Groningen Journal of International Law

International Law Under Construction

The ECtHR on Nagorno-Karabakh: Current Approaches and Future Prospects

Leave a comment

Gayathree Devi KT

In recent years, the European Court of Human Rights (ECtHR) has seen an influx in claims concerning human rights violations in contested territories. Nagorno-Karabakh is one such interesting territory, because it involves competing territorial claims not only from States – Azerbaijan (the internationally recognized territorial State) and Armenia (the occupying force) – but also from a non-State actor, the Republic of Artsakh. Although Artsakh lacks international recognition, it has been exercising de facto control over Nagorno-Karabakh to the exclusion of Azerbaijan since at least 1991. Its role in the human rights situation in Nagorno-Karabakh matters, because several human rights violations in the region are being committed by this de facto regime (DFR), whether with or without Armenia’s support. Against this backdrop, this blogpost analyses how the ECtHR has been establishing jurisdiction and responsibility for claims arising out of the Nagorno Karabakh conflict so far. It also considers the implications of the court’s 2021 decision in Georgia v Russia (II) for claims arising out of the recent resumption of hostilities in Nagorno-Karabakh.

Before 2020

The ECtHR’s most celebrated interventions in the Nagorno-Karabakh conflict came through Chiragov v Armenia and Sargsyan v Azerbaijan, which concerned the right of displaced persons to access their property in Nagorno-Karabakh. For a summary of the facts of these two cases, read this piece. This section explores some contradictions and inconsistencies in these two decisions.


In Chiragov, the ECtHR found Armenia’s jurisdiction over Nagorno-Karabakh on the basis that the military and financial support Armenia provided to Artsakh suggested that the two entities were ‘highly integrated’ (see Al-Skeini v UK on ‘jurisdiction’ under the ECHR). Even if ‘high integration’ is enough to establish jurisdiction, it is not sufficient to establish Armenia’s responsibility under international law – a distinction that the court fails to draw. As Marko Milanovic explains here, jurisdiction by itself does not lead to responsibility; instead, responsibility turns on whether the violation-establishing conduct (here, the DFR denying displaced persons the right to return) is attributable to the State.

Under general international law, the conduct of a DFR is attributable to a State if: (i) the DFR is completely dependent on the State in all fields of activity (‘complete dependence’); or (ii) the State in question exercised effective control over the specific wrongful conduct of the DFR (‘effective control’) (see Articles 4 and 8 of the International Law Commission’s Articles on State Responsibility, Military and Paramilitary Activities in Nicaragua and Bosnian Genocide Case). Although the ECtHR could not have established on the facts of Chiragov that either of these tests were met, it found Armenia responsible. One can only assume that, in doing so, the Court either: (i) created a special rule of attribution different from the rules under general international law; or (ii) held Armenia responsible for violating a positive obligation to prevent Artsakh authorities from violating rights, despite having the power to do so by virtue of the ‘high integration’. However, this is not clear from the language of the judgment (see Milanović and Papić), and therefore it appears that the ECtHR conflated jurisdiction and responsibility.

In any event, the practical desirability of the Court’s approach is questionable. The measures that Armenia was expected to implement in order to correct the applicant’s situation were not straightforward because they required action from Artsakh authorities too. The direct involvement of Armenian forces in the denial of the applicant’s property claims in Chiragov was never explicit; therefore, Armenia’s insistence, at the stage of judgment execution, that the ‘immediate addressees of the case are [the Artsakh] authorities’ is unsurprising. On the other hand, in Zalyan v Armenia and Muradyan v Armenia, the ECtHR’s finding of Armenia’s responsibility was not contested because the violations were committed by Armenian forces in Nagorno-Karabakh. Armenia was also forthcoming in executing these judgments (see here and here). Therefore, it is practically desirable for the ECtHR to follow rules of general international law on ‘attribution’.

Finally, the rigidity in the ECtHR’s approach towards finding Armenia’s jurisdiction needs addressing. For instance, although Zalyan concerned the wrongful conduct of Armenian authorities in Nagorno-Karabakh, the ECtHR strangely relied on its finding in Chiragov that Armenia andArtsakh authorities were ‘highly integrated’ to find Armenia’s jurisdiction over Nagorno-Karabakh. This approach of readily referring back to the pilot case involving a DFR to find jurisdiction in every other case concerning the DFR, without taking contemporaneous and underlying facts into account, is concerning.


The ECtHR’s holding in Al-Skeini suggests that States must have factual/ actual authority over territory or persons to have jurisdiction. However, in Sargsyan, although Azerbaijan did not exercise such factual authority over the village of Gulistan, the ECtHR found that Azerbaijan had jurisdiction on the ground that Gulistan fell within Azerbaijan’s internationally-recognised territory, and therefore, it had to secure human rights in that territory to ‘avoid a vacuum in Convention protection’.

While this logic is sound in the context of territories with undisputed internationally-recognised boundaries, it fails when the boundary is disputed, as in Sargsyan. As Judge Yudkivska indicates in her concurring opinion, Azerbaijan’s lack of effective control over Gulistan had everything to do with the exercise of factual authority by Artsakh and Armenian forces around Gulistan. Yet, the ECtHR failed to take this into account, holding that Azerbaijan had jurisdiction and full responsibility for the violation, purely based on its title to the territory.

Sargsyan’s finding on State responsibility also raises questions. Even if Azerbaijan is responsible, it is unclear why the ECtHR ignored the potential contribution of Armenia and Artsakh to the harm suffered by the applicants. As Judge Yudkivska points out, ‘the mere fact that the applicant…decided to lodge a complaint against only one [State] involved in the conflict and not both…should not automatically engage the full responsibility of Azerbaijan.’ Although the ECtHR cannot determine the responsibility of a State that is not a party to the case, it can, at the least, consider the possibility of shared responsibility (as it did in Matthews v the United Kingdom). As the ICJ observed in Nauru v Australia, determination of the responsibility of one State among several that contributed to the wrong does not amount to an adjudication on the legal position of other States. Holding Azerbaijan responsible only for its contribution to the applicant’s injury would not have automatically created a ‘vacuum’ or prevented the Court from holding Armenia responsible in a subsequent case.

2020: Resumption of Hostilities

After hostilities in Nagorno-Karabakh resumed in September 2020, the ECtHR called upon all parties to refrain from breaching the ECHR. However, this had no effect on the situation on the ground, and the hostilities ended only in November 2020, with Russia brokering a ceasefire agreement. Under this agreement, Azerbaijan regained control over parts of Nagorno-Karabakh. While the Republic of Artsakh continues to exist de facto, its territorial control has reduced substantially. Additionally, under the new arrangement, the Lachin corridor which forms the main link between Armenia and Artsakh falls under the control of Russian peacekeepers, with the rest of the Lachin district now being under Azerbaijan’s control. In light of this new factual situation, the ECtHR cannot sustain the ‘pilot judgment’ approach. In determining any claims that arise in connection with the 2020 hostilities and their aftermath, the Court will necessarily have to reassess the extent of control that the three parties to the conflict had during the hostilities and immediately thereafter.

The ECtHR’s recent decision in Georgia v Russia (II) could also be instructive in this assessment. According to this judgment, during the ‘active’ phase of an armed conflict, ‘the very reality of armed confrontation and fighting between enemy military forces seeking to establish control over an area in a context of chaos means that there is no control over an area.’ Strangely, the court applied this reasoning only to claims of violations of Article 2 of the ECHR (right to life), and not to Article 5 (right to liberty and security) claims. As Milanovic points out in this detailed break-down of the judgment, the court’s categorical characterisation of the active phase of all international armed conflicts as being ill-suited for a determination of control over territory (and in turn, spatial jurisdiction) is problematic. It betrays an incorrect assumption that it is not possible for a State to establish some level of stable control over a territory even as active hostilities continue in surrounding areas. The ECtHR also categorically rejected the possibility of the parties to the conflict exercising personal jurisdiction on the basis that such jurisdiction arises only in ‘isolated and specific acts involving an element of proximity’. This holding is, at best, a half-baked attempt at using legal reasoning to avoid deciding on a politically controversial point. Nonetheless, it is bound to influence how the ECtHR determines any claims arising out of the ‘active’ phase of hostilities in Nagorno-Karabakh in 2020.

Curiously, even in Georgia v Russia (II), the Court held Russia responsible for the wrongful conduct of South Ossetian and Abkhazian authorities (the DFRs), by conflating jurisdiction and responsibility. However, the Court, in passing, refers to the fact that ‘Russian troops were passively present during scenes of looting’. This may help establish that the State had knowledge of the wrongful conduct of the non-State entity and aided and assisted it nonetheless – a factor relevant under the test of ‘complicity’ adopted by the ICJ in the Bosnian Genocide Case to hold that Bosnia had a positive obligation to prevent genocide. It is unclear whether the ECtHR is adopting the same test, but should it, it will help harmonize its position on State responsibility with the position under general international law.


The ECtHR’s unchanging conflation of jurisdiction and responsibility is a source of concern for its future adjudication of violations arising out of protracted conflicts like the one in Nagorno-Karabakh. It risks fragmenting international law and creating conflicting obligations for States, which in turn could jeopardize State compliance with ECtHR judgments. Further, the court’s new refusal to deal with Article 2 violations during active hostilities is distressing. It creates a regrettable ‘vacuum in Convention protections’ for the people killed or maimed in the recent hostilities in Nagorno-Karabakh – a vacuum the Court is usually careful to avoid. One can only hope that the ECtHR will correct its course in the future.

Gayathree Devi KT is a DPhil student in public international law at the University of Oxford, writing her thesis on international responsibility for human rights abuses by de facto regimes. She co-convenes the Oxford Transitional Justice Research Group, and is also a Tutor in public international law at Hertford College. Before her DPhil, Gayathree completed the MPhil in Law and the BCL from Oxford, and also worked as a Research Assistant at the UN International Law Commission. She tweets @gayathreekt.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.