.

Turkish acquiescence to Euro-Atlantic coordination with Russia on the Karabakh conflict had long been an implicit assumption of the unipolar moment. Turkey’s NATO membership and status as a core US ally in the Middle East had determined Ankara’s cautious approach to its neighborhood in the post-Cold War period, despite Turkish elation at the prospect of a new Turkic world opening up in Eurasia after 1991. That sense of Turkic brotherhood was nowhere stronger than with Azerbaijan, embedded in a shared ethno-linguistic culture, and later in partnership in the multiple infrastructural initiatives involving pipelines, railways, and roads that link the two nations via Georgia. But that solidarity did not translate into an active policy of Turkish support for Azerbaijan in the Karabakh conflict, beyond the closing of the Turkish border with Armenia in 1993.

The change in this calculus was the most significant external factor in the onset of the war in 2020. Hardened by a failed coup attempt in 2016, Turkish President Recep Tayyip Erdoğan reoriented domestic politics toward a nativist populism and unanchored Turkey’s geopolitics from compliance with a unipolar order, positioning it as a rising regional power with aspirations of its own. Through a series of direct military interventions beginning in 2016 in Syria, leading up to a January 2020 operation in Libya that shifted the course of the conflict in that country, Turkey accumulated a wealth of experience in conflict theaters in what Ankara considers its “near abroad.”

In the aftermath of the July 2020 border clashes, Turkish–Azerbaijani contacts intensified. We now know that those discussions identified a key window for military action in the final weeks before the US presidential election on November 3. Turkish and Azerbaijani decision-makers evidently calculated that at a global conjuncture of preoccupation with the fate of American democracy, the ravages of a global pandemic, and a host of other issues from the Belarusian protests to Brexit, little more than expressions of concern would be forthcoming from the liberal multilateral world they sought to challenge.

Although many commentators had issued advance warnings, the war that actually broke out on September 27, 2020, was still in many respects unforeseen. Most external observers, including this author, expected if not another “four-day war” like that of 2016, perhaps a two- or three-week war, with Azerbaijan recapturing part of the occupied territories around Nagorny Karabakh.

What in fact transpired was a 44-day war, in which Azerbaijani forces, buttressed by numerous forms of additional Turkish capability, overwhelmed Armenian defenses in several locations, recaptured the entire southern belt of occupied territories, and advanced to the heart of Nagorny Karabakh itself to take the strategically and symbolically key town of Shusha (known as Shushi to Armenians) on November 8. At this point, Russia brokered the fourth of several cease-fires in a trilateral agreement with Armenia and Azerbaijan. The agreement further mandated the return of all other occupied territories to Azerbaijani jurisdiction, the deployment of a Russian peacekeeping mission to parts of Nagorny Karabakh that remained under Armenian control, and a new regional configuration of open borders and transit routes.

The human cost was devastating, with some 4,000 Armenians and 2,900 Azerbaijanis killed in action, 170 civilians slain, and tens of thousands displaced, mostly Armenians from the heavily bombarded towns of Nagorny Karabakh. Replicating the practices of the first Karabakh war, no territory changed hands without ethnic cleansing. The Armenian population of the southern region and town of Hadrut in Nagorny Karabakh was forced to flee to Armenia, and its community heritage was erased.

Armenian vulnerabilities were central to these outcomes. Armenia’s strategic posture was centered on a multifaceted deterrent intended to prevent war. First, the prospect of a costly highland war was widely assumed to constitute a core deterrent, combined with the perception of Azerbaijan as a “soft state” incapable of withstanding a lengthy and casualty-heavy war. But new forms of Azerbaijani military capability neutralized this deterrent. Recently purchased Turkish drones were deployed to devastating tactical effect, eliminating Armenia’s air defenses and then constricting its supply lines. Left without reinforcements, isolated Armenian frontline posts were overwhelmed with massive frontal assaults, some of them using mercenaries recruited from pro-Turkish militias in Syria to absorb casualties that would have been politically costly had Azerbaijani troops been sacrificed.

Another element of the Armenian deterrent was the prospect of punishment through countervalue strikes against non-military targets. Although Armenia does appear to have deployed Iskander-M precision ballistic missiles, their impact was minimal. Armenian forces also carried out missile strikes on the Azerbaijani cities of Ganja and Barda in an effort to deter further strikes on the towns of Nagorny Karabakh, but again the desired effect was not achieved despite a terrible cost in dozens of Azerbaijani civilian deaths.

Armenian vulnerabilities were compounded by a mythology of invincibility.

Additionally, Armenian balancing with Russia generated expectations that Moscow would step in to prevent a major war. When this did not happen, it prompted furious debate in the Caucasus and beyond over Russia’s “passivity” in the war’s early weeks. These expectations underestimated two factors.

First, Azerbaijan was no “ordinary” parent state. Affinities between the Russian and Azerbaijani elites run deep and are free of the tripwires presented by Georgian or Ukrainian aspirations to Euro-Atlantic membership. Second, even if Russia was a major stakeholder in a multilateral coalition to prevent a new war, it had never signed up for a liberal peace. Its reputation took a serious hit from the keenly observed international spectacle of an army almost exclusively armed and equipped by Russia being destroyed by one largely (if far from exclusively) trained and equipped by Turkey. Yet the war also created an opportunity for the Kremlin to exclude a liberal peace.

This required Russia to escalate its commitments, deploying a 2,000-strong peacekeeping mission to Nagorny Karabakh under uncertain and risk-laden conditions, and to tolerate the presence of a new actor, Turkey, in what Moscow takes to be its own near abroad. Yet these commitments were grounded in an authoritarian, post-Western consensus. For Moscow, this was a more attractive outcome than overt support of Armenia—only a nominal ally in a deeply asymmetric relationship, and a country experimenting with democracy to boot.

Armenian vulnerabilities were compounded by a mythology of invincibility, nurtured by the strongmen who had savored victory over a disorganized Azerbaijan in the first Karabakh war. Russian President Vladimir Putin has said he persuaded Aliyev to cease hostilities just over three weeks into the 2020 war, on October 20. Apparently refusing to believe in the possibility of defeat, Pashinyan decided to fight on, and lost more.

There are two core underlying outcomes of the 2020 Nagorny Karabakh war. First is the regionalization of the Armenian–Azerbaijani conflict. It is no longer a conflict that a multilateral coalition of states seeks to mediate and resolve, but the object of a Russian–Turkish condominium directed at managing conflict through negotiation among relevant regional powers. Russia dominates the security field and remains the guardian of the November 9, 2020, cease-fire declaration. Its security agencies will control existing and prospective transit corridors across Azerbaijani and Armenian territory. Turkey has a secondary role in the new security arrangements, with a nominal presence at a cease-fire monitoring center. Yet it has cemented its role as Azerbaijan’s patron. The Shusha Declaration of June 15, 2021, signed during a celebratory joint visit by Aliyev and Erdoğan to the recaptured city, formalized the partnership between the two states. Turkish firms will play a central part in the reconstruction of the de-occupied territories.

The second outcome is the conflict’s transformation into a test case for “illiberal peace,” or authoritarian conflict management (ACM). Multilateral mediation sought to resolve conflicts according to the principles of the liberal international order, identifying issues of rights and representation as core to disputes, and practices of participation and inclusivity as vectors to peace. ACM, by contrast, seeks to homogenize political communities, reinforce existing power hierarchies, and exclude nonstate actors.

Azerbaijan’s affirmations that the conflict is now “solved,” and its relegation to “history” of the core underlying issue—the political status of the Karabakh Armenian community—reflect this approach. These implications of an illiberal peace make it clear that only the OSCE’s Minsk Group provides a framework in which a genuinely comprehensive peace agreement addressing all issues contested by Armenians and Azerbaijanis is likely to be possible. The Minsk Group also retains the sole international mandate to negotiate an agreement.

Azerbaijan has emerged as the keystone of an illiberal regional configuration encompassing Russia, Turkey, and potentially Iran. The war allowed the Azerbaijani elite to pivot its legitimacy formula from petrodollar paternalism to strongman nationalism. This implies that the enemy image of Armenians may not recede in the Azerbaijani public sphere. Perhaps, like China’s “Century of National Humiliation” narrative, Azerbaijan’s national narrative will continue to focus on its quarter-century of humiliation as the foundation of its affective community.

For Armenia, its truncated defeat imposes the worst of all possible worlds. A residual Armenian Karabakh remains, but not one that Armenia can meaningfully influence, now that Russia has assumed all responsibilities in the territory. Yet Armenia’s domestic politics will continue to be shaped in substantial ways by the existential destiny of Armenians beyond the nation’s borders, whose fate is in other hands. Armenia is now delimited within borders it has never held as an independent state, with vulnerabilities that portend further embedding within a Russian security regime. A new era of compromised statehood beckons. Armenia will be torn between competing impulses to build a garrison state or maintain a path to what inevitably will be a fractious and imperfect democracy.

The framing of what is in conflict has also undergone a significant transformation. Although the interstate dimension has been salient over the past quarter-century, the conflict has always been multilayered. Armenia and Azerbaijan will continue to engage in strategic rivalry. Yet a fundamental outcome of the 2020 war that will become clearer over time is a shift in emphasis—from a territorial conflict focused on the regulation of a border between two states, to a conflict over the governance of majority–minority relations within a single state. The interlinking of the Armenian cause in Nagorny Karabakh with the Palestinian, Kurdish, Yezidi, Tigrayan, and other struggles in a new axis of twenty-first century resistance to coercive state-building can already be observed.

Overhanging the project of a regionalized illiberal peace is the question of how communities saturated with enemy images of each other will transact with, let alone trust, one another. With its stipulation to open borders, the November 2020 cease-fire statement constitutes a peculiar hybrid—more than a cease-fire but much less than a comprehensive peace agreement. Between the cessation of hostilities and the vision of a South Caucasus with open borders, latticed with transit routes and trade corridors, lies the vast middle ground of the incentives, motives, and fears of the communities living in the region. It is doubtful that Russia and Turkey have the will and capacity to sponsor the rehabilitation and reconstruction of vast swaths of western Azerbaijan de-occupied in 2020, as well as the kind of community-level confidence-building needed to transform Armenian–Azerbaijani relations and make regional connectivity plans viable.

These daunting, epochal agendas will make new conjunctures inevitable—between multilateral and multipolar, liberal and illiberal actors. But the unipolar moment of the 1990s will no longer be the lodestar for such conjunctures. Instead, the future of the Armenian–Azerbaijani conflict will reflect the mid-twenty-first-century dynamics of multipolarity, and interrogate the assumptions of its associated world order.