Self-Defence as an Exception to the Principle of Non-Use of Force: Debunking the Myth

Written by


Marko Milanovic recently wrote a blog for EJIL: Talk! examining whether Israel can rely on the right of self-defence to justify its military action against Hamas. Marko’s post addresses many difficult questions relating to the law of self-defence and is a must read. In this post I want to focus on one of Marko’s core claims, which potentially has significant implications for the law of self-defence and in particular its scope of application.

Marko questions the relevance of the right of self-defence under Article 51 of the UN Charter to the Israel/Hamas conflict because, unless Palestine is recognised as a State under international law (which Israel resists), the essentially internal character of the conflict means that Israel’s conduct does not engage the prohibition on the use of force under Article 2(4) of the UN Charter; and if Article 2(4) does not come into play, Article 51 ‘cannot logically apply’. But Marko goes on to explain that ‘[s]elf-defence is an exception to the prohibition on the use of force in Article 2(4) of the Charter’ (his emphasis) and concludes: ‘The point here is simply that Article 51 is not some kind of freestanding rule that can apply without Article 2(4)’. A possible implication of Marko’s argument – although not one that Marko has argued for – is that the right of self-defence under Article 51 only permits the use of forcible measures to halt and repel an armed attack.

The claim that self-defence as a primary rule (under Article 51) is parasitic on the principle of non-use of force (under Article 2(4)) is supported by many international lawyers. For example, Hans Kelsen wrote: ‘self-defense is the use of force by a person illegally attacked by another. The attack against which the use of force as an act of self-defense is permitted must have been made or must be intended to be made by force’ (p. 784). Similarly, Roger O’Keefe explains: ‘Article 51, as such, is an exception to the prohibition on the use of force laid down in Article 2(4) of the Charter and thus encompasses only forcible measures’ (p. 57). In her Separate Opinion in the ICJ’s Wall Advisory Opinion, Judge Higgins said she was ‘unconvinced that non-forcible measures (such as the building of a wall) fall within self-defence under Article 51 of the Charter as that provision is normally understood’ (para. 35). In its commentary on the Articles on State Responsibility (ASR) (2001), the International Law Commission (ILC) states: ‘Article 51 of the Charter of the United Nations preserves a State’s “inherent right” of self-defence in the face of an armed attack and forms part of the definition of the obligation to refrain from the threat or use of force laid down in Art 2, paragraph 4’ (Article 21, para. 2).

I reject this conceptualisation of the right of self-defence. In my view, self-defence under Article 51 is a general right and can be invoked to justify forcible and non-forcible responses to an armed attack, provided of course that they are necessary and proportionate to counter the attack (for a more detailed elaboration of my views on the status, content, and scope of self-defence see here). As I will explain, this claim is supported by State practice as well as the text and structure of the UN Charter and, perhaps, the use of non-forcible measures is even required by the principle of proportionality. By way of conclusion, I explain why characterising self-defence as a general right under international law is important for conflict de-escalation.

Before moving on, two points are necessary. First, casting self-defence as a general right under international law may not have any direct or immediate bearing on the Israel/Hamas conflict. As Marko says, the pressing question here is whether the right of self-defence is even relevant to the dispute given the ambiguous status of Palestine under international law. Rather, the purpose of this post is to trigger a wider debate as to scope of the right of self-defence under Article 51 of the UN Charter. Second, the focus of this post is on the scope of self-defence as a primary rule under Article 51, that is, what measures can be taken to halt and repel an armed attack. This post is not concerned with Article 21 of the ASR, where the ILC also identifies self-defence as a secondary rule (see Tsagourias). The purpose of Article 21 of the ASR is to preclude the responsibility of defending States for incidental breaches of international law that occur as a result of defensive action being undertaken under Article 51.         

State Practice

Self-defence has a long history in international law and originally took the form of a duty of self-preservation. As State practice at that time reveals, ‘[i]n the last resort almost the whole of the duties of states are subordinated to the right of self-preservation’ (Hall, p. 244). As the duty of self-preservation evolved into the right of self-defence during the mid-nineteenth century, States continued to interpret this right broadly and regularly engaged it to justify a wide range of forcible and non-forcible measures.

For example, during the 1860s, and in order to protect its essential economic interests (which, at that time, were covered by self-defence (Bowett)), the US invoked the doctrine of self-defence to justify its boarding and searching of vessels, and its subsequent apprehension of their crew, that were allegedly engaged in illegal fishing operations in the Bering Sea. The US explained that ‘the right of self-defense on the part of a nation is a perfect and paramount right to which all others are subordinate’ and could be engaged to justify an interference with the principle of the freedom of the high seas (see here, p. 10).

Evidently, the use of military force was bound up in these incidents insofar as the US Coast Guard boarded and searched the vessels without consent and apprehended their crew. But the point is that the US engaged the right of self-defence to justify an interference with the principle of the freedom of the high seas rather than the principle of non-use of force. Importantly, the reason for this was because, at the time, there was no specific prohibition on the use of force in international law. As Brownlie demonstrates, the use of force prohibition formed in customary law at some point between 1928 and 1945 (p. 110). What this shows is that the right of self-defence predates the principle of non-use of force and that these provisions developed at different points in time and in different contexts. Thus, historically, there was no inextricable relationship between the right of self-defence and the principle of non-use of force.

Moreover, State practice subsequent to the signing of the UN Charter illustrates that States have continued to cast self-defence as a general right under international law – in other words, while under Article 51 of the UN Charter States decided to limit the right of self-defence to situations in which an armed attack occurs, they did not limit self-defence to the use of forcible measures only. Take, for example, the North Atlantic treaty. Article 5 of that agreement, which aligns itself with Article 51 of the UN Charter, explains that NATO members commit to:

‘assist the Party or Parties so attacked by taking forthwith, individually and in concert with the other Parties, such action as it deems necessary, including the use of armed force, to restore and maintain the security of the North Atlantic area’.

The phrase ‘including the use of armed force’ clearly shows that collective self-defence under the NATO treaty can include non-forcible measures. Similar provisions can be found in other regional security agreements, such as Article 2 of the Treaty of Joint Defense and Economic Cooperation.

Finally, during the late 1990s and early 2000s, Israel constructed a security wall to protect itself from armed attacks emanating from Palestinian territories. Before the UN General Assembly, Israel justified this non-forcible measure in the following terms:

[A] security fence has proven itself to be one of the most effective non-violent methods for preventing terrorism in the heart of civilian areas. The fence is a measure wholly consistent with the right of States to self-defence enshrined in Article 51 of the Charter (p. 7).

Text and Structure of the UN Charter

The text of Article 51 does not limit self-defence to forcible measures. Moreover, State practice unambiguously entitles States to invoke the right of self-defence to justify the use of force. As a matter of interpretive logic, then, if self-defence covers forcible measures, it must also include non-forcible measures: in eo quod plus sit semper inest et minus (‘in the greater is always included the lesser’). Indeed, Israel invoked this maxim before the UN General Assembly when justifying its security wall:

International law and Security Council resolutions, including resolutions 1368 (2001) and 1373 (2001), have clearly recognized the right of States to use force in self-defence against terrorist attacks, and therefore surely recognize the right to use non-forcible measures to that end (p. 6).

In addition, the structure of the UN Charter points to the fact that self-defence is a general right under international law. Importantly, Article 51 is located in Chapter VII of the UN Charter rather than in Chapter I where the principle of non-use of force is situated. The reason for this is because self-defence is seen as an exception to the collective security system, namely, that self-defence is an emergency provision and can be invoked by States only up to the moment that the Security Council takes measures effective at maintaining international peace and security. For example, when commenting on a draft of Article 51, the Chinese delegate to the San Francisco conference explained that ‘this article amounts to an exception to the enforcement arrangements decided upon by the Security Council’ (p. 287). Thus, the structure of the UN Charter does not suggest that self-defence should be read as an in-built exception to Article 2(4).     


Customary international law requires that defensive action must be necessary and proportionate to counter an armed attack (ICJ, Oil Platforms, paras 76-77). There is a longstanding disagreement among States and scholars as to the meaning of the principle of proportionality in the jus ad bellum context (Buchan and Tsagourias, Chapter 3). One camp maintains that the principle of proportionality only limits States to do no more than is necessary to thwart an armed attack. Understood in this way, the ad bellum principle of proportionately does not require an assessment of the degree of harm caused by the defensive action; rather, the legality of this harm is assessed under other international legal regimes, such as the law of armed conflict, international human rights law, and international environmental law.

By contrast, the other camp resists this narrow interpretation and maintains that the principle of proportionality does require an assessment of the scale of harm inflicted by the defensive action. If one adopts this broader interpretation of the principle of proportionality, and for which there is significant support (although one that the present author does not necessarily share), it must be the case that the right of self-defence permits recourse to non-forcible measures because the principle of proportionality may actually require their use if they can effectively halt and repel an armed attack and cause less harm and suffering than forcible measures.        


States may decide to use non-forcible but otherwise unlawful measures to confront an armed attack, such as economic sanctions and cyber operations. Construing self-defence as an exception to the non-use of force principle would exclude these measures from the ambit of self-defence. As I have shown, this interpretation of the law of self-defence is not supported by international law. Rather, there is a wealth of evidence to indicate that self-defence covers all measures necessary to halt and repel an armed attack, regardless of whether they are forcible or non-forcible in nature. Moreover, this approach is desirable as a matter of policy for two reasons. First, it enhances the effectiveness of the right of self-defence by widening the response options available to States. Second, it can help de-escalate international crises – put differently, interpreting self-defence as an exception to the prohibition on the use of force may push States into using force to counter an armed attack when they may otherwise determine that non-forcible measures are sufficient and adequate.

Print Friendly, PDF & Email

Leave a Comment

Comments for this post are closed


Artem Ripenko says

November 29, 2023

Thanks for your post.
The right of self-defence should encompass a non-forcible response, preventing an unjust limitation of this 'inherent' right. Additionally, the UN Charter's logic, as you've noted, aligns with this perspective. Specifically, the sequencing of reactions is outlined in articles 39-42 of the Charter, akin to a potential model for a self-defensive response….
In my opinion, the right of self-defence also empowers a state to seize the aggressor’s assets as Ukraine did etc. (won't share the link to the blog post here to avoid accusations of self-advertising :))

Dr Mary Jane Fox says

November 29, 2023

Considering how utterly (but not surprisingly) disproportionate the Israeli response has been and is likely to continue to be, and how they have well exceeded the limits of any interpretation of the UN Charter's Article 51, and aside what has been done to Palestinian civilians, they have done themselves a shameful disservice.

Raphaël van Steenberghe says

November 30, 2023

Dear Russell

Many thanks for this post! I have two observations.

Firstly, self-defence before WWI was different from contemporary self-defence to the extent that it conflated with state of necessity – which is now considered as a circumstance precluding wrongfulness, that is, a secondary norm rather than a primary norm like Article 51 of the UN Charter. It was often invoked to justify minor uses of force (measures short of war), including the boarding and searching of vessels. Accordingly, precedents occurring at the time, in which self-defence was invoked – such as Pensacola and other forts in Florida (1815), the pursuit of predatory Indians and other marauders into Mexican territory (1836), the Caroline case (1837), the Mary Lowell case (1869), the Virginius case (1873), the incursions of Tunisian nomads in Algeria (1881) and the Bearing Sea case (1892) – do not prove relevant to the issue at stake.

Secondly, I am not sure that your view that self-defence under Article 51 can be used to justify non-forcible measures clearly fits with contemporary state practice and reflects the traditional view adopted by most states. The state practice you mention is quite limited compared to the numerous declarations expressed by states in various fora (UNSC, UNGA, ICJ, ILC) that self-defence under Article 51 is an exception to the prohibition on the use of force. In addition, a special attention must be paid to the state practice and discussions in relation to the ILC work on current Article 21 ARSIWA. States clearly linked Article 51 to the prohibition on the use of force, while admitting – like the Special Rapporteur – that self-defence could be used to justify non-forcible measures but as a circumstance precluding wrongfulness. In other words, when self-defence is not invoked to justify a use of force, otherwise prohibited by Article 2§4 of the UN Charter, it acts as a secondary norm when used to justify the violation of any other rule. It is also illustrative that Western states helping Ukraine by providing weapons did not expressly invoke Article 51 (collective self-defence) but asserted that they helped Ukraine in exercising its right of self-defence, although such support could be seen as breaching certain norms of international law. This is arguably because Article 51 is intrinsically linked to the prohibition on the use of force and that they do not want their support to be assimilated to a use of force.

I would be happy to have your view on this!

Katie Johnston says

November 30, 2023

Hi Russell. Thank you very much for this interesting post. Generally I tend to share Marko’s view that self-defence is an exception to/limitation on the scope of the prohibition on force, but in relation to the specific argument made here I had a couple of questions. If a state takes non-forcible measures in response to an armed attack against it, such as economic sanctions or cyber operations, then surely these kinds of actions can already find a legal basis as lawful measures of retorsion or, to the extent that they violate the defending state’s obligations, countermeasures. It is not clear to me why a further possible legal basis of non-forcible self-defence is needed, or what it adds?

If the argument is that the right of self-defence can provide a legal basis for non-forcible measures that would not otherwise be lawful then what kind of non-forcible measures might this apply to? Non-forcible measures that do not satisfy the conditions for lawful countermeasures e.g. reversibility, proportionality? Arguably the necessity and proportionality conditions of self-defence, if they could be applied to non-forcible measures, would impose conditions that are not substantially different in practice. Presumably invoking the right of self-defence also could not justify non-forcible measures that violate obligations for the protection of human rights, of a humanitarian character, or under a peremptory norm. I would be interested to hear your view as to what gap this extended right of self-defence would fill, as to me at first sight it seems that the only kind of measure that could be taken by a victim state in response to an armed attack that could not be justified as retorsion or as a lawful countermeasure, but that could be lawful as an exercise of the right of self-defence, is the use of force?

Or is your argument more significant for third states than for victim states? That is, is one implication that, subject to a request from the victim state, the right of collective self-defence might provide a legal basis for non-forcible measures taken against an aggressor state by states other than the state that is the victim of the armed attack? Such a right, if it indeed exists, could be invoked even where those states are unwilling to invoke/endorse the possibility of third-party countermeasures. If so, what (if any) would be the limits on the kinds of non-forcible self-defence measures that could be justified on this basis, other than necessity and proportionality? I know you develop some of these arguments in more detail in the article you linked to, but I’d be interested to hear your thoughts.

Russell Buchan says

November 30, 2023

Dear Raphaël,

Thanks for your comments. To your points:

1. I agree that, during the times you mention, there was no clear conceptual demarcation between self-defence, self-preservation, and necessity. I think Robert Jennings did a good job illustrating this in his AJIL article on the Caroline incident, namely, that different terms were bounded around by different states in different contexts. For me, the point is that states sought to protect their essential interests through a range of different measures, both forcible and non-forcible, thus demonstrating the broad scope of this justification. There is nothing in the text or structure of the UN Charter, nor anything definitive in post-Charter state practice, to suggest that the emerging right of self-defence should be limited to only forcible measures.

2. True, self-defence is typically invoked to justify forcible measures, and this can skew our assessment. Perhaps it was for this reasons that James Crawford and subsequently the ILC linked Arts 2(4) and 51. But I think this is a mistake, as I try to show. And yes, perhaps some states do see Art 51 as being a narrow exception Art 2(4). But again, I think they are in the minority and their practice is not concordant with general state practice nor nor consistent with the text or structure of the UN Charter.

On Art 21 ASR: I see Art 51 as being a justification for all those measures (forcible or non-forcible) that are aimed at halting and repelling an armed attack. Any collateral breaches of international law that do not fall under Art 51 (because they are not designed to counter the attack) have to be addressed under Art 21 ASR as a secondary rule.

On Ukraine, I think the water is muddied by the qualified neutrality doctrine. Thus, what it tells us about the scope of self-defence is limited.


Russell Buchan says

November 30, 2023

Dear Katie,

Many thanks for your comments. You raise an important question about the interplay between the doctrines of self-defence and countermeasures.

For me, countermeasures are law enforcement tools - they are designed to implement state responsibility for breaches of international law. Why would a state want to rely on this doctrine when it is defending itself from an armed attack? In other words, its actions are defensive not law enforcement.

As law enforcement tools, countermeasures are subject to a number of strict exceptions e.g., notification, reversible harm, cannot be anticipatory, not directed against non-state actors, etc.

As an act of self-preservation, self-defence is not subject to such limitations e.g., of course a state does not have to notify an aggressor that it is about to defend itself or that defensive harm has to be reversible.

SD is subject to the principles of necessity and proportionality, of course, but they take on different meanings in the jus ad bellum than in the law of countermeasures e.g., they are calibrated to meet the needs of self-preservation.

So, for me, CMs are a very poor fit for non-forcible defensive measures that are otherwise unlawful e.g., a cyber operation that is designed to halt an armed attack but in doing so prima facie breaches the principle of non-intervention.


Artem Ripenko says

December 1, 2023

Dear Katie,
An example could be the confiscation of assets of an aggressor state in self-defence (together with an armed response or, hypothetically, as a separate action). CM seem not to give grounds for that.

Artem Ripenko says

December 1, 2023

Dear Raphaël,

Perhaps the Western states were cautious of ‘co-belligerency’, the wording of the NATO treaty etc.

I may be wrong, but there seem to be few grounds for providing military assistance beyond collective self-defence.

Some of the assisting states were silent in justifying their assistance. Could this lack of formal declaration prove that collective self-defence was not invoked?

Anyway, even if the Western states confiscated Russian state assets, sending them as assistance to strengthen Ukraine’s military potential in collective self-defence, would that equal the use of force?