Is All Self-Defense Worth Exercising Under International Law? Revisiting the Use of Force in Nagorno-Karabakh from a Human Rights Perspective

The international law of self-defense is a central branch of international law implicating fundamental matters of sovereignty. In the 20^th Century, it also became associated with a broader defense of international peace and security, although these are notoriously hard to define. The jus contra bellum has emerged as premised on a simultaneously complex yet simple idea: states must not attack others; if they do, the latter are entitled to self-defense.

In this post, we argue that the project, in its broad brush, is not only famously difficult to apply to actual cases, but also occasionally normatively problematic. We suggest that the jus contra bellum may need updating in light of the evolving role of human rights in structuring our understanding of the effects of violence in the international order. As a case study, we consider how arguments about the use of self-defense fit into the analysis of the war that opposed Azerbaijan to Armenia over Nagorno-Karabakh a year ago. This is especially relevant since the Russian-brokered ceasefire of November 10^th, 2020, has been losing its hold, with Azerbaijan now setting its sights on Armenia proper (see also here and here) and escalating border tensions giving cause for concern of renewal of hostilities.

The tendency to formulate the jus ad bellum in human rights terms has long been particularly apparent when it comes to the pursuit of hostilities, for example, in Theodor Meron’s “humanization” of the laws of war thesis, as well as contemporary human rights approaches to war (see here or here). But what if the jus ad bellum itself were understood more fundamentally in human rights terms? The idea that aggression is best analyzed as not simply entailing a violation of another state’s sovereignty but as a form of massive human rights violation directed against that state’s population has begun to gain ground (see here and here). It has received at least passing recognition in the Human Rights Committee’s General Comment 36 on the Right to Life (see here, paras. 69, 70) or dicta at the European Court of Human Rights, which has led to a renewed (albeit guarded) scholarly openness to entertain unorthodox “humanization of jus ad bellum” arguments. This is against the background of an elaboration of the human right to peace (see here and here), as a welcome antidote to ideas that have traditionally mobilized human rights to justify a more liberal resort to force, for example in the name of humanitarian intervention.

What remains remarkably neglected, however, despite these theoretical advances, is the possibility that self-defense itself might be subject to human rights limitations; in other words, that a presumptively legitimate exercise of self-defense under general international law might nonetheless encounter specific limitations in human rights (for a skeptical view which probably expresses the discipline’s consensus, see here). In some ways, this is the more interesting scenario, however: with aggression, human rights merely serve to recast the crime’s particular evil, adding further (but somewhat redundant) reasons to condemn aggression beyond those concerned with sovereignty or inter-state stability; with self-defense, human rights may act as a more obvious counterpoint to an otherwise generally plausible statist case for using violence. On this point, we suggest that the Nagorno-Karabakh conflict presents a compelling case.

In 1988, a secessionist war erupted between Azerbaijan and the majority ethnic Armenians of Nagorno-Karabakh backed by Armenia, until a ceasefire in 1994 left the Armenian side victorious. On September 27^th, 2020, Azerbaijan launched an armed attack invoking the right to self-defense to reclaim Nagorno-Karabakh and its surrounding territories. As soon as the November 10^th ceasefire came into effect to end six weeks of heavy fighting, this time leaving Azerbaijan victorious, a debate arose in the international legal blogosphere over whether Azerbaijan legitimately exercised a right of self-defense.

Assuming a best-case argument for Azerbaijan in which it holds valid title over the region of Nagorno-Karabakh (an issue on which this article takes no position), in what ways might a human rights analysis help nuance our understanding of the applicable jus ad bellum? We are well aware that the jus ad bellum is classically structured by considerations that have little to do with human rights. But, in an effort to think outside the box and understand why the jus contra bellum often appears tone-deaf and disconnected from the complex reality of armed conflict, including in the region, we offer the following analysis as an attempt to move beyond a kind of knee-jerk promotion of self-defense.

First, a human rights analysis might draw attention to the very real human costs of self-defense, even outside international humanitarian law violations. Accusations of war crimes on both sides in the Nagorno-Karabakh conflict have been rife, but even beyond such violations, the overall harm to rights – whether as lawful collateral damage or more general structural harm and disruption of civilian life – has been considerable. There is an argument that even wars of self-defense can have a huge impact on various rights (to life, to be free from torture and/or arbitrary detention, etc.) in ways that should impact our assessment of their presumptive legality. That harm typically falls through the cracks of the jus in bello and jus ad bellum in ways that are increasingly problematic and have reactivated recent interest in a “human right to peace.”

In this context, even states that have a presumptive right of self-defense to reclaim territory are required, according to General Comment 36 (para. 70), to “take all reasonable measures to settle their international disputes by peaceful means” or else “fall short of complying with their positive obligation to ensure the right to life.” Although there is a debate as to whether the use of force in Nagorno-Karabakh was a step towards settling a dispute or a more conventional exercise of self-defense, that may ultimately not be decisive if one considers that there is an obligation in very limited cases to even refrain from self-defense that seems unjustifiable from a human rights perspective. This human rights turn in recourse to dispute settlement, then, arguably raises the threshold Azerbaijan has to clear to invoke self-defense, above and beyond its traditional bare international law justification. Although Azerbaijan might also invoke General Comment 36’s obligation for states to protect the life of nationals/co-ethnics from violations at the hands of third parties (language that seems, in turn, to provide an added human rights justification for self-defense), it did not use that kind of argument and it is unclear on what factual basis it could have or what the particular urgency was in 2020.

Second, human rights might help evaluate the legality of Azerbaijan’s use of force by fleshing out the notoriously underdeveloped principle of proportionality in the jus ad bellum. The dominant understanding of that requirement is that resorting to force should be proportional to the attack that is being repelled or the goal sought (sometimes described as “necessity”: free Kuwait but do not invade Iraq) and also that the amount of force used be proportional to that goal (what Kretzmer refers to as “narrow proportionality”: invade Panama or attack Lebanon but do not destroy their infrastructure more than necessary).

Most discussions focus on distinguishing proportionality in the jus ad bellum and jus in bello. But what if proportionality were understood from a human rights perspective as further specifying the kind of costs that can be inflicted on populations for the sake of self-defense? After all, international human rights law has its own notion of proportionality, and if we take seriously the idea that rights should not be limited except when necessary and proportional, then surely jus ad bellum proportionality could be read in a broad human rights lens. This would be distinct from, although related to, the question of whether individual uses of force are violations of the right to life and pertain to the macro question of whether self-defense itself is legal under international law.

The significant anticipated human cost of the war in a context where non-violent alternatives exist might thus make use of force appear all the less necessary. Imagine that, in order to re-conquer a small parcel of territory, the self-defending state anticipates that thousands of nationals of the opposing state will be killed, including a high proportion of civilians and including in ways that are not unlawful under international humanitarian law (e.g. collaterally). Should the right of self-defense, clear as its existence may be otherwise in such a scenario, readily be exercised? Clearly, re-establishing one’s sovereignty on a given piece of territory at any cost cannot – on the basis of foundational human rights principles – be the be-all and end-all of international law. It is also clear that some of the force deployed in Nagorno-Karabakh went beyond the need to reassert Azerbaijani sovereignty, in ways that may be disproportionate in the second, means-oriented sense.

Third, a human rights approach to the jus ad bellum might help us assess one element that is quite specific to the Nagorno-Karabakh conflict, which is the considerable passage of time between the original “attack” and the claimed exercise of self-defense by Azerbaijan. This issue gave rise to a spirited debate in the aftermath of the conflict a year ago. Akande and Tzanakopoulos (see also Heller) argued that an occupation that is a direct consequence of an armed attack by another state is a “continuing armed attack”, and that the attacked state therefore never loses its right to self-defense, regardless of how much time passes. According to Ruys and Silvestre, by contrast, the right of self-defense stops where there exists a territorial status quo, characterized by a prolonged absence of fighting and peaceful administration of the territory concerned, as in the case of Nagorno-Karabakh. For Knoll-Tudor and Mueller, “[a]ny other result would challenge the overall architecture of peace preservation”.

What is striking is how little consideration of human rights is involved in this otherwise sophisticated discussion where human rights might shed new light on the problematic character of last year’s resumption of hostilities. Compared to an exercise of self-defense as an immediate response to an attack (the scenario imagined in the classic Caroline case), the exercise of self-defense after 30 years of relative peace without any novel armed provocation is likely to appear relatively more “aggressive,” especially when it occurs despite an obligation to negotiate peace superimposed in the meantime. The decision to re-capture Nagorno-Karabakh by force created a huge, renewed disruption in terms of rights for both Azerbaijanis and Armenians affected by it. It suggests that populations who benefit objectively from a situation of absence of war in fact live under a permanent Damocles’ sword because the right to self-defense can never be suspended.

This suggests a chronic uncertainty about the very possibility of rights that seems deeply antithetical to an international culture of rights. It might be argued that the problem lies with the state that used force against and occupied another state’s territory in the first place and thereby exposed its own population to the self-defending riposte of the occupied state. In this vein, Armenia and Nagorno-Karabakh might conceivably be blamed for having failed to promptly return the territories surrounding Nagorno-Karabakh, thus putting their populations in a vulnerable situation decades later. It does not seem, however, that any possible agency or responsibilities of Armenia or Nagorno-Karabakh in this respect negate those of Azerbaijan nearly 30 years later. That is, if one adopts a perspective that seeks to maximize human rights and the interests of populations rather than picking winners in a zero-sum game between sovereigns.

From an international human rights law point of view, it might be argued that the attacking state has no “jurisdiction” over individuals in the territory it seeks to reclaim. This seems excessively formalistic, however. Force will severely and directly affect the rights of those impacted the minute it is exercised, a fact which alone should be understood as creating human rights responsibilities. General Comment 36 (para. 63) indeed specifies that a state’s obligation to respect and ensure the right to life “includes persons located outside any territory effectively controlled by the State, whose right to life is nonetheless impacted by its military or other activities in a direct and reasonably foreseeable manner.”

Finally, there is a concern that a human rights inspired jus ad bellum, imposing further limits to self-defense as it does, could blunt the jus contra bellum and its attempt to impose a strict prohibition on the use of force (and its attendant permission of self-defense as a necessary corrective). There may be a moral hazard in telling attacked states that they will lose their right of self-defense if they let so much time pass as to alter the human rights equation in their disfavor. Yet that is in a sense precisely the point if a far-subsequent exercise of self-defense is hard, especially for those directly affected by it, to distinguish from an “out of the blue” aggression.

At any rate, failing to admonish Azerbaijan for using force in Nagorno-Karabakh last year, including by failing to uncover the very real human rights costs of the war (except through the narrow lens of war crimes), is in turn creating its own very real moral hazard: the apparent risk that Azerbaijan considers that it has an open license to resort to further force – for instance to claim the remainder of Nagorno-Karabakh or invade Armenian sovereign territory – at whatever human cost.

These risks point to rather than away from the need for a more fundamental restructuring of the jus ad bellum from a rights perspective. We have sought to bring more nuance to this debate than the invocation of self-defense typically does, in ways we think can lead to a renewed appreciation of the jus contra bellum. Although we would not claim that this human rights approach to self-defense is currently backed by state practice, we hope that it at least captures the centrality of peace to the human rights project.