The Nagorno-Karabakh Conflict and the Exercise of “Self-Defense” to Recover Occupied Land

(Editor’s Note: This is the latest in a series of articles on the Nagorno-Karabakh conflict.)

The heavy fighting in Nagorno-Karabakh that erupted on the morning of Sept. 27 pitted troops of Azerbaijan, supported by Turkey, against the forces of the self-proclaimed “Republic of Artsakh” and Armenian forces. The conflict, which broke out along the line of contact established in the aftermath of the 1988-1994 war over the region, has included the deployment of drones and heavy artillery, with substantial casualties on both sides. Third States (Russia, France, and the United States) have brokered a number of ceasefires, including, most recently, yesterday’s agreement between the governments of Armenia, Azerbaijan, and Russia, which aims to put a lasting end to more than six weeks of hostilities.

In spite of the conflict’s intensity and inter-State dimension, very few States have commented on the compatibility of the protagonists’ conduct with the international law on the use of force (the jus ad bellum). Even the legal blogosphere has remained silent on the matter.

This silence may be due partly to the difficulty in ascertaining the facts on the ground – with both Armenia and Azerbaijan accusing each other of triggering hostilities. Still, the events raise a fundamental question of jus ad bellum – and one that is surprisingly overlooked in legal doctrine.

The question is: when part of a State’s territory is occupied by another State for a prolonged duration, can the former still invoke the right of self-defense to justify military operations aimed at recovering its land? Put differently: can unlawful occupation be regarded as a “continuing” armed attack permitting the recourse to self-defense at any given point in time — possibly years after the occupation commenced?

The relevance of the question is clear: ever since the 1994 ceasefire agreement, the position widely held by the international community is that the Nagorno-Karabakh region is occupied by Armenia, a status quo which has remained unaltered for the last decades. Although the region is populated mostly by persons of Armenian ethnicity, various United Nations Security Council and General Assembly resolutions have called upon Armenia to end the occupation and withdraw its forces. The “Republic of Artsakh” has not been officially recognized by any State, and is instead regarded as a puppet regime under the control of Armenia (as confirmed by the European Court of Human Rights).

If we assume that the region indeed belongs to Azerbaijan and is unlawfully occupied by Armenia, can Azerbaijan claim self-defense to lawfully recover it, even though the current territorial status quo in the region has existed for a quarter of a century? Some statements – including by Turkey and Pakistan – would seem to respond in the affirmative. A similar position was voiced by Olivier Corten, a jus ad bellum expert who is sometimes placed in the ‘restrictivist’ camp.

We doubt the ayes have it. Here’s why.

Exercising Self-Defense to Recover Occupied Land?

As a preliminary point, some might argue that a situation of occupation inevitably entails the continuation of an international armed conflict (IAC) between the States concerned (as per Common Article 2 of the Geneva Conventions), and that this renders moot the application inter se of the jus ad bellum. Such a position (expressed in particular by Dinstein (here at 61-64)) would hold that an Azeri offensive against Armenian forces might perhaps breach the ceasefire agreement, but not the prohibition on the use of force proper. However, the view that, as long as there is no formal peace agreement, either side to the conflict may resume hostilities at any point in time without the need for a justification under the jus ad bellum, appears to be a minority position that finds no support in State practice. Suffice it to note that neither Armenia nor Azerbaijan have used this argument: Azerbaijan explicitly framed its actions by referring to the right of self-defense.

Turning to self-defense, the argument in support of a continued right of self-defense as long as unlawful occupation persists essentially revolves around the idea that occupation amounts to a “continuous armed attack.” After all, Article 3(a) of the U.N. General Assembly Definition of Aggression lists as one example of an act of aggression the “military occupation … resulting from [an] invasion or attack, or any annexation by the use of force of the territory of another State or part thereof.”

To this can be added the teleological argument: inasmuch as the prohibition on the use of force seeks to protect States’ territorial integrity, isn’t this purpose defeated if we allow aggressor States to “get away” with invasion and occupation by providing only a limited timeframe within which the “victim State” must act before forfeiting its right to respond in self-defense in perpetuity? Doesn’t this encourage aggression and surrender to the “realist” position that the strong do as they can and the weak suffer as they must? And how can this be reconciled with the principle of ex iniuria jus non oritur (legal rights cannot arise from unlawful acts)?


The authors do not pretend that the above arguments are without merit — far from it. Yet, the counter-arguments would, in our view, appear more compelling.

Let’s start with the idea of a continuous armed attack. Of course, when one State invades another and goes on to occupy part of its territory, that entails a breach of an international obligation (e.g., the prohibition on the use of force, or the right of self-determination) of a “continuing character” in the sense of Article 14 Articles of State Responsibility. Yet, this does not necessarily imply that there is also a continuing armed attack as per Article 51 of the U.N. Charter. The latter provision indeed refers to the scenario where an armed attack “occurs,” which suggests an event taking place at a specific point in time (read: when the occupation commences).

Furthermore, it is generally agreed that there must be a close proximity in time between the start of an attack and the response in self-defense (see here, at 99 ff.). Practice admittedly accepts that this “immediacy requirement” (part of the necessity criterion) must be construed flexibly, taking into account the context, and leaving leeway for States to exhaust peaceful means, conduct investigations or make military preparations (as in the case of the 1982 Falklands War). Nevertheless, the lapse of time cannot be extended indefinitely. Otherwise, the ratione temporis dimension of self-defense included in article 51 would be rendered meaningless in this context.

Let’s skip the uncertain relevance of the U.N. General Assembly Definition of Aggression for purposes of interpreting the notion of “armed attack,” and move instead to the teleological aspect. Indeed, the purpose of the prohibition on the use of force is not solely to protect territorial sovereignty. One might well add the protection of human life (especially in light of the U.N. Human Rights Committee’s General Comment No. 36).

More importantly for present purposes, the prohibition also serves to protect international peace and stability, particularly in a world that is regrettably rife with territorial disputes and (semi-)frozen conflicts. Crucially then, the Friendly Relations Declaration confirms that “[e]very State has the duty to refrain from the threat or use of force… as a means of solving … territorial disputes,” and that States must likewise refrain from the threat or use of force “to violate international lines of demarcation, such as armistice lines.”

The fundamental principle that States cannot settle territorial disputes through military means – a corollary of the broader principle enshrined in Article 2(3) of the U.N. Charter – has been repeatedly confirmed, including by the Security Council (e.g., here). Of course, the principle makes sense only if it operates irrespective of whether a State holds a valid title over land or not (if not, it would simply restate the prohibition of aggression).

This was precisely the position of the Ethiopia-Eritrea Claims Commission in its Partial Award on the Jus ad Bellum. In particular, the claims commission could not accept that Eritrea’s recourse to force would have been lawful on the basis that it held a valid claim over the land it sought to recover. Rather, State practice and legal doctrine were deemed to confirm that “self-defense cannot be invoked to settle territorial disputes.” This was only logical and desirable, since “border disputes between States are so frequent that any exception to the prohibition of the threat or use of force for territory that is allegedly occupied unlawfully would create a large and dangerous hole in a fundamental rule of international law” (para. 10).

We broadly agree with that conclusion. And while some will disagree, we find ourselves in fine company. In 1984, the renowned Columbia University Law School Professor Oscar Schachter noted how an exception for recovering “illegally occupied” territory would render Article 2(4) of the U.N. Charter meaningless in many cases (here, at 1627-8). In turn, Quincy Wright of the University of Chicago found that “[a] State that neglects to defend its frontiers against hostile encroachments soon loses its right to do so, and can rely only on negotiation or action by the United Nations to restore its rightful possession.”

A more recent and elaborate confirmation of this position is found in the 2018 report of the British Institute of International and Comparative Law on “The Use of Force in relation to Sovereignty Disputes over Land Territory.” According to the report’s authors: “if State A attacks neighboring State B and occupies part of State B’s territory and State B does not respond within a reasonable period of time, it loses the ability to rely on self-defense to retake its territory by force at a later stage. … Thus, unless the attacked State [responds in] self-defense within a reasonable time, it may be confronted with the newly established territorial status quo which it will not be legally permitted to alter retroactively” (at 79-80). For the sake of clarity, the foregoing does not alter the fact that further military action by the occupying state, for instance with the intention of expanding the portion of occupied territory, would constitute a new armed attack and a modification of the territorial status quo, thus allowing the attacked state to react in self-defense.

What about State practice? Admittedly, when Egypt and Syria triggered the Yom Kippur War in 1973 with the aim of recovering the land lost to Israel in the 1967 Six-Day War, many non-aligned and communist States were sympathetic to its actions. While most reactions were not formulated in clear legal terms, at least some States expressly approved the legality of military action to liberate occupied territory.

Overall, however, in those few cases where States invoked an alleged title to territory to justify military action against (territory controlled by) other States (e.g., India in respect of Goa (1961), Argentina in respect of the Falklands/Malvinas (1982) and Iraq in respect of Kuwait (1990)), these arguments were not embraced by other States or by the U.N. organs. For instance, while many States supported Argentina’s territorial claims in 1982, they nonetheless condemned its attempt to recover the Falklands through force.

Choosing Peace Over Justice? A Balancing Exercise

Some will argue that the prohibition against settling territorial disputes through force does not “protect” every established territorial status quo, but applies only when there is a “genuine” territorial dispute (as in the case of Ethiopia-Eritrea). And so, the argument goes, it may have been irrelevant in the Yom Kippur War, since Israel had not claimed sovereignty over the Sinai and the Golan Heights.

In a similar vein, it would be irrelevant vis-à-vis Nagorno-Karabakh due to the absence of an Armenian claim of sovereignty over the area. This qualification, however, would seem both futile and potentially counterproductive. Futile, since what amounts to a “genuine” territorial dispute will in many cases of course itself be heavily disputed. Counterproductive, because an occupying State could exclude a possible reaction in self-defense by the other State simply by making a formal claim – and might well have an incentive to do so in this scenario. It hardly makes sense to suggest that Syria had a continuing right to self-defense but then lost it simply because Israel proclaimed the annexation of the Golan Heights.

Of course, defining where the right of self-defense stops and the prohibition on settling territorial disputes through military means kicks in remains an extremely difficult exercise, given the competing interests at stake. What seems to matter most is the existence of a territorial status quo, characterized by a prolonged absence of fighting, and whereby the occupying power peacefully administers the territory concerned. The latter factor may partly explain reactions to the Yom Kippur War: as some have noted, in view of all the incidents that took place between 1967 and 1973, “it can hardly be suggested that the occupied Arab territories were under the peaceful administration of Israel” (here, at 199).

One might object that Nagorno-Karabakh, too, has hardly been peaceful over the past few years. It is certainly true that there have been confrontations in recent times, including in 2016. Yet, this arguably does not alter the initial finding that a stable and relatively peaceful territorial status quo emerged after the 1994 ceasefire, prevailing for an undeniably prolonged period of time.

On a final note, one could argue that any prohibition on the use of military force to recover unlawfully occupied land must be construed in light of the political and diplomatic context. Specifically, one might be tempted to suggest that the possibility to resort to force “revives” when peaceful means of dispute settlement are exhausted. Against this, however, the obligation to settle territorial disputes by peaceful means is an absolute one, which is not made contingent on the likelihood of success of further diplomatic efforts.

An alternative theory also runs into great difficulty with regard to how to establish that peaceful means have been fully exhausted. By way of illustration, could one say 46 years after the invasion of Turkish forces in northern Cyprus and the de facto partition of the island, that the peaceful route has hit a dead end? And if so, are we willing to accept that the military option is again on the table…?

Surely these are hard questions which merit further discussion. Returning to the recent fighting over Nagorno-Karabakh, one might regret – as a scholar – that States have not taken the opportunity to express their legal views on this issue, and that the U.N. Security Council has completely remained on the sidelines during this recent crisis. That being said, the prohibition against settling territorial disputes by force remains a fundamental pillar of the international legal order, and one that argues against re-opening frozen conflicts and questioning the territorial status quo through military action (yes, also with regard to Crimea). Such conflicts may well trigger the duty of third States to not recognize an illegal state of affairs such as an unlawful occupation (see Article 41 of the Articles of State Responsibility) and may prompt those third parties to impose economic sanctions, but the dispute must necessarily be resolved through diplomatic means.

All things considered, we do not believe that the latest confrontation between Armenia and Azerbaijan lends support to the concept of unlawful occupation as a “continuing armed attack.” It is telling indeed that Azerbaijan itself did not invoke this idea, but instead argued that it was engaging in a “counter-offensive” after Armenia had first used force, allegations that were in turn denied by Armenia (note that Egypt and Syria followed the same approach in 1973). Nor did the idea that Azerbaijan was entitled to use force to end the occupation of Nagorno-Karabakh attract substantial support from third States. The United States, France, and Russia expressly “condemn[ed] in the strongest terms the recent escalation of violence along the Line of Contact.”

While the latest agreed ceasefire between the parties alters the previous status quo, and stipulates that Azerbaijan will retain the territories of Nagorno-Karabakh recovered during the course of recent hostilities, this modification is founded solely on the agreement reached by the parties and does not constitute a precedent in support of the legality of self-defense in response to unlawful occupation as a “continuing armed attack.”

Finally, we acknowledge that, at first glance, an occupied state may feel unfairly disadvantaged if it has limited time to react to occupation. Yet, the protection of territorial integrity cannot be pursued at all costs or operate in a vacuum, disregarding other core values enshrined in the U.N. Charter, such as the peaceful settlement of disputes and the maintenance of peace and stability between nations. Put differently, in the context of occupation, practice indicates that the objective behind the prohibition of the use of force is better accomplished by protecting the territorial status quo instead of granting an open-ended right to self-defense with no time constraint.

IMAGE: A picture taken on Oct. 16, 2020 shows a destroyed tank in the city of Jabrayil, where Azeri forces regained control during the fighting over the breakaway region of Nagorno-Karabakh. (Photo by BULENT KILIC/AFP via Getty Images)


About the Author(s)

Tom Ruys

Professor of International Law at Ghent University, Founder of the Ghent Rolin-Jaequemyns International Law Institute (GRILI) Follow him on Twitter (@TomRuys_Law).

Felipe Rodríguez Silvestre

Doctoral researcher at the Ghent Rolin-Jaequemyns International Law Institute (GRILI); former legal officer of the Ministry of Foreign Affairs of Chile in a dispute between Chile and Bolivia before the International Court of Justice. Follow him on Twitter (@frsilvestre).