1 - The Use of Force in Self-Defence against Non-State Actors, Decline of Collective Security and the Rise of Unilateralism: Whither International Law?

Summary

This Trialogue concerns the use of force in self-defence against non-State actors in the territory of third States under international law. International law relating to the use of force, in particular international law on the rules of self-defence, is immensely important. It is immensely important because one of the most fundamental objectives of modern international law, and an overriding objective of the UN Charter, is the prevention of war.¹ International law on the use of force provides the framework within which the objective recited in the preamble of the UN Charter to ‘save succeeding generations from the scourge of war’ is pursued by the United Nations. Having clear and precise scope and limits of this law is necessary to achieving this lofty objective. Yet the importance of the international law on the use of force is matched by the controversy surrounding the very scope and limits of the law on the use of force.

I. Introduction

A. Importance and Controversy of the Law on the Use of Force

This Trialogue concerns the use of force in self-defence against non-State actors in the territory of third States under international law. International law relating to the use of force, in particular international law on the rules of self-defence, is immensely important. It is immensely important because one of the most fundamental objectives of modern international law, and an overriding objective of the UN Charter, is the prevention of war.¹ International law on the use of force provides the framework within which the objective recited in the preamble of the UN Charter to ‘save succeeding generations from the scourge of war’ is pursued by the United Nations. Having clear and precise scope and limits of this law is necessary to achieving this lofty objective. Yet the importance of the international law on the use of force is matched by the controversy surrounding the very scope and limits of the law on the use of force.

In many ways, as will become clear in the course of this chapter, the controversy is centred not so much around the use of force per se, but rather on the right to use force unilaterally. Collective use of force in accordance with the Charter of the United Nations – in principle this refers to the use of force authorised by the UN Security Council – does not raise the same controversies as those raised in this chapter. This is not to say that collective use of force does not raise controversies at all, for it surely does. Whether the Council has unlimited power to authorise force, who may be the addressees of the authorisation, under what circumstances force may be authorised and for what purpose are some of the issues that have been the subject of much intellectual discourse over the course of the UN Charter’s existence.² Similarly, the very question of what constitutes ‘collective use of force’, beyond explicit authorisation of the UN Security Council, remains an issue of debate. Can, for example, action under the North Atlantic Treaty Organisation (NATO), or the African Union (AU) without UN Security Council authorisation qualify as collective measures under the Charter?³ These issues, interesting though they are, fall beyond the scope of this contribution. Instead, the controversy that this chapter concerns itself with is that relating to the unilateral use of force.

The controversy surrounding the law on the unilateral use of force is manifested in different ways. None, however, is as pronounced as the controversies concerning the unilateral use of force in self-defence. The debate concerning humanitarian intervention is, for example, largely passé, since the concept is rejected by a large majority of States.⁴ The controversy concerning the right to use force unilaterally in self-defence raises questions concerning the requirements for self-defence, the precise limits of which appear sufficiently malleable to attract widely divergent approaches.⁵ Some scholars and States adopt a generous and permissive approach to the law on self-defence, granting States a comparatively large margin to decide when, how and under what circumstances force may be used in self-defence, which in turn means a narrow approach to the prohibition on the use of force.⁶ Others, on the other hand, adopt a more restrictive approach to the use of force in self-defence, which, in turn, implies a generous approach to the prohibition of the use of force. Commenting on the contrast between the restrictive and permissive approaches, Arimatsu and Schmitt state that there is

inherent ambiguity in the law governing the use of force by States. Such ambiguity should not be surprising because States are caught on the horns of a dilemma. On the one hand, clear and restrictive norms serve to enhance international peace and security by limiting what States may do in terms of force. Yet, on the other, opacity benefits the State when it is facing threats.⁷

Very often, the question whether particular acts of self-defence are in conformity with international law is considered from the perspective of legal rules on proportionality and necessity,⁸ and rules concerning anticipatory self-defence.⁹ The latest, and perhaps most complicated, controversy surrounds the question of whether force may be used in self-defence against non-State actors in the territory of a third State. There is a growing trend, including in the writings of some commentators and the views and actions of some States, suggesting that force may be used against non-State actors in the territory of innocent third States.¹⁰ It is this question that this Trialogue is concerned with.

When these various aspects affecting the scope of the prohibition on the use of force are placed in a single crucible, the possible interpretations of what is permissible under international law on the use of force become many – almost indeterminable. At the narrowest end of the spectrum, the use of force in self-defence would only be permitted in cases where an armed attack by a State has already commenced. At the other, more permissive, end of the spectrum, the use of force would be permitted where there is a possibility of an attack by a non-State actor operating on the territory of a third State. Between these possibilities, and taking into account the fact or circumstance specific dimensions pertaining to proportionality and necessity, there is a continuum of other options on the scope of the right to use force in self-defence. The argument in this contribution is that there is an objectively correct interpretation of the rules on the use of force which can be discovered by the proper application of the tools of interpretation and methodology of international law.¹¹

B. The Role of Policy Considerations

While there is an objectively correct interpretation of the rules on the use of force, it is also true that law does not exist in a vacuum. Law exists in the real world, to curb real-world problems. Policy considerations, therefore, tend to play a significant role in the approach that States and commentators take with respect to law, in particular international law. This also applies to the law on the use of force and, in particular, the right to use force in self-defence. That policy considerations play a role, however, should not be taken to mean that law does not provide objectively correct answers or that the law is, so to speak, in the eye of the beholder. Such approaches risk undermining the rules of international law and the UN’s objective to ‘save succeeding generations from the scourge of war’.

Policy considerations have a direct influence on the content of objective law in a number of ways. Policy considerations may provide context for the interpretation of rules of law, whether this is through the ‘object and purpose’, ‘context’ or ‘relevant rules of international law’ elements of interpretation, or the assessment of practice, either for the purpose of treaty interpretation (cf. Article 31(3) of the Vienna Convention on the Law of Treaties –VCLT)¹² or formation of customary international law. More importantly, policy considerations influence how States behave and, consequently, will affect practice, which is central to both the evolution and interpretation of law. It is important to stress, however, that the policy considerations do not replace law, but rather influence its development and assessment.

One important policy consideration that may well influence the evolution of the law is the change in the nature of conflict from purely inter-State to intra-State conflict involving non-State actors and the growth of terrorism as a new threat to international stability. The terrorist attack of September 2001 in the territory of the United States has, largely because of US dominance in international affairs, caused a shift in strategies to combat terrorism.¹³ More broadly, the emergence of non-State entities such as Al-Qaeda (responsible for those attacks in September 2001), Boko Haram, Al-Shabaab and more recently the Islamic State of Iraq and Syria (ISIS), has changed the nature of terrorism, and this may necessitate new strategies for confronting terrorism, including the use of force as a response strategy.¹⁴ The rise in terrorism could be a basis for turning away from the prohibition of force as a legal norm. As a policy matter, it may well be asked why a State like Nigeria should not have the right to use force against Boko Haram if it is operating from Cameroon or the Central African Republic to launch attacks in its territory. As a policy question it may also be asked whether non-State actors operating in failed States should be permitted to carry out ‘armed attacks’ against innocent third States without consequences because rules of international law permit the use of force in defence only against culpable States. Should, in the light of these realities (situation on the ground), States be afforded a greater margin to unilaterally use force ‘in self-defence’?

These policy considerations, flowing from ‘facts on the ground’, can have three different possible impacts on the law.¹⁵ First, they may suggest a need to elastically interpret the rules of the Charter to enable a broader interpretation of the rules. Second, they may require us to jettison altogether the rules of law relating to the use of force to permit a completely unrestrained right to use force unilaterally. Glennon, for example, has asserted that given the realities on the ground, the insistence on the rules of international law on the use of force is ‘illusory’.¹⁶ A third possibility is the argument that, independent of the growth in terrorism, the rules of international law on the use of force in self-defence have always been permissive.

These realities on the ground could influence policy positions concerning how broad or narrow to interpret the law on the use force in self-defence. To influence the development (or evolution) of international law, however, requires that they also influence the practice of States. In other words, it is not enough to show that policy or normative considerations require a particular – narrow or broad – interpretation of the rules on the use of force in self-defence. To influence the development of the law, it is necessary to show that these policy and normative considerations have influenced the practice of States, either for the purpose of the formation of customary international law or treaty interpretation.

C. The Purpose of the Chapter

This chapter of the Trialogue focuses on a specific problem,¹⁷ namely whether international law permits the use of force in self-defence in the territory of a State that has not consented and to which the actions of the non-State actors cannot be attributed – referred to in a previous contribution by the author as ‘innocent States’.¹⁸ The phrase ‘innocent State’ in this context means nothing more than a State on which an armed attack cannot be attributed. The relevant States may be less than innocent in general (they may even be generally ‘evil’). Their ‘innocence’ in this regard is a shorthand denoting only that they are ‘innocent’ of the acts of the non-State actor, or that the acts of the non-State actor are not attributable to them.

Most authors agree that the starting point for any attempt to decipher the rules under international law relating to the use of force must be Article 2(4) and Article 51 of the UN Charter.¹⁹ However, it is also important to understand these provisions as part of the larger system of collective security established under the Charter. Interpreting their precise scope and limits must, therefore, be informed by the context of the system of collective security. The collective security system will therefore be described with a view to providing context for the interpretation of the rules on the use of force in self-defence.

The next section seeks to identify core elements of the prohibition of the use of force. Section III provides a sketch of the collective security system established by the Charter, of which the prohibition of the use of force and the exception permitting the use of force in defence in Article 51 are a part. Section IV presents the arguments often advanced for an expansive reading of self-defence which would permit the right to use force against non-State actors in the territory of innocent third States. These arguments are based mainly on an assessment of practice. Section V evaluates the arguments for an expansive interpretation of self-defence based on the tools and methodology of interpretation under international law. Section VI considers the legal and policy implications of the opposing interpretations of the right to self-defence. This contribution makes the following conclusions. First, the practice often referred to in support of the expansive interpretation of self-defence cannot sustain the argument for an expansive interpretation. Second, the rules of interpretation support a narrow reading of self-defence in which attribution of the acts of non-State actors to the third State is required in order to allow for self-defence on the territory of that third State. Third, the collective security system, established by the Charter, which seeks to prevent unilateral use of force, provides further support for an attribution-based approach to self-defence.

It is worth setting out, at the outset, the approach and normative premises from which the chapter proceeds. First, this chapter is grounded in a positivist approach to the identification of the law. In other words, as described above, policy considerations are relevant only to the extent that they are supported by the traditional materials for the identification and interpretation of law. Second, powerful States tend to seek to ‘de-constrain’ themselves from the shackles of law while leaving the illusion of the constraining power of law in order to constrain the less powerful.²⁰ I am thus concerned about – and thus seek to avoid – an interpretation of law that facilitates the ‘de-constraining’ through an expansive interpretation that may permit the limits of the right to use force in self-defence to be in the eye of the beholder and thus benefit the (militarily) powerful. The position, however, is not merely a policy preference, it is also one which is grounded in the law.

A final preliminary point concerning the approach in this chapter: I adopt a decidedly doctrinal approach to the question of the use of force against non-State actors and refuse to be engaged with what may be termed a ‘fact or circumstance specific’ approach. ‘Fact or circumstance specific’ approaches suggest that the dangers of a broad interpretation of the right to use force in self-defence against non-State actors can be addressed by reliance on concepts such as proportionality and necessity as requirements of self-defence. The problem with these approaches is that they assume, incorrectly in my view, that, in principle, international law permits the use of force in self-defence against non-State actors in the territory of non-consenting third States. The ‘fact or circumstance specific’ approaches obviate the need to address the principles of the prohibition on the use of force and places us on a slippery slope. It is a reflection of a trend to chip away at the legal principle itself and to leave us at the point where the law is really in the eye of the beholder. Where there are no principle constraints against the use of force, and the facts and circumstances constrain behaviour, it becomes easier, in politically convenient circumstances, to suggest that the law permits a particular course of action. The reliance on facts or circumstance specific arguments is, in my view, the road to erosion of the law prohibiting the use of force.

II. Prohibition on the Use of Force

A. A Brief Historical Context

It is now beyond dispute that international law prohibits the use of force and Article 2(4) of the Charter reflects the basic rule concerning the use of force. As with any other provision of any treaty, Article 2(4) of the Charter is to be interpreted using the rules of interpretation of international law. Nonetheless, to grasp the breadth (and limits) of the prohibition on the use of force requires an understanding of the historical evolution of the prohibition under international law. This historical evolution provides context to the rule in Article 2(4). Moreover, as will be seen below, a historical event, namely the Caroline incident, forms a central element of the argument in favour of permitting the use of force in self-defence against non-State actors in the territory of third States. It is therefore apposite to begin the analysis by a description of the history of the prohibition of the use of force in international law.

International law did not always prohibit the use of force as an instrument of foreign policy.²¹ The Caroline incident should be seen in the light of the nature of pre-twentieth-century international law, based on a bundle of bilateral relations, in contrast to the new international law based on community values.²² In this classical international law, States, as sovereign entities, were free to determine the nature of their international relations including by relying on the comparative advantage of military might.

The view that international law has always prohibited the resort to force is often based on the just war theory. Yet as Randelzhofer and Dörr correctly observe, ‘the medieval theory of bellum iustum had been developed by theologians and was never a valid rule of public international law.’²³ For others, the widely cited Caroline incident– involving an exchange in 1842 between the United States and the United Kingdom on the use of force by the latter on the territory of the former – is often advanced as an illustration that the use of force was prohibited in the nineteenth century.²⁴ However, it is not at all clear that the exchange between Lord Ashburton, for the United Kingdom, and Daniel Webster, for the United States, was meant to convey legal propositions. Justification for the use of force, in particular self-defence, was often advanced for ‘political expediency’ and in ‘order to secure moral high ground’, rather than as a shield against legal wrongfulness.²⁵ In this regard, in the exchange Lord Ashburton declares that it should not be lightly assumed that the United Kingdom would ‘provoke a great and powerful neighbour’ – a clear appeal to the political and expediency.²⁶ Brownlie similarly states that the reference to the Caroline incident ‘as the critical date for the customary international law’ on self-defence is both ‘anachronistic and indefensible’.²⁷ The position advanced by Randelzhofer and Dörr, that up and until 1919 public international law ‘did not know any rules about when it [was] permissible to wage war’ and that the use of force was ‘permitted in the relations between States without any conditions’,²⁸ is therefore the correct position in law.

B. The Content and Status of the Prohibition

The right to use force in self-defence is an exception to a general rule, namely the prohibition on the threat or use of force. The content of the exception – the right to use force in self-defence – can only be understood in the light of the general rule – the prohibition on the use of force.²⁹ While Randelzhofer and Dörr state that the two provisions ‘do not exactly correspond to one another in scope’, they do suggest the phrase ‘armed attack’, used in Article 51, is narrower than the notion of ‘use of force’.³⁰ Determining the scope of Article 2(4) can therefore be useful in identifying at least the limits of what is meant by ‘armed attack’ in Article 51. As a treaty rule, the prohibition on the use of force is to be found in Article 2(4) of the Charter of the United Nations. It provides as follows:

All members shall refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any State, or in any other manner inconsistent with the Purposes of the United Nations.

The ICJ, in the Military and Paramilitary Activities case, identified several key aspects of the prohibition. First, the Court made the distinction between, on the one hand, ‘the most grave forms of the use of force’ and, on the other hand, ‘other less grave forms’ of the use of force.³¹ According to the Court, it is the former forms of the use of force – the most grave forms – that constitute an armed attack.³² Other ‘less grave forms’ of the use of force remain, nonetheless, prohibited by Article 2(4). It appears that the significance of the distinction by the Court is that the right to use of force in self-defence, in terms of Article 51 of the Charter, applies in respect to an armed attack, i.e. ‘the most grave forms of the use of force’ and not to ‘other less grave forms’.³³ As a consequence of this distinction, the Court states that ‘assistance to rebels … in the form of provision of weapons or logistical support … may be regarded as threat or use of force’ but not armed attack.³⁴ Examples identified by the Court, and drawn from the UN Declaration on Principles of International Law concerning Friendly Relations and Cooperation among States, of ‘less grave forms of the use of force’, include ‘organizing, or encouraging the organization of irregular forces … for incursion into the territory of another State’, ‘instigating, assisting or participating in acts of civil strife or terrorist acts in another State or acquiescing in organized activities within its territory’.³⁵ These actions by a third State would amount to a violation of the prohibition of the use of force but would not rise to the level of an armed attack. The consequence of this appears to be that while such acts would be unlawful and thus entitle the affected State to invoke the responsibility of the State providing assistance, it would not entitle the affected State to use force in response.

Other international instruments contributing to an understanding of the prohibition on the use of force include Resolution 3314 (XXIX) on the definition of aggression³⁶ and the International Law Commission’s 1996 Draft Code of Crimes against the Peace.³⁷ In its commentary to Draft Article 16 on the crime of aggression, the Commission notes that Article 16 refers to ‘aggression committed by a State’ and that the ‘rule of international law which prohibits aggression by a State applies to the conduct of a State in relation to another State’.³⁸

The prohibition on the use of force has been described variously as ‘the central rule’ of international law,³⁹ ‘one of the cornerstones of the modern international law legal order’,⁴⁰ and a ‘basic rule of the Charter’.⁴¹ The International Court of Justice has itself recognised the principle, not only as a ‘cornerstone of the United Nations Charter’,⁴² but as a rule of customary international law.⁴³ Indeed, the prohibition on the use of force is generally accepted as a norm of ius cogens.⁴⁴ Already in 1966, the International Law Commission, while not venturing to provide details on ius cogens, was able to declare boldly that ‘the law of the Charter concerning the prohibition of the use of force in itself constitutes a conspicuous example of a rule’ of ius cogens in international law.⁴⁵ More ambiguously, the International Court of Justice in the Military and Paramilitary Activities case stated that the International Law Commission ‘expressed the view that “the law of the Charter concerning the prohibition of the use of force itself constitutes a conspicuous example”’ of a ius cogens norm.⁴⁶ There has been some debate as to whether, by this language, the Court was endorsing the view expressed by the Commission or whether the Court was being non-committal.⁴⁷ Whatever the reluctance of the Court to unambiguously declare the prohibition of the use of force as a peremptory norm of international law, it is now beyond dispute that the prohibition on the use of force is a norm of ius cogens.⁴⁸

The questions of whether the prohibition on the use of force is ius cogens and, more to the point, whether it matters, are addressed in the two other chapters in this Trialogue. For Mary Ellen O’Connell, the prohibition is ius cogens and the characterisation of the prohibition as ius cogens matters for the question under consideration, i.e., whether a State may use force in self-defence in the territory of third States. Christian Tams does not offer a view as to whether the prohibition is or is not ius cogens, but states merely that it is irrelevant for the determination of the question. I share, without hesitation, the view expressed by O’Connell that the prohibition falls within the ambit of norms of ius cogens. That said, whether identifying the prohibition as ius cogens or not matters for the question at issue in the Trialogue depends on what is characterised as ius cogens. If what is characterised as ius cogens is the prohibition of force itself, then it clearly matters. In the resolution of tension (rather than conflict) between a ius cogens norm (in this case the prohibition on the use of force) and another norm (in this case self-defence), it would be expected that the ius cogens norm would be given priority. Thus if the prohibition of the use of force is ius cogens, the right to use of force in self-defence would have to be strictly circumscribed so that it does not affect the core of the ius cogens norm. For that reason, if the ius cogens norm is the prohibition itself, then it would clearly have an important role to play in the interaction between the prohibition and the right to use of force in self-defence. However, if the ius cogens norm is, as the International Law Commission has described, ‘the law of the Charter concerning the prohibition of the use of force’, then the characterisation as ius cogens is significantly less important for the purposes of the question under consideration in the Trialogue.⁴⁹ This is because the ‘law of the Charter concerning the prohibition of the use of force’ includes not only the prohibition but all of the Charter provisions relating to the use of force, including the right to use of force in self-defence. In those circumstances, the characterisation of the law as ius cogens has no effect on the interaction between the prohibition of the use of force and the right to use force in self-defence.

While the prohibition of the use of force is central to the UN system and modern international law, it forms part of a broader law on peace and security. This broader law includes, as an exception to the prohibition on the use of force, the right to use force in self-defence. However, the ‘law of the Charter’ relating to the prohibition of the use of force is underpinned by the collective security framework of the United Nations. This collective security system is also relevant in circumscribing the relative scope of the right to use force in self-defence vis-à-vis the prohibition on the use of force. It is this collective security system that is addressed next.

III. International Peace and Security Architecture

A. The Charter as an Instrument for Collective Security

While most contributions (correctly) assert that the starting point for the assessment of the law on the use of force in self-defence is Articles 2(4) and 51 of the UN Charter, these provisions themselves, and the customary international law rules they reflect, must be understood in their context. That context, I suggest, is the collective security system established under the Charter. It should be taken into account in trying to identify the limits of the right to use of force in self-defence against non-State actors in the territory of third States.

The Charter attempts to give life to its inspirational call to ‘save succeeding generations from the scourge of war’ by providing both normative rules and institutional mechanisms for overseeing the normative rules. It begins, in Chapter I, by laying out a series of purposes and principles peppered with the objective of peace and security. It is not surprising that the first purpose, and arguably the most important purpose, of the United Nations, identified in Article 1(1), of the Charter is to ‘maintain international peace and security’.⁵⁰ More significant is that to achieve this primary purpose, the Charter sets out as a secondary purpose the taking of ‘effective collective measures for the prevention and removal of threats to the peace’. Thus, under the Charter the principal way to remove threats to peace, including armed aggression, is through ‘collective measures’. As Kohen observes, in the system set up by the Charter ‘the motto was “peace through collective security”.’⁵¹ Similarly, just after the adoption of the Charter and the creation of the United Nations, the opening line in an article by Hans Kelsen declared that collective security ‘is the main purpose of the United Nations’.⁵²

According to this general framework, the prohibition of the use of force in Article 2(4) is the primary normative rule for the maintenance of international peace and security, and the principal response to its breach is through collective measures.⁵³ Danchin defines collective security as the agreement by States to ‘abide by certain norms and rules and to maintain stability and, when necessary, band together to stop aggression’.⁵⁴ This definition has three elements: first, the reliance on legal norms; second, the appropriate forms of response to breaches of (or threats to) security; and third the rejection of self-help in favour of collective responses when breaches do happen.⁵⁵ The third element, ‘the rejection of self-help’, is particularly noteworthy.⁵⁶ Danchin draws his definition from Kupchan and Kupchan’s definition.⁵⁷ It is also noteworthy that Kupchan and Kupchan, in setting out their definition, emphasise a clear contrast between collective security, on the one hand, and, on the other hand, ‘self-help balancing predicated on the notion of each for his own’⁵⁸ and ‘anarchy [where] states fend for themselves’.⁵⁹ Danchin places the system of collective security between a system of balance of power – ‘every “man” for himself’ – and a system of global government where States are deprived of their power to the use of force in favour of a global government that has complete monopoly over force.⁶⁰ Collective security, he states, ‘sits uneasily between and incorporates elements of both’ sides.⁶¹ The system, he continues

involves a centralization of authority over the use of force to the extent that states are deprived of the legal right to use violence at their own discretion. That states give up this discretion and agree to follow objective rules governing the threat and use of force requires an international organization with the authority not only to determine when resort to force is illegitimate but also to require states to collaborate under its discretion in suppressing such use of force.⁶²

B. Institutional Framework for Peace and Security under the Charter

The view taken in this chapter, namely that the key to understanding the scope of the right in Article 51 is the collective security system set up by the Charter, necessitates that the organs and entities created by the Charter for carrying out its purposes are discussed. This is particularly true since one of the arguments for an expansive interpretation rests on the impotence or ineffectiveness of the collective security system.

The UN Charter establishes three principal organs relevant to the maintenance of international peace and security, namely the UN Security Council, the General Assembly and the International Court of Justice. While arguments can be made for a recognition of the role that other organs of the United Nations can play in the maintenance of international peace and security, it is realistic to limit ourselves to these three principal organs. Indeed, arguments can be made for the exclusion of the International Court of Justice, whose primary mandate is the resolution of disputes, since its contribution to peace and security is, in a sense, indirect.⁶³ Nonetheless, given the important contribution that the Court has made to understanding the content of the rules, it is worth including here. Its particular contribution, however, will be considered not in this section, but in section IV on the prohibition of the use of force and the section on self-defence.

The Security Council is the lynchpin of the collective security system established under the Charter. It has the primary responsibility for the maintenance of international peace and security, a responsibility conferred on it by and exercised on behalf of the members of the United Nations. In some sense, as observed by Anne Peters, Article 24(1) of the UN Charter operationalises, at least partly, the primary purpose of the United Nations to maintain (and restore in case of breach) international peace and security.⁶⁴ The specific powers granted to the Council to discharge its mandate for the maintenance of international peace and security are to be found in Chapters VI and VII.

While the Council has wide powers, for the purposes of the use of force and self-defence, two powers are of particular importance. First, the Security Council has the power to make a determination ‘about the existence of any threat to the peace, breach of the peace, or an act of aggression’.⁶⁵ This means that, over and above whatever measures the Council may decide upon to restore or maintain peace, a determination by the Council of aggression is relevant in making a determination of whether the use of force is consistent or not with the prohibition in Article 2(4) of the Charter. Second, the Security Council is empowered to ‘take such action by air, sea or land forces as may be necessary to maintain or restore international peace and security’.⁶⁶ Thus, as a measure of collective security, where use of force in contravention of Article 2(4) has taken place, the Council is empowered to make use of force necessary in response to the aggressive use of force. Of course, in practice, the Security Council hasn’t used force itself but rather has authorised willing States to take action on its behalf.⁶⁷ And the idea, certainly as originally envisaged by the drafters of the Charter, was that the Council’s use of force (or authorisation of the use of force) would ‘be the main exception to the prohibition on force’.⁶⁸ The unilateral use of force in self-defence, in response to an armed attack emanating from the territory of a third State, should be understood in this context, i.e. as an exception to the exception where the main exception, namely collective use of force in response to aggression, has not occurred.

While it has sometimes been contended that the Security Council possesses exclusive powers in relation to peace and security, it is clear that the General Assembly also plays a role.⁶⁹ Moreover, the International Court of Justice has explicitly held ‘that Article 24 refers to a primary, but not necessarily exclusive competence’ and continued to note that the General Assembly does have powers in relation to peace and security.⁷⁰ Articles 10, 11, 12 and 14 make plain that the General Assembly may consider any matter ‘within the scope of’ the Charter, including matters of peace and security. Indeed, by virtue of the fact that the General Assembly has the power to discuss ‘any matter’, while all other organs of the United Nations are limited to specific areas, it may be said that the General Assembly ‘is given pride of place’ in the structure of the United Nations.⁷¹ While the scope of its mandate is wide, the General Assembly’s powers in the areas in which it may ‘act’ are limited, first in the sense that it may only discuss and make recommendations, and second in the sense that it may not make recommendations where the Council is seized with the matter. Thus, while the General Assembly has pride of place by virtue of the breadth of the scope of its mandate, the Security Council can claim preeminence by virtue of the depth of its powers in the narrow area of its mandate.

To start with, the fact that the General Assembly can, according to the text of the Charter, do no more than discuss and make recommendations is a significant limitation on the mandate of the General Assembly with regards to peace and security. The second limitation is that the General Assembly may not make recommendations while the Security Council is exercising its functions in respect of any dispute or situation. This requirement may raise questions about precisely when the Council ‘is exercising in respect of any dispute or situation [its] functions’ (Article 12(1) of the Charter). Does this include any situation on the agenda of the Security Council, or only those situations in which the Council has adopted measures?⁷² If it is the latter, do the measures have to be effective or recent? Moreover, does the fact that the Security Council is seized with a situation preclude recommendations on any issue relating to that situation, even issues tangential to the particular situation?

The practice of the UN reveals a rather flexible approach to the limitations on the General Assembly. The most direct example of the flexibility adopted by the General Assembly is the Uniting for Peace Resolution of November 1950.⁷³ While the Charter provides that the General Assembly may not make a recommendation on a matter under consideration by the Security Council, the Uniting for Peace Resolution provides that it may make such a recommendation where the Council ‘fails to exercise its primary responsibility’ due to ‘the lack of unanimity of the permanent members’.⁷⁴ Furthermore, the resolution permits the authorisation by the General Assembly through recommendation of the use of force in response to aggression.⁷⁵ The International Court of Justice has already affirmed the legality and validity of the Uniting for Peace Resolution.⁷⁶ If, as appears to be the view of the Court, the Uniting for Peace Resolution is legal and valid, then presumably the use of force pursuant to an authorisation given under the Uniting for Peace Resolution would be justified, notwithstanding the apparently clear language of the Charter. Moreover, even outside of the Uniting for Peace Resolution, the practice of the General Assembly has been to consider and make recommendations on matters that are still on the agenda of the Security Council.⁷⁷ The situation in Syria – discussed more fully further below – provides an example of the General Assembly’s assertion of powers even while the Security Council is seized with a situation and without relying on the Uniting for Peace Resolution.⁷⁸ In these resolutions the General Assembly did more than discuss; it made ‘demands’, ‘called upon’ and ‘decided’ various things concerning the situation in Syria.⁷⁹ Importantly, in each of the resolutions on the situation, the General Assembly makes a determination about the lawfulness of the conduct of the relevant parties to the conflict.⁸⁰ While the General Assembly’s mandate to authorise through recommendations the use of force may be in question,⁸¹ what is not in question is that, in practice the Assembly, together with the Security Council, is able to make determinations about the lawfulness of the use of force in given circumstances and, in this way, contribute to an understanding of the law on the use of force.⁸²

The effectiveness of the system of collective security described above is of course based on particular geopolitical considerations. In particular, the dynamics on the Security Council, especially between the permanent members of the Security Council, often determine whether the collective security system established by the United Nations is able to function effectively. Some have thus suggested that the scope of the rules relating to the right to use force unilaterally should be determined by the effectiveness of the collective security system at any given time. Gray, for example, observes that

Many US commentators argued during the Cold War that the interpretation of Article 2(4) depended on the effective functioning of the UN collective security system, and therefore that the inability of the Security Council to act because of the veto of the permanent five meant that Article 2(4) should be read to allow the use of force to further ‘world public order’ or the principles of the United Nations.⁸³

This view was also advanced by Judge Jennings in his dissenting opinion in the Military and Paramilitary Activities case, in which he stated that

the original scheme of the United Nations Charter, whereby force would be deployed by the United Nations itself, in accordance with the provisions of Chapter VII of the Charter, has never come into effect. Therefore, an essential element in the Charter is totally missing. In this situation it seems dangerous to define unnecessarily strictly the conditions for lawful self-defence, so as to leave a large area where both a forcible response to force is forbidden and yet the United Nations employment of force, which was intended to fill the gap, is absent.⁸⁴

The Court itself, however, rejected this view, holding that the prohibition in Article 2(4) is ‘not as such conditioned by the provisions relating to the collective security’.⁸⁵ The problem with the interpretation making the full scope of Article 2(4) dependent on the effective functioning of the Security Council is that it deprives the prohibition on the unilateral use of force of an objective, independent content. Its content will always oscillate between broad and narrow depending on the effectiveness of the collective security system at a particular point in time. As stated at the beginning, the rules prohibiting the use of force have an objective and determinable content and are not in the eyes of the beholder. At any rate, the argument must fail since, as described above, in the event that the Council fails to act because of the veto, the General Assembly, by virtue of the Uniting for Peace Resolution and/or the practice described above, can fulfil the collective security functions. At the very least, given the controversy of interpreting the Charter as permitting the General Assembly to authorise the use of force, the General Assembly can make a determination that the use of force in self-defence is permissible because a situation constitutes an armed attack or aggression.

This framework of collective security established by the United Nations, and centred around the Security Council with the General Assembly playing a secondary role, provides the principal mechanism for addressing threats to international peace and security, with the collective use of force under Article 42 being the principal exception to the prohibition on the use of force. Reliance on self-help, as a means to respond to aggressive use of force, should therefore be seen as an exception to this principal exception. As Kelsen observes, ‘self-defense as a case of decentralized use of force is an exceptional and provisional interlude between an act of illegal use of force … and the collective enforcement action.’⁸⁶ This framework for collective security, underpinned by the Security Council’s primary responsibility, provides the context within which the rules relating to the use of force should be seen.

IV. The Law on Self-Defence

A. General Framework

The right to unilaterally use force in self-defence should be understood in the broader context of the collective security framework set up under the Charter. Article 51 provides as follows:

Nothing in the present Charter shall impair the inherent right of individual or collective self-defence if an armed attack occurs against a Member of the United Nations, until the Security Council has taken the measures necessary to maintain international peace and security. Measures taken by Members in the exercise of the right to self-defence shall be immediately reported to the Security Council and shall not in any way affect the authority and responsibility of the Security Council under the present Charter to take at any time such action as it deems necessary in order to maintain or restore international peace and security.

That Article 51 is situated in Chapter VII, the chapter wherein the primary powers of the Security Council are provided for, serves to reinforce that the right to use force in self-defence should be seen in the context of the collective security framework described in Section III above. Under this framework, there is a prohibition on the use of force reflected in Article 2(4) of the Charter. In the event that this prohibition is breached, collective action, principally under Chapter VII, is foreseen. Unilateral self-defence, therefore, is an exceptional and ‘temporary right’.⁸⁷ Indeed the International Court of Justice has suggested that the requirement to report to the Council is a factor to be considered, even under customary international law, in assessing the lawfulness of self-defence.⁸⁸ As with the prohibition on the right to use of force, the right to unilateral use of force has also been recognised to exist in customary international law.⁸⁹ The content of both Article 51 and that customary international law on self-defence are co-extensive.

B. The Permissibility of Unilateral Use of Force against Non-State Actors

At face value one would expect that the law on the use of force, including rules on self-defence, is applicable to inter-State conflict. As Gray observes, the two provisions in the Charter regulating the unilateral use of force, Articles 2(4) and 51, ‘are very much a response to the Second World War and are accordingly directed to inter-state conflict’.⁹⁰ Moreover, Article 2(4) of the Charter expressly applies to inter-State use of force. It prohibits ‘Members’, which are States, from using force ‘against the territorial integrity or political independence of any State’. As suggested above, if Article 51 was intended to be an exception to Article 2(4), one would expect it to regulate similar conduct, i.e. permit what is ordinarily prohibited in Article 2(4) subject to the particular requirements laid down in its provision.⁹¹ As Kelsen observed shortly after the adoption of the Charter, Article 51 ‘restricts the right to self-defence to the case of an “armed attack” actually made by one state against another’.⁹² Even Tams, who supports the idea that under international law, as it now stands, the use of force in the territory of another State in response to an attack by a non-State is permissible, accepts that the dominant theory used to be contrary to the proposition he now supports.⁹³

In recent years, however, there has been a trend, mainly in European and American literature, that advances a broad and permissive interpretation of Article 51 so as to permit the use of force in self-defence against non-State actors in the territory of third States. According to this trend, international law, as it currently stands, permits States to use force unilaterally, i.e. outside of the collective security framework of the United Nations, against a non-State actor in the territory of a third State without the third State’s consent.

While this broad and permissive interpretation has been propagated in literature,⁹⁴ it has also been advanced in at least three widely discussed policy-oriented documents. The first of these, the Chatham House Principles, was adopted in October 2005 by the International Law Programme at Chatham House, based on consultation with an eminent group of British international lawyers.⁹⁵ The second document, the Leiden Policy Recommendations (Leiden Recommendations), was adopted in April 2010 and ‘emanated from a consultative process’ led by leading Dutch jurists Nico Schrijver and Larissa van den Herik.⁹⁶ The participants in the adoption of the Leiden Recommendations were predominantly from Europe and North America,⁹⁷ and all of the participants in the group addressing the use of force for the Chatham House Principles were from Europe and North America.⁹⁸ Finally, the Bethlehem Principles are a set of principles put forward by Daniel Bethlehem.⁹⁹ Bethlehem declares that the Principles ‘do not reflect the settled view of any state’ and are ‘published under [his] responsibility alone’.¹⁰⁰ He states, however, that they have ‘been informed by detailed discussions over recent years with foreign ministry, defense ministry, and military legal advisers from a number of states who have operational experience in these matters’.¹⁰¹ Given the views expressed above about the significance of geopolitical considerations, it will come as no surprise that the source of these documents – that is the geographical representativeness of its drafters – is an issue of concern for me.

The Chatham House Principles, while acknowledging that ‘the law in this area is politically and legally contentious’, declare that the Principles ‘are intended to give a clear representation of the current principles and rules of international law’.¹⁰² Principle 6 of the Chatham House Principles states that ‘Article 51 is not confined to self-defence in response to attacks by States’ and that the ‘right to self-defence applies also to attacks by non-State actors.’¹⁰³ The commentary to Principle 6 asserts that there ‘is no reason to limit a state’s right to protect itself to an attack by another state’. In the view of the drafters, the ‘source of the attack, whether state or non-state actor, is irrelevant to the right’. In particular, the commentary states that the ‘right to use of force in self-defence is an inherent right and is not dependent upon any prior breach of international law by the state in the territory of which defensive force is used’. This would suggest that attribution of the conduct of non-State actors to the territorial State is not a necessary condition for the exercise of force in self-defence. The commentary to the Chatham House Principles attempts to mitigate the implications of Principle 6 by stating that while normal rules, such as proportionality and necessity, apply to the application of the Principles, special considerations are relevant in cases where the State on whose territory force is being used is not responsible for the armed attack. In particular, the commentary to Principle 6 states that, more so than with respect to an attack by a State, the attack by non-State actors ‘must be large-scale’ to qualify as an armed attack for the purposes of Article 51.¹⁰⁴ This mitigation amounts to the same fact or circumstance specific approach that I earlier rejected.

The use of force against terrorists is dealt with in Part II of the Leiden Recommendations.¹⁰⁵ As with the Chatham House Principles, it assumes that the position it postulates on the use of force reflects current international law. The introduction to the policy recommendation on the use of force states the principal aim of the policy recommendations on the use of force against terrorists as being ‘to clarify the state of international law on the use of force against terrorists against the backdrop of recent practice’.¹⁰⁶ It declares that there ‘is no need for new rules’ in this area and that the ‘framework laid down in the Charter is the cornerstone of the international legal regime on the use of force’.¹⁰⁷ However, it states that the Charter ‘is not a static instrument’ and must be ‘interpreted in the light of practice and taking states’ expectations into account’.¹⁰⁸ The Recommendations state that the right of self-defence in Article 51 ‘makes no reference to the source of the armed attack’.¹⁰⁹ It then proceeds to state that it ‘is now well accepted that attacks by non-state actors, even when not acting on behalf of a state, can trigger a state’s right’ of self-defence.¹¹⁰ As with the Chatham House Principles, the Leiden Recommendations make some attempt to mitigate against the implications of permitting a wide reading of ‘armed attack’, by stating that only ‘large-scale attacks’ of the non-State actor can form the basis of the right to use force in self-defence.¹¹¹

Unlike the Leiden Recommendations and the Chatham House Principles, the Bethlehem Principles do not, at least not expressly, purport to represent the law as it stands. On the Principles’ relationship with law, Bethlehem states that they are ‘intended to work with the grain of the UN Charter as well as customary international law, in which resides the inherent right of self-defence’.¹¹² The stated objective of the Bethlehem Principles is, it seems, to stay clear of the doctrinal debate, which he says ‘has yet to produce a clear set of principles’, in favour of a more ‘strategic’ and ‘operational’ approach.¹¹³ He accepts that some of the principles he advances will ‘undoubtedly prove controversial’.¹¹⁴ It is interesting though, that the examples of those principles that are likely to prove controversial do not include the proposition that force may be used in the territory of a third State. Bethlehem offers, as examples of the more controversial principles, the question of what is meant by imminence and who may be ‘properly targeted within the non-state actor continuum of those planning, threatening, perpetrating, and providing material support to an armed attack’.¹¹⁵

While the Chatham House and Leiden Recommendations depart from the assumption that, as a general rule, a State can use force unilaterally in self-defence against non-State actors on the territory of a third State, the Bethlehem Principles move from the opposite premise. Principle 10 of the Principles state as follows:

Subject to the following paragraphs, a state may not take armed action in self-defense against a non-state actor in the territory or within the jurisdiction of another state (‘the third state’) without the consent of that state.

Principles 11 and 12 provide exceptions to the basic rule in Principle 10. Principle 11 states that the requirement for consent does not apply ‘in circumstances in which there is a reasonable basis for concluding that the third state is colluding with the non-state actor’ or in cases where the third State is ‘unwilling to effectively restrain’ the non-State actor. Principle 11, in its essence, removes the requirement of consent where there is some attribution¹¹⁶ of the armed attack by the non-State actor on the third State or, to put it another way, in cases where the third State is not innocent. Principle 12, in contrast, does apply to what we may term the ‘innocent State’. It provides that the requirement for consent in Principle 10 does not apply where ‘there is a reasonable and objective basis for concluding that the third State is unable to effectively restrain’ the non-State actors. There are of course factual questions that may be asked about this principle. For example, it is not clear who determines whether there is a ‘reasonable and objective basis’ for making a decision that the third State is unable to restrain the non-State actors. It is also not clear whether the criteria ‘unable to effectively restrain’ is one of result or conduct. If it is the former, then in most instances the third State will be deemed to be unable unless the third State is able to stamp out attacks from the non-State actor. At any rate, for the purposes of this contribution, the essential point is that, according to the Bethlehem Principles, like the Chatham House Principles and the Leiden Recommendations, a State that is a victim of an attack or a series of attacks from a non-State actor may launch military attacks in self-defence against that non-State actor on the territory of an innocent non-consenting third State if that third State is unable to restrain the non-State actor.

While the Bethlehem Principles have a different point of departure from the other documents, the basic tenets of the three sets of documents are the same. All three advance a proposition that permits, in principle, the use of force against non-State actors, in the territory of non-consenting third States on whom the armed activities of the non-State actors cannot be attributed – what this chapter has termed the ‘innocent State’. Moreover, all three documents suggest that this proposition reflects the law under the Charter and customary international law or, in the case of the Bethlehem Principles, is at least not inconsistent with it. It is also worth pointing out that these instruments offer scant legal justification for the proposition. The Leiden Recommendations simply state the proposition ‘is now well accepted’ and that Article 51 ‘makes no reference to the source of the attack’.¹¹⁷ The commentary to Principle 6 of the Chatham Principles does, however, offer some basis, albeit without much explanation. First, similar to the Leiden Recommendations, the commentary to Principle 6 states that there ‘is nothing in the text of Article 51’ to suggest that the right to self-defence applies only in the case of an armed attack from a State. It suggests that this conclusion can be supported with reference to the Caroline incident – the relevance of this incident is considered below. Second, it is suggested in the commentary to Principle 6 that ‘State practice in this field, including the recent practice of the Security Council, gives … support’ to the notion that self-defence is permissible in the territory of an innocent third State even if the attacks by non-State actors are not imputable to the State on whose territory force is to be used. Although the commentary to Principle 6 does not expand on these bases, they are, by and large, reflected in the literature supporting the proposition that force may be used in the territory of a non-consenting innocent State. I turn now to the literature in support of the central proposition advanced in these three documents.

C. The Proposition that Unilateral Force can be used Extraterritorially in Self-Defence against Non-State Actors

Several authors have sought to justify the proposition that, under international law as it stands, a State may use force in self-defence against a non-State actor in the territory of an innocent State without that State’s consent. A survey of the literature and the reasons advanced reveals, similar to the commentary to Principle 6 of the Chatham House Principles, that the proposition is based on three mutually reinforcing pillars. First, the proposition is based on a literal reading of Article 51. Second, the proposition is based on the idea that pre-Charter customary international law continues to exist side by side with the UN Charter and that under the rules established under customary international law, the use of force against non-State actors was (and still is) permissible. Finally, the literature suggests that post-Charter practice supports the proposition that force may be used against non-State actors in the territory of a non-consenting innocent State.

The literal interpretation pillar of the argument is based on the fact that the words in Article 51 of the Charter do not limit the right of self-defence to an armed attack from a State. Paust captures the argument succinctly as follows:

Article 51 of the Charter expressly affirms the right of a State to respond defensively ‘if an armed attack occurs’, and nothing in the language of Article 51 restricts the right to engage in self-defense measures to circumstances of armed attacks by a ‘state’.¹¹⁸

According to Murphy, ‘the “ordinary meaning” of the terms of Article 51 provide no basis for reading into the text a restriction on who the attacker must be.’¹¹⁹ But in addition to ‘ordinary meaning’, Murphy also advances what might be seen as a contextual reading of Article 51 as supporting the proposition that a State may unilaterally use force in the territory of innocent third States without such States’ consent. He suggests that since Article 2(4) of the Charter explicitly prohibits the use of force against ‘the territorial integrity or independence of any State’,¹²⁰ if the drafters of the Charter intended to limit Article 51 to inter-State relations that construct would be repeated in Article 51.¹²¹ Paust adds that any other interpretation would be ‘lacking in common sense’ since it would suggest that a State being attacked by a non-State actor may ‘only defend itself within its own borders’.¹²² This ordinary meaning argument underlying the proposition is also reflected in various ways in certain instruments.¹²³

The second argument based on customary international law also proceeds from the text of Article 51 of the Charter. The argument proceeds from the fact that the right to self-defence in Article 51 is described as ‘inherent’ and takes this to mean that the rules on self-defence under customary international law continue to apply notwithstanding the UN Charter. Murphy recalls that to fully understand the rule in Article 51 ‘requires looking beyond the language of that article’ because ‘Article 51 did not create a right on self-defence.’¹²⁴ Rather, he argues, Article 51 ‘preserved’ a right ‘that existed in customary international law’ even before the adoption of the Charter.¹²⁵ Authors raising this point in support of the broad interpretation of the ‘armed attack’ requirement – and therefore in support of a permissive approach to the unilateral use of force against non-State actors – invariably point to the Caroline incident as the foundation of the argument on customary international law.¹²⁶

The Caroline incident is described in many texts, so only a summary of the relevant facts is presented here.¹²⁷ In short, in the nineteenth century, while the territory of Canada was still under British control, insurgents trying to overthrow British rule often carried out attacks, sometimes from the territory of the United States. In 1837, the Caroline, a vessel used in one such attack, was destroyed by the United Kingdom while it was in US waters. The content of the exchange of letters between Daniel Webster, on behalf of the United States, and Lord Ashburton, on behalf of the United Kingdom, is what is referred to as establishing the law on self-defence in the Caroline incident.¹²⁸ In his letter to Webster offering an explanation for the United Kingdom’s use of force against the Caroline while it was on US waters, Ashburton declares that there are ‘possible cases in the relations of nations as of individuals, where necessity … may be pleaded’.¹²⁹ In response, Webster, while admitting the existence of self-defence, notes that it is ‘confined to cases in which the “necessity of that self-defence is instant, overwhelming, and leaving no choice of means, and no moment for deliberation”’.¹³⁰ The Caroline incident is more often than not raised in the context of the requirements of necessity and imminence. However, in the context of the proposition relating to the use of force against non-State actors, it is pointed out that Webster in his retort made no issue of the fact that the actors responsible for the attacks in British Canada were non-State actors. Nor is there any indication that the United Kingdom had imputed the acts on the United States government. Therefore, it is suggested, both the United States and the United Kingdom accepted that the use of force in self-defence against non-State actors in the territory of a third State was possible even if the attack by non-State actors were not imputable to that third State.¹³¹

The third pillar supporting the proposition is to refer to certain post-Article 51 State practice which, it is argued, supports the proposition that current international law permits a State to use force in self-defence against a non-State actor in the territory of an innocent third State without that third State’s consent. Explaining the central element of this pillar, Murphy states that ‘whatever the original meaning of Article 51 might have been, subsequent state practice appears to support’ the proposition that a State may use force against non-State actors in the territory of a third State.¹³² Similarly, van Steenberghe states that ‘recent practice greatly contributes to the evolution of the law of self-defence’ and ‘implies the recognition of the right to respond in self-defence to private armed attacks’.¹³³ The examples of State practice in these areas are too many to fully discuss or cover at any length. I will focus, therefore, only on those examples of practice common to all publications supporting the proposition that the use of force in self-defence against non-State actors in the territory of innocent third States is permissible. These can be usefully divided into two categories.

The first category involves practice following (and in response to) the attacks of 11 September 2001 in the United States.¹³⁴ These attacks on US territory were carried out by a non-State actor, Al-Qaeda. In the aftermath of those attacks, the United States notified the Security Council on 7 October 2001 that it would be deploying forces in Afghanistan in the exercise of its right to self-defence. It is now a historical fact that the United States did indeed attack Afghanistan in what was known as Operation Enduring Freedom. Tams, in other works, suggested that Operation Enduring Freedom provides ‘the most obvious piece of evidence’ to support the proposition that unilateral force may be used in self-defence against non-State actors in the territory of innocent non-consenting third States.¹³⁵ Similarly, Paust states categorically that the October 2001 use of force by the United States in Afghanistan against Al-Qaeda ‘was justified, and justifiable, as self-defence against on-going processes of armed attack on the United States … ’.¹³⁶ He clearly identifies the use of force against Al-Qaeda as an act of ‘self-defence’ and in response to an ‘armed attack’, thus bringing them within the ambit of Article 51.

Practice flowing from the 11 September 2001 attacks is also said to include the UN Security Council resolutions adopted in the aftermath of those terrorist attacks.¹³⁷ In the aftermath of the attacks, the UN Security Council adopted Resolutions 1368 and 1373 on 12 September and 28 September respectively. Both Resolutions include a preambular paragraph recognising ‘the inherent right of individual or collective self-defence in accordance with the Charter’. These references, it is said, amount to a recognition by the UN Security Council that unilateral force may be used in self-defence against non-State actors in the territory of a non-consenting innocent third State.¹³⁸ Explaining the position, Murphy states that there is no language in the Resolutions that indicates that the initial attack ‘must first be imputed to a State’ before the right to use force in self-defence can be triggered.¹³⁹

The second category of examples often advanced to justify the use of force in self-defence against non-State actors in the territory of non-consenting innocent third States involves other uses of force by States, claimed to be in self-defence, against non-State actors in the territory of third States.¹⁴⁰ The most common examples of such other cases include the use of force by Israel against Hezbollah in Lebanon in 2006 and the incursion by Turkey into northern Iraq in response to attacks by the Kurdistan Workers Party in 2006 and 2008 respectively.¹⁴¹ There are, of course, many other cases of use of force against non-State actors that have been considered, but time and space permit only the consideration of these two cases as illustrative of the discussion.¹⁴² Tams, for example, refers to the US bombardment of a pharmaceutical plant in Sudan in response to attacks on its embassies in Kenya and Tanzania, the Russian airstrikes in Georgia in response to Chechen rebel activity, and the Colombian use of force in 2008 against the FARC.¹⁴³ In this Trialogue, Tams (pp. 152–3) lists, without discussion, the following additional cases of ‘cross-border raids in pursuit of rebels’: Tajikistan in Afghanistan; Rwanda against the DRC; Senegal in Guinea Bissau; and Thailand in Myanmar. In her chapter in the Trialogue, O’Connell (pp. 176–7), for her part, identifies a series of US conduct, not so much as law-changing practice, but rather as examples of unlawful use of force. More recently, the use of force against ISIS provides a significant example of practice.¹⁴⁴

D. The Rules for Interpreting and Identifying the Scope of Self-Defence in Respect of Non-State Actors

1. The General Rules of Treaty Interpretation

Given that the rules on self-defence exist both under customary international law and the Charter, the question arises whether the proper analysis should focus on one or the other. In my view, because the content of the customary and charter rules on self-defence are co-extensive and identical, it makes little difference where the focus lies, and the outcome of the analysis will be the same.

The question, though, is how to construct the relationship between the prohibition in Article 2(4) and the right in Article 51 of the Charter. My co-participants in the Trialogue adopt largely divergent approaches. Tams, on the one hand, employs a positivist approach, seeking to find meaning in the text and the practice of States. O’Connell, on the other hand, adopts a historical account, based on the tenets of natural law, to her analysis and places the right of self-defence in the context of the important and fundamental nature of the prohibition on the use of force. Her position is that because the prohibition on the use force is ius cogens, the right to self-defence, as an exception to that prohibition, cannot be subject to the ordinary rules of interpretation, such as the application of subsequent practice in a way that would reduce the scope of the ius cogens norm.¹⁴⁵ I adopt the approach that the traditional rules of interpretation, including practice, provide the path for finding the scope of self-defence.

The rules relevant to the interpretation of Article 51 are to be found in Article 31 of the VCLT. The primary rule of interpretation, found in Article 31(1) of the VCLT, is that words in a treaty must be given their ordinary meanings, in their context and in the light of the object and purpose of the treaty. The literal meaning of the words can, therefore, not be dispositive of the question whether a State can use force in self-defence against non-State actors in the territory of innocent non-consenting third States. The context in which those words are used, as well as the object and purpose of the treaty, give the text its meaning. Moreover, the VCLT provides for subsequent agreements and subsequent practice of the parties to a treaty to contribute to finding the meaning of the treaty provisions. Article 31(3)(a) provides that, in the interpretation of a treaty text, ‘subsequent agreement between the parties regarding the interpretation of the treaty’ shall be taken into account. Article 31(3)(b), on the other hand, provides for the taking into account of ‘subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’. Finally, Article 31(3)(c) provides for the taking into account of ‘relevant rules of international law applicable in the relations between the parties’.

As shown above (Section IV.C), the proposition that the right to use force in self-defence extends to force against non-State actors operating on the territory of an innocent non-consenting State is based on three pillars. First, that the word ‘inherent’ in Article 51 means that the rules of self-defence in customary international law continue to apply and that these customary international law rules permit the use of force in self-defence against non-State actors in the territory of innocent non-consenting States. The second pillar is that the phrase ‘armed attack’ in Article 51 is not qualified by the requirement that the armed attack must emanate from a State. Finally, this proposition is based on post-Article 51 practice that is said to support the idea that the use of force in self-defence against non-State actors in the territory of innocent third States is permitted.

All three pillars, in some way, depend on an assessment of practice. The pillar based on the word ‘inherent’ seeks to invoke customary international law, in particular practice in the form of, inter alia, the Caroline incident, to support the proposition. The pillar based on the phrase ‘armed attack’, though essentially advancing a literal interpretation of the Charter, must also, out of necessity, rely on practice. Practice would be important to establish the meaning of the phrase ‘armed attack’. Finally, and most obviously, the third pillar on post-Article 51 practice is, of course, based on practice in the form of post-Article 51 practice. Given the centrality of practice for the proposition, a brief description of the role of practice for the determination of the rules of international law on self-defence is appropriate before embarking on an evaluation of the three pillars of the proposition.

2. The Role of Practice in Evaluating the Proposition

Practice can, potentially, serve three roles in the evaluation of the proposition that international law permits the unilateral use of force in self-defence in the territory of an innocent non-consenting third State. First, practice can be advanced in support of a rule of customary international law, existing independent of (perhaps even in competition with) Article 51 of the Charter. Second, practice can be advanced in support of an existing rule of customary international law to be considered in the interpretation of Article 51 of the Charter under Article 31(3)(c) of the VCLT. Finally, practice can be advanced as subsequent practice in the interpretation of Article 51 under Article 31(3)(b) of the VCLT. To fulfil these respective roles, however, the relevant practice must meet certain requirements.

For practice to contribute to the formation of customary international law, it has to be widespread or general and must be settled.¹⁴⁶ In 2018, the ILC adopted a set of Draft Conclusions on the Identification of Customary International Law on first reading.¹⁴⁷ According to Draft Conclusion 8, the ‘relevant practice must be general, meaning that it must be sufficiently widespread and representative, as well as consistent’.¹⁴⁸ Thus, in instances where the practice is limited to a few States, condonation by other States, if sufficiently wide, might be sufficient to establish that there is widespread practice. Conversely, widespread objection to or criticism of a practice would prevent the establishment of widespread practice for the purposes of formation of customary international law. It is worth pointing out in this regard that the ILC, in its Draft Conclusions on the Identification of Customary International Law, considered, and decided not to include, a provision to the effect that the practice of ‘specially affected States’ is more significant in the establishment of customary international law.¹⁴⁹ The attitude of the ‘silent majority’ – those States that neither engage in a practice nor respond to it – is thus of particular importance. Inaction on its own cannot, as a rule, be taken to mean acquiescence to a practice.¹⁵⁰ It has to be shown that inaction was deliberate and aimed at expressing acquiescence.

In addition to the above, for practice to form the basis of a rule of customary international law, that practice must be accepted by the community of States as a whole as a rule of law – opinio iuris sive necessitas.¹⁵¹ Whether evidence of opinio iuris is sufficient to meet this standard requires not only the belief of a few States that the practice constitutes a rule of customary international law, but this belief must also be widely held. The question of the silent majority is again important here. In the commentary to Draft Conclusion 9 of the Draft Conclusion on the Identification of Customary International Law, the ILC observed that the requirement of acceptance of law ‘is to be sought with respect to both the States engaging in the relevant practice (action or inaction) and those in a position to react to it’.¹⁵² More to the point, paragraph 3 of Draft Conclusion 10 states that silence will only constitute opinio iuris if the States ‘were in a position to react and the circumstances called for some reaction’. Thus, international law as it currently stands requires widespread practice and acceptance as law. For these purposes, it is important, particularly where the practice is limited in extent, to also consider the reactions of other States to that practice.

Similarly, there are important requirements for reliance on subsequent practice as a tool for interpreting Article 51 of the Charter. The VCLT provides that subsequent practice ‘shall be considered’ in the interpretation of a treaty provision. However, it is only practice which ‘establishes the agreement of the parties regarding [the treaty’s] interpretation’ that is to be considered under Article 31(3)(b). Thus, in order to constitute an authentic interpretation of parties under Article 31(3)(b) of the VCLT, it is not sufficient to show widespread agreement. Rather, it must be shown that the practice in question establishes the agreement of all the parties to the UN Charter as to the interpretation of Article 51. Moreover, the practice must establish the agreement of all the parties regarding the interpretation of Article 51. Political support of particular conduct in which force has been used is insufficient; the support should be linked to the interpretation of Article 51. Thus, even more than with the formation of customary international law, the attitude of the silent majority is an essential element of any attempt at establishing that a particular conduct, or series of actions, constitutes subsequent practice for the purposes of Article 31(3)(b) of the VCLT. More to the point, any expression of criticism or objection will most certainly inflict a deathblow on any attempt to advance conduct, or series of actions, as constitutive of subsequent practice for the purposes of Article 31(3)(b) of the VCLT.

Thus, in assessing practice, whether for the purposes of customary international law or treaty interpretation, it is important to also assess the response of other States. Finally, it may also be pointed out in this respect that judicial practice, particularly that of the International Court of Justice – though not State practice – may also serve as an important element for weighing the evidence for an interpretation of rules of international law, including the weight of certain conduct as practice, either for the purposes of customary international law or as an authentic interpretation of Article 51 of the Charter under Article 31(3)(b) of the VCLT.¹⁵³

V. Evaluation of the Scope of the Right of Self-Defence

I proceed now to consider each of the three pillars on which the proposition that States may use force unilaterally in self-defence against non-State actors in the territory of an innocent third State without that State’s consent rests in turn. The main thrust of the first pillar is dependent on the identification of customary international law before the Charter and not on the interpretation of the Charter provisions. The second and third pillars both rely on the interpretation of the provisions of the Charter concerning the use of force.

A. The ‘Inherent Right’: Pre-Existing Rules of Customary International Law

The cogency of the first pillar – pre-Charter customary international law permitted the use of force against non-State actors – depends on several assumptions. The first, and least controversial assumption, is that a rule of customary international law continues to exist even if embodied in a treaty. First, this assumption seems to be borne out by the text of Article 51 of the Charter, which refers to the ‘inherent right’ of self-defence, suggesting that the right exists independent of the Charter. Second, in connection with the right to self-defence, the ICJ has recognised this principle stating that there ‘can be no doubt that the issues of the use of force and collective self-defence … are issues which are regulated both by customary international law and by treaties, in particular the United Nations Charter’.¹⁵⁴

A second assumption necessary for the cogency of the argument based on the pre-Charter customary international law is that the exchange between Ashburton and Webster in the Caroline incident constituted customary international law. However, this assumption is based on flawed logic. For the exchange in the Caroline incident to be formative (or even reflective of a rule of customary international law), the resort to force had to be prohibited under international law at the time of the exchange. Self-defence, as an exception under international law, can only make sense where international law prohibits the resort to force. Yet, as observed in Section II.A above, at the time of the Caroline incident international law did not prohibit the use of force. The justifications advanced by Ashburton for the attack on the Caroline appear, therefore, to be for ‘political expediency’ and ‘in order to secure moral high ground’ and not to provide a shield against legal wrongfulness.¹⁵⁵ If the use of force was not prohibited under international law and the justifications offered by Ashburton were not intended as legal justification, then the Caroline incident cannot be said to have been constitutive, or reflective, of customary international law. This is not to say that it could not have inspired, at a later date, the establishment of specific rules of customary international law on the right to self-defence. However, any such rules would need to be established through the normal process of the formation of customary international law and not through some nostalgic recollection of some incident that took place before the development of the law on self-defence. In other words, if the Caroline incident inspired the establishment of a new rule of customary international law, it would be necessary to show the existence of such a rule through evidence of the practice and opinio iuris inspired by the exchange.

A third assumption underlying the pre-Charter customary international law argument is that a single incident, involving only two States, can generate a rule of customary international law. It is not clear how the Caroline incident meets the requirements of a widespread or general practice. Those advancing it as constitutive of customary international law make no attempt to show acquiescence on the part of other States. The Caroline incident does not meet the generally accepted criteria for customary international law. As most recently captured by the ILC: ‘the practice must be followed by a sufficiently large and representative number of States.’¹⁵⁶ Moreover, even if it could be said that the exchange between Ashburton and Webster constituted widespread, representative and consistent practice, no evidence of acceptance of law has been illustrated by those relying on the exchange to suggest a rule of customary international law. Quite the contrary if, as illustrated in Section II.A, international law did not prohibit the use of force, then the exchange could not be evidence of an acceptance of law. Indeed it is not even clear that the authors of the exchange themselves consider themselves to be stating rules of international law.¹⁵⁷

A fourth assumption underlying the argument that pre-Charter customary international law permitted the use of force in self-defence against non-State actors is that customary international law is static, and an exchange which may have reflected (or formed) a rule of customary international law in the period around 1840 when the incident occurred would still reflect customary international law of the twenty-first century. This assumption seems to ignore the dynamic nature of customary international law and that, whatever the law may have been at the time of the incident, world events since then have impacted on customary international law on the use of force so much so that to refer to an exchange of letters in 1842 as definitively reflecting rules of customary international law is highly questionable.¹⁵⁸

The argument that pre-Charter international law permits the unilateral use of force in self-defence against non-State actors on the territory of a non-consenting innocent State is based on four assumptions about customary international law, three of which are erroneous. The standard in the Caroline incident could, of course, reflect a rule of customary international law if it can be shown to meet the normal requirements for the formation of customary international law. It must thus be assessed in the light of practice, in particular developments in the twentieth and twenty-first century.

B. Armed Attack does not mean Armed Attack by a State

1. A (fairly) Consistent Line of Reasoning by the International Court of Justice

Before engaging with the interpretation of the words ‘armed attack’ in Article 51, it is useful to describe what appears to be a consistent line of reasoning from the International Court of Justice concerning the meaning of those words. Judicial practice, particularly from the ICJ, may be useful as a subsidiary means of determining the appropriate interpretation of Article 51. Over the years there have been several judgments, advisory opinions and individual opinions of members of the Court that have shed light on the proper interpretation of the rules of Article 51 and the corresponding rule of customary international law. In general, it may be said that the Court has, over the years, interpreted the phrase ‘armed attack’ as meaning an attack by one State against another.

It should be said, however, that the proposition that an armed attack from a non-State may permit the attacked State to use force against that non-State actor in the territory of a third State has received some support in separate and dissenting opinions of some members of the Court. Although Judge Schwebel’s dissenting opinion is, at least insofar as concerns the meaning of ‘armed attack’, a little less generous than the proposition, he does, based on the views on Sir Humphrey Waldock’s Hague Lectures, conclude that the use of force in self-defence is permitted even in cases not involving armed attack – a view even broader than the proposition.¹⁵⁹ More relevant for the present discussion, Judge Buergenthal, in his declaration in the Construction of the Wall opinion, declares that the United Nations Charter in Article 51 does not make the exercise of the inherent right to self-defence ‘dependent upon an armed attack by another State’.¹⁶⁰ Similarly, Judge Higgins, in her separate opinion, noted that there is ‘nothing in the text of Article 51 that thus stipulates that self-defence is available only when an armed attack is made by a State’.¹⁶¹

The Court itself, however, has been fairly consistent in rejecting the proposition that a State may unilaterally use force in the territory of an innocent third State without that State’s consent. This consistent line of reasoning can be traced back to the Nicaragua case. There the Court states that for a State to rely on self-defence under international law, that State must be a victim of an armed attack.¹⁶² In addressing what constitutes an armed attack, the Court famously makes the following statement, which has been the subject of much discussion:

There appears now to be general agreement on the nature of the acts which can be treated as constituting armed attacks. In particular, it may be considered to be agreed that an armed attack must be understood as including not merely action by regular armed forces across an international border, but also ‘the sending by or on behalf of a State of armed bands, groups, irregulars, or mercenaries, which carry out acts of armed force against another State of such gravity as to amount to’ an actual armed attack conducted by regular forces, ‘or its substantial involvement therein’. This description … may be taken to reflect customary international law.¹⁶³

It is clear that the Court considers that, as a general rule, an armed attack means an attack by one State against another. However, the Court considers that it is possible for an armed attack, justifying the use of force in self-defence, to occur in the case of attacks by non-State actors. However, for this possibility, the acts must be arising from the ‘sending by’ or must be ‘on behalf of [the third] State’.¹⁶⁴ In other words, the acts in question must be imputable to a State to constitute an armed attack capable of forming the basis of unilateral use of force in self-defence. The Court, moreover, continues to significantly reduce the possibility of imputability. The Court states that the notion of armed attack does not include cases where acts of non-State actors result from ‘assistance to [non-State actors] in the form of the provision of weapons or logistical or other support’.¹⁶⁵ For the Court, therefore, armed attacks by non-State actors, will, for the purposes of the law on self-defence, be imputable to a State if that State exercises effective control over the non-State actors.

The basic principle, articulated in the Nicragua case has been reaffirmed in other judgments and opinions of the International Court of Justice. Although the issue of the author of the ‘armed attack’ was not central in the Oil Platforms case, the Court in that case also contributed to this consistent line of jurisprudence when it noted that for the United States to successfully claim self-defence against Iran, it had to show that ‘attacks had been made upon it for which Iran was responsible’.¹⁶⁶ The issue was also addressed in the Wall advisory opinion, in particular on the question of whether Israel was entitled to rely on Article 51 in response to attacks from non-State actors in the Occupied Palestinian Territories. In the Wall advisory opinion, the Court confirmed that Article 51 of the Charter applies to cases of an ‘armed attack by one State against another’.¹⁶⁷ The language in these cases is clear and unambiguous.

The Armed Activities in the Territory of the Congo case was seen, by those supporting the proposition that States can use force unilaterally in self-defence in the territory of third States, as an opportunity to ‘correct’ the judgments in the Nicaragua, Oil Platforms and Wall jurisprudence.¹⁶⁸ The Court, however, confirmed the fundamental element in its reasoning in these decisions, namely that the use of force in self-defence was impermissible against non-State actors on the territory of a non-consenting, third State. Dismissing Uganda’s claim of self-defence, the Court observed that Uganda ‘did not ever claim that it had been subjected to an armed attack by the armed forces of the DRC’.¹⁶⁹ The Court notes that, rather, the armed attack which was purportedly the basis of the unilateral use of force response was from a non-State actor – the Allied Democratic Forces (ADF).¹⁷⁰ Therefore, having found ‘no satisfactory proof of the involvement in these attacks, direct or indirect, of the Government of the DRC’, or that the attacks emanated ‘from armed bands or irregulars sent by the DRC or on behalf of the DRC’, the attacks by the ADF were not attributable to the DRC and therefore could not form the basis of unilateral use of force in self-defence in the territory of the DRC.¹⁷¹ The Court, thus, clearly requires ‘proof of the involvement’ of the third State, whether direct or indirect, or proof that the non-State actors were ‘sent by’ or acted ‘on behalf of’ the State in question before self-defence would be permissible. A statement by the Court that it is unnecessary ‘to respond to the contentions of the Parties’ concerning ‘conditions under which … international law provides for a right of self-defence against large-scale attacks by irregular forces’ might suggest a departure from the strict attribution standard.¹⁷² This latter statement, however is hardly a declaration of law. Moreover, reliance on it ignores the fact that the Court makes this statement in the context of its finding that ‘the preconditions for the exercise of self-defence’ did not exist in that case.¹⁷³ These preconditions are precisely that the acts of the non-State actors ‘remained non-attributable to the DRC’.¹⁷⁴ But more than that, the quoted phrase is not a doctrinal statement about law but is rather, on its face, concerned with the fact/circumstance specific issues that I have already deemed beyond the scope of this contribution.

It has also been suggested by Trapp that the decision of the Court in Armed Activities in the Territory of the Congo can be read as not requiring attribution for reliance on the right to use force on the territory of a third State.¹⁷⁵ This conclusion is based on an interesting, yet ultimately misleading, reading of the Court’s decision in the Armed Activities in the Territory of the Congo case. The conclusion highlights that in Armed Activities in the Territory of Congo ‘the Court emphasized that Uganda’s defensive measures were carried out against the DRC.’¹⁷⁶ This argument suggests that had Uganda’s use of force been directed at the ADF and not the DRC, then the Court would have found such unilateral use of force, in principle and subject to other requirements, permissible. With respect, this interpretation is only possible if one ignores a number of important factors.

First, it ignores the fact that the decision of the Court in Armed Activities in the Territory of the Congo is not given in a vacuum but rather is part of a consistent line of reasoning in which the Court has defined ‘armed attack’ as either an attack emanating from a State or one which can be attributed to a State. This definition in a consistent line of findings by the Court would be inconsistent with the novel interpretation advanced by Trapp, and it is inconceivable that the Court would make such a significant departure from its previously declared position in such a furtive manner. Second, this interesting reading of the Court’s decision ignores the fact that the Court is fully aware of the criticism from some quarters – mainly ‘Anglo-Saxon scholars’¹⁷⁷ – that has been levelled against its definition of armed attack. If the Court wanted to depart from its jurisprudence in response to this criticism it would have done so explicitly. Third, this novel interpretation would ignore the judicial dialogue involved in the International Court of Justice’s deliberative process. Several judges wrote separate opinions in the Armed Activities case, in particular Simma and Kooijmans, specifically criticising the Court for sticking with its Military and Paramilitary Activities case definition of armed attack.¹⁷⁸ The Court was therefore well aware that its dicta in Armed Activities in the Territory of the Congo case would be seen as endorsing its previous decisions, and it continued in that direction. Sean Murphy, in the context of the Wall opinion, also considers the possibility that the Court had departed from its position in the Military and Paramilitary case and correctly observes that assuming that the views of the judges writing separate opinions ‘were made known to the Court in the course of the deliberations, as would normally be the case, the Court understood how this language was going to be interpreted’ and it nonetheless ‘chose to adopt that interpretation’.¹⁷⁹ If there were any ambiguity in the Court’s decision – and I do not believe there is – such an ambiguity would have to be seen in the light of previous judgments. Given the practice of the Court of following its previous judgments, it has to be assumed that it has done so also in Armed Activities in the Territory of the Congo case. The correct description of the Court’s position after the Wall opinion, Oil Platforms case and the Armed Activities case seems to be neatly summarised by Tom Ruys, who observes that in these cases and the advisory opinion, the Court ‘sticks to the restrictive threshold spelled out in Nicaragua’.¹⁸⁰

It is worth noting that attribution of the armed attack, as the basis for unilateral use of force in self-defence on the territory of a third State, was not questioned, even in Judge Schwebel’s strongly worded dissent in Nicaragua.¹⁸¹ In his dissent, Judge Schwebel states that the Court’s conclusion ‘that a State’s assistance to foreign armed irregulars who seek to overthrow the government of another State’ does not amount to an armed attack ‘is inconsistent with generally accepted doctrine, law and practice as well’.¹⁸² Judge Schwebel’s conclusion in the Nicaragua case that there was an armed attack necessitating the use of force in self-defence is based on his view that there was a ‘pervasive and prolonged support by the Nicaraguan Government of the insurgencies in El Salvador’ which has ‘been a major, perhaps the critical factor in the transformation of what … were largely sporadic if serious acts of insurgent terrorism into an organized and effective army of guerrillas’.¹⁸³ It is clear that in his view, aiding or supporting, including through the provision of weapons or logistical support, or the provision of a safe haven, may constitute the basis of attribution. Thus attribution as such was not questioned in his dissent. What he did question was the standard required by the majority, which he deemed unreasonably high. This view was similarly advanced by Judge ad hoc Kateka in the Armed Activities in the Territory of the Congo case.¹⁸⁴ Similarly Judge Koroma makes a distinction between two scenarios, namely a ‘State’s massive support for armed groups, including deliberately allowing them access to its territory’ and a ‘State’s enabling groups of this type of act against another State’.¹⁸⁵ The former, he claims, constitutes an armed attack while the latter may not. While Judge Koroma suggests that this distinction is consistent with the standard set forth in Nicaragua, there is a significant difference. In the Nicaragua case the Court expressly excluded assistance and support from the ambit of armed attack, while Judge Koroma expressly includes assistance if it is ‘massive’ and done with deliberate intent to allow access. Moreover, even the second category of Judge Koroma should not be automatically excluded from the scope of armed attack. In placing this distinction in context, Judge Koroma states that ‘if a State is powerless to put an end to the armed activities of rebel groups despite the fact that it opposes them, that is not tantamount to’ an armed attack.¹⁸⁶ Clearly, then, Judge Koroma is excluding only innocent States from the ambit of States in whose territory force in self-defence may not be used.

In previous works, our co-participant in the Trialogue, Christian Tams, has similarly called for something between ‘strict adherence’ to attribution and ‘a radical departure’ from the Nicaragua test.¹⁸⁷ He suggests discarding the ‘effective control’ test of Nicaragua to a rule permitting the use of force in self-defence in the territory of a third State if that third State has supported or provided a safe haven for non-State actors engaged in acts of terrorism.¹⁸⁸ In many ways, this approach appears similar to the language of the US National Security Strategy of 2002, which declares that the United States will not distinguish ‘between terrorists and those who knowingly harbour or provide aid to them’¹⁸⁹ – it is one that is still based on some attribution, even if not rising to the level of the Nicaragua test. Dismissing this approach, Trapp states that this approach is ‘a long-rejected basis for attributing non-State conduct to a state’.¹⁹⁰ Advancing her test, which she seems to posit as more in line with an inter-State reading of Article 51, she suggests that if the third State ‘is doing everything possible to prevent its territory from being used as a base for terrorist operations’ then the use of force in self-defence in the territory of that State is not permissible.¹⁹¹ If the third State is not doing everything possible, then unilateral force in self-defence on its territory is permitted. There is an internal inconsistency in the approach. On the one hand, according to Trapp, assisting and supporting is too loose a standard for attribution – after all, that is the principal reason why it is ‘long-rejected’ – while on the other hand an even looser standard of attribution, namely ‘not doing everything possible’ is advocated.

The approach of the Court, which clearly rejects that force can be used in the territory of a third State in response to attacks from non-State actors, has been fairly consistent. The Court’s interpretation, particularly since it has not met with widespread objection from other States, should be highly persuasive in assessing whether an armed attack includes an attack from non-State actors. More importantly, the Court’s approach is consistent with the interpretation of Article 51 under the VCLT.

2. Interpretation According to Article 31 of the VCLT

The second pillar on which the proposition that the unilateral use of force in self-defence against non-State actors in the territory of an innocent third State is permitted under international law rests is the text of Article 51, which provides that the right of self-defence may be exercised in the event that ‘an armed attack’ occurs. It is pointed out by those supporting the proposition that Article 51 does not provide that it has to be an armed attack by ‘a State’. At first blush, this argument seems convincing. Treaty text however, as Article 51 is, cannot be interpreted only with reference to the literal meaning of the terms. The basic rule of treaty interpretation is that the terms of a treaty must be given their ordinary meaning, in their context and in the light of the object and purpose of the treaty.

We can find the context of Article 51 in several places in the Charter. The Charter as a whole forms part of the context within which Article 51 ought to be interpreted. In this regard, the fact that Article 2(4) and Article 51, ‘are very much a response to the Second World War and are accordingly directed to inter-state conflict’, is a contextual factor that suggests that Article 51 applies to an armed attack from a State.¹⁹² In this regard, Article 51 is an exception to Article 2(4). Article 2(4) is concerned with inter-State use of force. As pointed out in Section IV.B, if Article 2(4) is concerned with inter-State use of force, it would be expected that Article 51, the exception thereto, would cover the same terrain. More to the point, the broad interpretation of ‘armed attack’ to include attacks from non-State actors with no State involvement would significantly reduce the scope of the prohibition on the use of force in the Charter. After all, it should be recalled that the right to use force in self-defence is an exception to the prohibition on the use of force. Exceptions should not be interpreted so broadly that they render meaningless the primary rule, in this case the prohibition on the use force.¹⁹³ As Randelzhofer put it, while the UN Charter did not ‘intend to exclude self-defence entirely’, it ‘restricted its scope considerably’.¹⁹⁴ The proposition that self-defence may be used against non-State actors in the territory of innocent third States would significantly expand, not restrict, the scope of the self-defence and, as a consequence, significantly restrict the prohibition on the use of force.

Moreover, if Article 2(4) applied to use of force between States, while Article 51 permitted the use of force against non-State actors in the territory of third States, this would result in the anomalous situation that acting consistently with Article 51 would result in a violation of Article 2(4) since, by definition, any use of force in the territory of a third State would infringe the territorial integrity of that State. The commentary to the Chatham Principles is illuminating in this respect. It states that the ‘right to use of force in self-defence is an inherent right and is not dependent upon any prior breach of international law by the state in the territory of which defensive force is used’.¹⁹⁵ This reasoning seems to assume that somehow a non-State actor on the territory of a third State can be targeted without violating the territorial integrity and sovereignty of the third State. But of course that is not possible. Moreover, this seems to ignore that Article 51 does not exist in a vacuum but rather in the context of other principles of international law which are also contained in the Charter. How any interpretation of Article 51 interacts with and impacts on these other principles is important for determining the scope and limits of the right to self-defence.¹⁹⁶ These principles, apart from the prohibition on the use of force itself, include the respect for the territorial integrity and sovereignty of other States. It should be recalled, in this regard, that the use of force in the territory of another State, even if minimal, implies a violation of the third State’s territorial integrity and sovereignty. These principles of sovereignty and the respect for territorial integrity remain central and foundational to the UN system and international law and are held as fundamental by the vast majority of the international community.¹⁹⁷ To the extent that we have, in recent decades, seen the erosion of these principles, this has been to prevent States from using the principles to harm their own peoples. Indeed, world leaders have, in the Declaration of the High Level Meeting of the General Assembly on the Rule of Law at the National and International Level, rededicated themselves to:

support all efforts to uphold the sovereign equality of all States, to respect their territorial integrity and political independence, to refrain in [our] international relations from the threat or use of force in any manner inconsistent [with the Charter].¹⁹⁸

An interpretation of Article 51 that permits the use of force in self-defence against the territory of an innocent non-consenting third State would be countenancing the violation of the third State’s sovereignty and territorial integrity and would significantly impair these principles. Self-defence, when permitted under international law, amounts to an act of self-help in response to ‘a violation of rights … against the state responsible for the violation’.¹⁹⁹ To hold otherwise would impair the principles of territorial integrity and sovereignty. A contextual interpretation of the words ‘armed attack’ in Article 51 should take into account these principles of territorial integrity and sovereignty. This context, in my view, supports an inter-State reading of Article 51 of the Charter.

In his chapter in the Trialogue, Tams advances an alternative contextual factor that he describes as ‘more powerful’, namely Articles 39 and 42.²⁰⁰ He suggests that because Articles 39 and 42 have, through interpretation and practice, shed the original ‘inter-State construction’, Article 51 should be interpreted in a similar fashion. With respect, the argument simply does not follow. The expansive reading of Article 39 and 42 has been facilitated in part by the special powers of the Council and its primary mandate for the maintenance of international peace and security, powers and a mandate that States, acting unilaterally, simply do not have. To suggest, as this contextual interpretation might imply, that the scope of the rights of an individual State under Article 51 is the same as or even comparable to the scope of the powers of the Security Council in Chapter VII is wrong and dangerous. As Corten notes, to confer ‘on every state the power to implement unilaterally its own conception of the necessities of the war against terror’ would be to bypass ‘if not simply ignore’ the core ideal of collective security.²⁰¹

Tams, also in this Trialogue, has suggested that object and purpose are less helpful for the interpretation of Article 51 because teleological arguments ‘simply do not offer compelling arguments either way’.²⁰² However, the object and purpose of the Charter also, in my view, suggest a restrictive interpretation of the phrase ‘armed attack’. The object and purpose of the Charter, simply put, are the prevention of wars and conflict. Interpreting Article 51 to permit the use of inter-State force in the territory of another State – thereby violating the third State’s territorial integrity – is contrary to this purpose. It promotes rather than prevents war. Presumably, a third State on whose territory force is used would feel that the ‘act of self-defence’ against the non-State actor on its territory is a violation of the protection afforded to it by Article 2(4) potentially justifying (counter) resort to Article 51. This is precisely what the UN Charter seeks to prevent. Rather – and this contributes to both context and object and purpose – the Charter promotes a collective approach to peace and security. It is worth repeating that the very first purposes of the United Nations mentioned in the UN Charter are, firstly, to ‘maintain international peace and security’ and, secondly, to do so by taking ‘effective collective measures for the prevention and removal of threats to peace’. Unilateral action which the broad interpretation of ‘armed attack’ facilitates has a serious potential for escalation and therefore undermines this objective.

Tams scoffs at this, suggesting that it is ‘too good to be true’.²⁰³ He suggests that this is because, like many treaties, the Charter does not have ‘one single, undiluted object and purpose’ – it rather ‘seeks to integrate competing goals into one overarching framework’.²⁰⁴ Yet he offers no evidence of an object and purpose conflicting with the object and purpose to remove threats to peace through collective action. The reference in his contribution of the Trialogue to the changing nature of conflict, supported by the Secretary-General’s report, In Larger Freedom, in no way diminishes or contradicts this object and purpose. The right to self-defence is a provision in the Charter – an exception at that – and is not, as appears to be suggested by Tams, an object and purpose.

Finally, until recently, the undisputed understanding of Article 51 was that it applied to a right of a State to respond to an armed attack from another State.²⁰⁵ That this was the generally accepted view until recently seems to be accepted by Tams in his chapter where it is suggested that there has been a ‘gradual shift’ from the restrictive interpretation of ‘armed attack’ to a more expansive interpretation.²⁰⁶ This is significant for interpretation, because it suggests that if there has been a shift, it is for those who are advocating a shift to show either that the context of the words in the Charter and/or its object and purpose have changed to such an extent as to warrant a new meaning being ascribed to the words. A reliance on the literal meaning of the words ‘armed attack’ is simply inadequate because the words themselves have not changed. To illustrate, if in 1980 ‘armed attack’ was understood as ‘armed attack by a State’, shifting to a broader meaning would require more than pointing out that taken literally ‘armed attack’ does not exclude ‘armed attack from a non-State actor’. Something else would need to be advanced to show why these words have now acquired a new meaning. By the same token, it is inadequate to rely, as Murphy does,²⁰⁷ on the differences in the choices of words between Articles 2(4) and 51. After all, the language on which this argument is based has not changed from the days when ‘armed attack’ was understood as an ‘armed attack from a State’.

It would seem that the only plausible factor that could have resulted in a shift is subsequent practice by States, so that ultimately the fate of the proposition that a State may use force in self-defence against a non-State actor depends on whether subsequent practice, which is the third pillar of the argument, has been such as to be evidence of an agreement between members of the United Nations that ‘armed attack’ now applies to armed attack from non-State actors. It is this third pillar to which the chapter now turns.

3. Contemporary Practice regarding the Proposition that Unilateral Force against Non-State Actors in the Territory of Third States is Permissible

The third pillar on which the proposition is based is that there are examples in contemporary practice of States using force against non-State actors, in self-defence, in the territory of innocent third States without the latter’s consent. The examples mentioned in Section IV.C can be separated into two categories. The first category concerns practice emanating from and linked to the US response to the terrorist attacks on 11 September 2001. These, as discussed above, include the adoption of UN Security Council Resolutions 1368 and 1373 and the US war in Afghanistan. Examples of the second category include the Turkish incursion into northern Iraq in 2008 in response to attacks by the Kurdistan Workers Party, Israel’s bombardment of Lebanon in response to attacks by Hezbollah and the more recent use of force in Syria. I turn now to these examples of practice, beginning with the 9/11-related practice.

The 9/11-related practice – both the US intervention in Afghanistan and the Security Council Resolutions 1368 and 1373 – have been advanced as sufficient to change the course of law.²⁰⁸ Similarly, after noting that early on Article 51 was seen as being applicable to armed attacks by States, Hilpold declares that ‘the situation changed radically’ after 9/11 and that ‘[w]ithout doubt, Resolution 1368 … was a watershed.’²⁰⁹ Indeed, several individual opinions of the Court suggesting a departure from the Nicaragua test appear to indicate that the change in the law occurred as a consequence of the practice of the international community in response to the 9/11 attacks. This view can be observed in the approach of Judge Kooijmans in the Wall opinion.²¹⁰ He begins by noting that the statement by the majority that Article 51 applies ‘in the case of an armed attack by one State against another State … is undoubtedly correct’.²¹¹ However, he continues, Security Council Resolutions 1368 and 1373 have added ‘completely new elements’ which would permit the use of force in response to acts of terrorists ‘without ascribing these acts of terrorism to a particular State’.²¹² This new element, Judge Kooijmans observes, ‘marks undeniably a new approach to the concept of self-defence’.²¹³ Judge Simma, in his separate opinion in the Armed Activities in the Territory of the Congo case, viewed this permissive approach to be new under international law.²¹⁴ These positions, reflected in individual opinions of the Court, seem to accept that prior to 9/11, international law required attribution – using the strict standards set in Nicaragua – for a State to use force against non-State actors on the territory of third States.

The post 9/11 attacks have provided the most solid basis for departing from the test set by the International Court of Justice in the Nicaragua case largely because the US intervention in Afghanistan was undertaken with virtually no condemnation by other States.²¹⁵ It is possible, however, that the lack of condemnation, particularly of the initiation of the US attacks on Afghan territory, was more out of a sense of solidarity with the United States than any belief in the legality of the action.²¹⁶ However, for the purposes of this chapter, I will assume that the lack of response was not merely out of politeness. Nonetheless, a careful assessment of the 9/11-related practice does not support the expansive interpretation of self-defence.

I begin, first, by addressing the attacks by the United States in Afghanistan – the most direct form of practice. As mentioned above, these attacks were scarcely, if at all, condemned. Yet it is not at all clear that the United States was attacking what it saw as an innocent State, pouring serious doubt on the applicability of that intervention as an example of practice in support of the proposition. It is now not seriously questioned that the 9/11 attacks were committed by Al-Qaeda, a non-State actor, and that the US responded by launching attacks in Afghanistan. Yet speeches by the then US President, George W. Bush, suggest that, at the time, the United States, rightly or wrongly, did not view Afghanistan as an innocent State as we have defined it. In a speech on the day of the attacks, Bush said that the US ‘will make no distinction between the terrorists who committed these acts and those who harbour them’.²¹⁷ Nine days later, in an address to a joint sitting of Congress, he reaffirmed those comments, stating that by ‘aiding and abetting murder, the Taliban regime is committing murder’.²¹⁸ More to the point, in this speech, Bush went as far as to say ‘we condemn the Taliban regime’²¹⁹ – at the time the government of Afghanistan. Finally, emphasising that the Taliban itself would be targeted by US operations, Bush said the ‘Taliban must act and act now’ and hand over the terrorists ‘or they will share in their fate.’²²⁰ It is clear that the US, at the time of the discussion, saw Afghanistan as, to some extent, responsible for the attacks. There seems to be, thus, some form of attribution, at least in the view of the United States at the time. It is true that the standard of attribution is probably inconsistent with the standard put forward by the International Court of Justice in the Nicaragua case. The standard, rather, appears more akin to that advanced by Judge Schwebel in his dissenting opinion in the Nicaragua case. It will be recalled that Judge Schwebel criticised the Court’s decision to exclude ‘State’s assistance to foreign armed irregulars’ as a basis for permitting the use of force in self-defence in the territory of the third State.²²¹

The above analysis does not concern the lawfulness or unlawfulness of the use of force in Afghanistan, or the correctness or incorrectness of the standard of attribution applied by the United States in the case of Afghanistan. The analysis serves only to show that US attacks in Afghanistan, in response to the 9/11 attacks, do not constitute appropriate evidence of State practice in support of the use of force against non-State actors in the territory of a non-consenting innocent State. To relate this to the Bethlehem Principles, the events surrounding the use of force by the United States in Afghanistan in response to Al-Qaeda’s terrorist attacks in the United States on 9/11 are more closely explained in terms of Principle 11, i.e., where the third State is unwilling to act against the terrorist and is a ‘colluding or harbouring’ State.²²²

The UN Security Council resolutions that are often advanced to support the proposition, Resolutions 1368 and 1373, also do not, on proper construction, constitute evidence of practice supporting an expansive reading of the right to self-defence against non-State actors. First, these Resolutions simply, on their face, do not support the proposition. Both Resolutions include a preambular paragraph recognising the ‘inherent right of individual or collective self-defence’. Murphy is, of course, correct to state that ‘no language in the resolutions indicates that the Security Council believes that terrorist acts must first be imputed to a State so as to trigger the right to self-defence.’²²³ This view was also expressed in the declaration by Judge Buergenthal in the Construction of the Wall advisory opinion. In that case Judge Buergenthal stated that in ‘neither of these resolutions did the Security Council limit their application to terrorist attacks by State actors only, nor was an assumption to that effect implicit in these resolutions’.²²⁴ Nonetheless, it is also true that these resolutions do not state, or imply, that the inherent right of self-defence permits the use of force in the territory of a State where the original attack could not, in any way, be attributable to the third State. The resolutions contain a condemnation of the terrorist attacks and a reaffirmation of the right to self-defence. In the context of the present discussion they do nothing more than that. They do not suggest that the inherent right of self-defence is exercisable against non-State actors – indeed they do not mention non-State actors at all, certainly not in the context of the inherent right to self-defence. Second, these resolutions do provide some context that might be read as setting a standard for attribution. Resolution 1368, for example, states that those ‘responsible for aiding, supporting or harbouring the perpetrators, organisers and sponsors of these acts will be held accountable’.²²⁵ The preamble of Resolution 1373, for its part, recalls that ‘every State has the duty to refrain from organizing, instigating, assisting or participating in terrorist acts in another State …’ This does not mean that the resolutions set a new standard of attribution different from that set in Nicaragua. The simple point being made here is that the speculative arguments concerning the import of the preambular paragraph ignore that these provisions might have assumed the role of a State so that the reference to the inherent right of self-defence in the preamble could be linked to this possibility. This of course is also speculation. It suffices to say that we can learn nothing about the permissibility of the use of force against non-State actors in innocent third States from the provisions of these resolutions.²²⁶

It also bears mentioning that the judgments of the International Court of Justice in the Armed Activities in the Territory of the Congo case and the Oil Platforms case, as well as the Wall advisory opinion, were all delivered after 9/11 and yet all maintained the reasoning in the Military and Paramilitary Activities case. Against this background, the observations of Hilpold that ‘subsequent practice is inconclusive’ and that the international community ‘has demonstrated an extreme reluctance to depart from the existing cautious and restrictive approach in the interpretation and application of Article 51’ appear to be correct.²²⁷

Of the Turkish use of force in the territory of Iraq in 2006 and 2008, van Steenberghe asserts ‘most international reactions (sic) remained vague regarding the legal justification.’²²⁸ However, a cursory troll of news reports at the time suggests that most States and international actors questioned the legality of the use of force even though expressing sympathy for the Turkish security concerns. The EU, for example, through its then High Representative for Foreign Affairs, Javier Solana, expressed understanding for the concerns of Turkey but stated that its chosen course of action was ‘not the best response’.²²⁹ This statement by the EU High Representative was not just a political statement, or one concerned only with policy preferences. Confirming that the statement was one informed by the applicable legal principles, Solana explained that the ‘territorial integrity of Iraq [was] very important for’ the EU.²³⁰ Then United Nations Secretary-General, Ban Ki Moon, similarly said that while he was ‘conscious of Turkey’s concern’, he appealed for the ‘utmost restraint, and for respect of the border between Iraq and Turkey’.²³¹ Similarly, Stephen Smith, then Australian Foreign Minister, also called on ‘Turkey to respect Iraq’s Sovereignty and [to] retreat back over the border as soon as possible’.²³² A theme running through all of these statements is the acceptance of the seriousness of the attacks (by non-State actors) against Turkey, but a declaration that the attacks by the Kurdistan Workers Party cannot justify the violation of the territorial integrity of Iraq. Incidentally, Turkey itself never declared that it has the right to use force in self-defence against the Kurdistan Workers Party in Iraq for acts not attributable to Iraq.²³³ In the light of this reaction, it is hard to see how this example, often advanced as State practice in support of the proposition, can be seen either as reflective of customary international law or as constituting subsequent practice for the purposes of interpretation.

With respect to the use of force by Israel against Lebanon in 2006, van Steenberghe explains that many States explicitly recognised the right of Israel to use force to defend itself.²³⁴ The statements to which van Steenberghe refers were delivered in the course of a UN Security Council debate and can be easily accessed.²³⁵ Argentina, for example, while acknowledging Israel’s right to self-defence, insists that Israel should exercise the right ‘in accordance with the law’.²³⁶ Van Steenberghe suggests that this reference to international law was directed only at the elements of proportionality and necessity.²³⁷ Yet there is nothing in Argentina’s statement to indicate that its concern is directed only at proportionality and necessity. It is true that the statement referred to – ‘in particular, excessive force’ – suggests a link with proportionality. However, the statement also listed, separately, ‘the imposition of sea, air and land blockade’ and the ‘violations on the part of Israel along the Blue Line’ as concerns.²³⁸ These elements were not linked with the excessive use of force but rather concerned the use of force as such. The Russian statement similarly attacked both the proportionality and the use of force as such by describing the Israeli action as both ‘disproportionate and inappropriate use of force that threatens the sovereignty and territorial integrity of Lebanon’.²³⁹ Other statements were even clearer and more direct. The statement on behalf of Ghana declared that the use of force against ‘a sovereign state is unacceptable’.²⁴⁰ Even more explicit, Qatar recognised the right of Israel to self-defence, but nonetheless described Israeli attacks in Lebanon as a ‘flagrant violation of the Charter of the United Nations and all international custom’.²⁴¹ China used similarly strong terms, describing the attacks as an act of ‘armed aggression’.²⁴² It is true that in its statement, as with most other statements, there is reference to proportionality, but there is nothing in its statement to suggest its description of the act as ‘aggression’ is based solely on proportionality. Indeed, the statement by China identifies the violation of sovereignty as such as a major cause of concern.²⁴³ Given the criticisms of the use of force by Israel and, in particular, its characterisation as a violation of the UN Charter and an aggression, it cannot form the basis of subsequent practice for the purposes of interpretation of the UN Charter nor can it be regarded as practice constitutive or reflective of customary international law.

Other incidents, not fully discussed here, are similarly not reflective of customary international law, nor can they constitute subsequent practice, since they have been similarly criticised. The US attacks against a pharmaceutical factory in Sudan, in response to bombings of American embassies in Kenya and Tanzania, are a case in point. The Arab League, for example, declared the attack as ‘an attack on Sudan’s sovereignty’.²⁴⁴ Tams, in addition to the examples discussed above, refers to the Iranian attacks in Iraq against the Mujahedin-e Khalq Organisation, Russian attacks in Georgia against Chechen rebels as well as Colombian forces against FARC rebels in Ecuador.²⁴⁵ He concludes that for the most part these acts of force were met with ‘mixed reactions’ from third States.²⁴⁶ If Tams’ assessment is correct that the examples he cites were met with ‘mixed reactions’ then that already precludes their interpretative value as subsequent practice because they cannot be deemed as establishing the agreement of the members of the United Nations as to the interpretation of Article 51. Indeed, in the case of the Colombian use of force, the Permanent Council of the Organisation of American States determined that that action amounted to ‘a violation of the sovereignty and territorial integrity of Ecuador and of principles of international law’.²⁴⁷ Thus, there may well have been examples of States using force on the territory of another in response to attacks by non-State actors. However, given the at best ‘mixed’ and at worst ‘negative’ reaction, the acts cannot form the basis of relevant practice, either for the purposes of sustaining a customary international law rule permitting such use or as subsequent practice for the interpretation of Article 51.

The situation in Syria provides the most recent example of practice that may be advanced in support of the proposition for an expansive interpretation of the right to self-defence against non-State actors. That situation is a very complicated one, involving interrelated conflicts implicating a number of non-State actors and several States.²⁴⁸ The non-State actors include the Syrian anti-government forces, which itself consists of several separate groups, while the States involved include Syria, Iraq, Turkey, a US-led coalition of States and Russia. The Russian use of force in Syria (like the coalition use of force in Iraq) against ISIS is therefore excluded from the analysis below.

ISIS is a non-State actor that operates, in addition to other places, in Syria. It has been said to be responsible for terrorist attacks in several places, in particular in Europe, the United States and the Middle East. However, its acts are not attributable to Syria. Thus any use of force without the consent of Syria, against ISIS on the territory of Syria would fall squarely within the parameters of the proposition, namely the use of force against non-State actors in the territory of innocent, non-consenting third States. In addition to the United States, other States, like the United Kingdom, France, Belgium, Germany, Canada and Australia, have been engaged militarily in Syria against ISIS.²⁴⁹ While the number of States involved in the attacks in Syria is impressive, the question is whether this constitutes practice sufficient to advance a particular interpretation of Article 51. In addition to these acts of force undertaken by individual States, the UN Security Council passed a resolution, UN Security Council Resolution 2249, invoking language similar to language used when the Council authorises the use of force. This Resolution has already attracted some attention.²⁵⁰ What contribution, if any, do Resolution 2249 and the various acts by individual States operating in Syria make to the question on the use of force against non-State actors on the territory of third States?

It is useful to begin by reflecting on the key use of force-related language in Resolution 2249. As Wood notes, the language is complicated, indicating ‘tough negotiations’.²⁵¹ While it is true that the Council does not expressly state that it is acting under Chapter VII, this is, in the greater scheme of things, immaterial. What is important is the language used in operative paragraph 5. There the Resolution

Calls upon Member States that have the capacity to do so to take all necessary measures, in compliance with international law, in particular with the United Nations Charter, as well as international human rights … on the territory under the control of ISIL also known as Da’esh, in Syria and Iraq …

A few preliminary points about some of the language in the paragraph are necessary. ‘All necessary means’ or, in this case, ‘measures’ as a rule includes the use of force. The ‘territory under the control of [ISIS]’ includes Syria. The Resolution therefore ‘[c]alls upon States that have the capacity to do so’ to undertake military action in Syria against ISIS. However, the Resolution does not authorise the use of force. In UN language, ‘calls upon’ is, at best, encouragement. To authorise the use of force, the Resolution would use the verb ‘Decide’ or ‘Authorize’.²⁵² Nonetheless, it has been suggested that, though not authorising the use of force, Resolution 2249 provides ‘political support for military action’.²⁵³ Wood goes further, suggesting that the Resolution is not only politically important but that it is legally important.²⁵⁴ Even more pertinently for the purposes of the central question posed in this contribution, he suggests that ‘it is difficult to read the resolution otherwise than as an endorsement … of the use of force in self-defence against’ a non-State actor in the territory of a third State – in this case against ISIS in Syria.²⁵⁵

It is hard to argue that an encouragement to use force does not amount to the endorsement of such use of force. However, the Resolution does not encourage or endorse all uses of force against ISIS on the territory under the control of ISIS. The Resolution only encourages and endorses the use of force ‘in compliance with international law’. It is not clear why the Resolution cannot be read as encouraging the use of force with the consent of the territorial States in question, i.e. Iraq and Syria. If, as I have argued, international law at the time of the adoption of Resolution 2249 did not permit the use of force against non-State actors on the territory of innocent third States, then Resolution 2249 would have no impact on that law, nor would it endorse, even as an exceptional measure, the use of force without the consent of the territorial State unless the conduct of ISIS were attributable to that State. That would be the effect of the phrase ‘in compliance with international law’. The statement of Russia made it abundantly clear that Resolution 2249 was to have no impact on the current state of the law, stating that the Resolution was ‘a political appeal’ and did not ‘change … the legal principles underlying the fight against terrorism’.²⁵⁶ Of course, States have already relied on the Resolution in justifying the use of force on the territory of Syria.²⁵⁷ The Resolution itself, however, has no effect whatsoever on the law regarding the use of force against non-State actors in the territory of third States.

While Resolution 2249 did not, itself, authorise the use of force, many States have used force in Syria against ISIS without the consent of Syria. In determining whether this use of force is relevant for identifying the scope of self-defence, it should be noted that the relevance of this operational ‘conduct on the ground’ depends not only on reactions to it, but also on the justifications for such use of force.²⁵⁸ It is only if the justifications advanced are consistent that we can say that the practice ‘establishes the agreement of the’ States ‘as to the interpretation of’ Article 51 of the UN Charter.²⁵⁹ For surely, if the various States advance different bases for their conduct, even if the conduct is the same, it cannot be evidence of established agreement as to the meaning of Article 51 and therefore cannot affect the interpretation of Article 51.²⁶⁰

It is clear that all the States using force in Syria adopt the position that taking such action is lawful. Some of the States operating there, it seems, have adopted Bethlehem’s ‘unwilling or unable’ test.²⁶¹ The United States, for example, declares that ‘when … the government of the territory where the threat is located is unwilling or unable to prevent the use of its territory’ for attacks, then it is entitled to have recourse to the right to use force in self-defence.²⁶² Australia has, similarly, advanced the ‘unwilling or unable’ test as the basis for the use of force under Article 51.²⁶³ However, other States have advanced altogether different approaches that would seem to suggest disagreement, or at best discomfort, with the ‘unwilling or unable’ test. Germany, for example, without referring at all to the ‘unwilling or unable’ test in the knowledge that other States before it had relied on that test, chose to rely, instead, on the fact that ISIS ‘occupied a certain part of the Syrian territory’, thus treating that part of Syrian territory as the territory of ISIS and not of Syria.²⁶⁴ To be sure, this itself is a controversial assessment, but one that falls beyond the scope of this chapter. Belgium’s letter is even more revealing of a distancing from the ‘unwilling or unable’ test.²⁶⁵ Not only does Belgium rely, like Germany, on the fact that ISIS has ‘occupied a certain part of the Syrian territory’ and the government of Syria ‘does not exercise effective control’ over these parts, it also states that ‘this is an exceptional situation.’²⁶⁶ Like the German statement, the Belgian letter seems to suggest that it is using force against the government in control of a territory. But more importantly, the invocation of the exceptional nature of the circumstances is an indication that it does not view its conduct as a law-making or law-transforming conduct, whether for the purposes of customary international law or treaty interpretation. In addition there are those States that have taken action in Syria and, while fully aware that there are different interpretations to Article 51 and that States taking action have advanced different legal bases for their actions, that have advanced no legal interpretation of Article 51, suggesting that they are non-committal about the legal interpretation of Article 51.²⁶⁷ And other States, notably Norway and Denmark, referred ambiguously to the fact that ISIS has established ‘a safe haven’ in Syria.²⁶⁸

To form the basis of subsequent practice in the interpretation of Article 51 of the Charter under Article 31(3)(b) of the VCLT, it is not sufficient that there is a common political purpose with respect to action to be taken in a particular case – as there seems to be, at least with respect to the States taking measures in Syria. There should also be a clear meeting of minds as to the interpretation of Article 51. While, at least in the case of Syria, there seems to be a number of States that have converged on the ‘unwilling or unable’ standard for the use of force in self-defence, this is not the case for all States. Some States also taking action in Syria, notably Germany and Belgium, have sought to clearly distance themselves from the broad ‘unwilling and unable’ test and have, instead, referred to a more narrow ‘where the State does not have effective control of its territory’ test. Furthermore, other States have not advanced an interpretation of Article 51, suggesting that their conduct is not based on an interpretation of Article 51. When one considers that many more States have not expressed themselves at all, the disparate understandings of the basis for action in Syria by those that have taken measures in Syria hardly serve as ‘establishing the agreement of’ UN member States as to the interpretation of Article 51.

Finally, some States have questioned, at least implicitly, the legality of the use of force in Syria without the consent of the Government. Corten identifies, for example, in addition to Syria, a number of States, including Venezuela, Russia, Ecuador, Cuba, Iran, China, Chad, South Africa, India, Belarus and Argentina, that criticised the military action in Syria without its consent as a violation of its sovereignty.²⁶⁹ For its part, the Non-Aligned Movement, a group comprising 120 States, though not commenting specifically on the use of force in Syria, has expressed concern about the trend to use extra-territorial force against non-State actors as part of the ‘war on terror’. The Non-Aligned Movement (NAM) accordingly

Reject[s] actions and measures, the use or threat of use of force in particular by armed forces, which violate the UN Charter and international law … under the pretext of combating terrorism.²⁷⁰

The language, of course, is not express. But it does serve to highlight the discomfort of a large number of States – States which have not participated in the actions in Syria – with the extraterritorial use of force to combat terrorism. More directly, a number of States, including the thirty-three States making up the Community of Latin American and Carribean States, have questioned the consistency with international law of the various letters sent to the President of the Security Council in support of the use of force in Syria.²⁷¹ This suggests that this large group of States does not share the interpretation that Article 51 permits the use of force against terrorists (non-State actors) in the territory of innocent non-consenting States. Indeed, in 2012 the NAM confirmed that

consistent with the practice of the United Nations and international law, as pronounced by the ICJ, Article 51 of the UN Charter is restrictive and should not be re-written or re-interpreted.²⁷²

The views of such a large majority of parties to the UN Charter, expressed subsequent to the operations in Syria, can surely not be ignored in determining whether those operations reflect the agreement of the parties to the Charter as to its interpretation.

The conflict in Syria is important for at least two other reasons. First, it shows the fallacy of the arguments that a broad interpretation of self-defence is necessary because of the weakness of the collective security architecture. That within a matter of days following the attacks in Paris on 18 November 2016 the Security Council was able to adopt a resolution containing language referencing the use of force illustrates that the Council can and has acted in response to international threats. There were, of course, dynamics that prevented the adoption of a full use of force authorisation. Whatever these were, it is clear that when the Council is committed against a common threat, even well-documented differences between its permanent members will not prevent action to address the common threat.

Second, as a policy point, ISIS does not only operate from Iraq and Syria. To the extent that it has operatives in the territories of various European territories and that the relevant European States are not able to stamp out these activities, it would not be far-fetched to suggest that such States are ‘unable’ to effectively deal with the threat from ISIS. Would that (should that) justify any State that has suffered attacks by ISIS to use force on these European States’ territories against ISIS? Further, does that not risk the type of escalation that our collective security system – underpinned by the prohibition on the use of force, with the exception of collective measures under Chapter VII of the Charter and, in urgent cases, the use of necessary force in self-defence – was designed to avoid? Is it not better, in such circumstances, to allow the Security Council or the General Assembly under the Uniting for Peace Resolution to exercise the mandate under the Charter to the best possible extent? It is of course true, as is arguably the case with terrorism and ISIS in particular, that measures adopted by the Security Council may not be sufficient, or be deemed sufficient by other States. But is that not the nature of the collectivity? That sometimes measures are more and sometimes less than what we, individually, may desire? To decide that we reserve the right – when such measures do not go as far as we would wish, or when they go further than we would prefer – to engage in unilateral action is to undermine that collective security system. This is so particularly when the very States that reserve the right to ‘go at it alone’ without the collective also have disproportionate influence on the Security Council – in terms of the States exercising force on the territory of Syria against ISIS without the consent of Syria, this would include the United States, France and the United Kingdom. This is the very constellation described by the proverb of eating one’s cake and having it.

While there are many examples of incidents of States using force in the territory of third States in response to attacks from non-State actors, many of these qualitatively are not able to form the basis of either interpretative State practice or form the basis of a rule of customary international law because of the negative reaction from other States. Other cases advanced in support of a broad interpretation of self-defence, in particular the 9/11-related practice, appear to be based on attribution and can therefore not form the basis of the right to use force in the territory of an innocent third State. The Syria-related cases cannot form the basis of the expansive interpretation because they do not establish the agreement of the actors concerning the interpretation of Article 51. Contemporary practice does not, therefore, as is sometimes argued, form the basis of an expansive interpretation of Article 51 which would permit the use of force against non-State actors in the territory of innocent third States.

VI. Unilateral or Collective Security: The Intersection of Law and Policy

A. Scope and Limits of the Law of Self-Defence against Non-State Actors

The above analysis has illustrated a number of points concerning the law on the use of force against non-State actors. First, as a rule, a State may only use force in self-defence against another State in response to an armed attack from that State. Thus, at least as a general rule, a State may not invoke the right to self-defence as a justification for the use of force in the territory of a third State in response to acts performed by non-State actors. This view is supported by an interpretation of Article 51 of the Charter based on Article 31(1) of the VCLT, and there does not appear to be subsequent practice within the meaning of Article 31(1)(b) of the VCLT to lead to a contrary conclusion. Second, the fact that, as a general rule, a State may not use force in self-defence in response to an armed attack, does not mean that a State is prohibited from using force against non-State actors in the territory of a third State in all circumstances. In addition to the case where the third State has given consent, a State may use force in self-defence against non-State actors in the territory of the third State if acts of the non-State actor can be attributed to the third State.

While the basic rule articulated above is, to my mind, relatively clear, the application of the rule is less clear. In particular, an issue that arises in connection with this basic articulation of the law as it stands is: what is the standard of attribution that is required? Must the non-State actor be acting on behalf of the third State concerned or be sent by the third State for the armed acts of that non-State actor to be the basis for self-defence under Article 51, as seems to be the requirement under Nicaragua? Or is it sufficient for the third State to have assisted or harboured the non-State actor? The law articulated by the International Court of Justice, and which has yet to be overturned by any clear, consistent practice of States, appears to be that the non-State actors must have been sent by, or acting on behalf of, the third State. Whether, as a normative question, this standard ought to be reconsidered, is discussed below (section VI.B). Whatever the standard of attribution is, what should not be debatable, as a matter of law at least, is that some level of attribution is required before a State may exercise the right of self-defence against a non-State actor on the territory of a non-consenting third State. This basic rule reflects an equilibrium in the Charter that ought not to be lightly disturbed: allowing one State to exercise its right under Article 51 means that another State will lose protection offered under Article 2(4), and this requires that the latter State has done something to justify this loss.

B. Do Current Circumstances Call for a New Approach?

As discussed above, there are new realities on the ground, in particular the rise in terrorism and the ability of terrorist groups to carry out large scale attacks, coupled with the existence of territories over which States may not have control.²⁷³ These new realities might, it may be suggested, require that States should have greater leeway to protect themselves from incidents of terrorism. Schachter states, in this regard, that terrorists operating abroad ‘should not enjoy immunity’ and ‘[c]ounter force is the most obvious remedy [to terrorist attacks]; one uses fire to fight fire.’²⁷⁴ Kaufman succinctly describes the arguments as follows:

We find ourselves, it is widely claimed, in a new and unprecedented age of terrorism. The rules of the game, we are told, must change in order for us to respond to a new kind of threat from shadowy terrorist organizations not affiliated with any particular state, and with possible access to weapons of mass destruction.²⁷⁵

According to Bethlehem the ‘reality of the threats, the consequences of inaction, and the challenges of both strategic appreciation and operational decision-making in the face of such threats frequently trump doctrinal debates’ about the scope and limits of the right to self-defence.²⁷⁶ It is important, he continues, that the legal ‘principle is sensitive to the practical realities’ on the ground ‘even as it endeavours to prohibit the excess and egregious pursuit of national interest’.²⁷⁷ These are, no doubt, important considerations that should bear, at the very least, on the normative question of what the law ought to be in this area.

In assessing Bethlehem’s views, it is useful to point out that there are policy considerations that pull in an altogether different direction. The broad interpretation of the right to use force against non-State actors on the territory of third States would, in the first place, bring us back the law of the jungle where might was right – if we are, as an empirical fact, already there, it would serve to legitimise that empirical fact. Militarily powerful States would be free to use force unilaterally in the territory of other States wherever they perceived, rightly or wrongly, that terrorist actors were operating. Corten warns that if this broad interpretation were allowed, in principle, ‘every state would be allowed to launch a military campaign on another state’s territory, under the sole pretext of the “inability” of this State’ to put an end to the activities of the non-State actors.²⁷⁸ This interpretation would counter the attempt identified by Bethlehem to ‘prohibit the excess and egregious pursuit of national interest’. It risks bringing to fruition Franck’s view of a world in which the UN architecture for peace and security had failed and wherein ‘the national interest of the super-powers has usually won out over treaty obligations.’²⁷⁹ Second, and more importantly, such a broad interpretation could undermine the very collective security system established by the UN Charter. A third State on whose territory force is being used against non-State actors in response to an attack from the non-State actors not imputable to that State could respond by forcible measures of its own against the State using force on its territory, with the potential for an all-out war.²⁸⁰ Furthermore, consider that terrorists operate from potentially everywhere in the world. That means that potentially any State that has suffered an attack by a terrorist group may launch attacks against that group in any State if, in its view, that would prevent further attacks against it. It does not take an oracle to see that the slope to pre-twentieth-century international law, where the use of force was fully permitted under law, is a slippery one. These are all policy considerations that should be taken into account in determining, as a normative proposition, whether the law should shift towards a more expansive view of the right to use force in self-defence.

Much of the arguments in support of a broad right to use force often appeal to people’s fears. This appeal to the fear of the citizenry, in fact, permits States with the power to use force to liberate themselves from the need to abide by international law.²⁸¹ The fear-inducing idea that not expanding the right to use force in self-defence against non-State actors on the territory of innocent States would place States at great peril is, as a matter of law, an overstatement. It ignores that the Charter architecture for peace and security is based, first and foremost, on collective security. In particular, that a State is rendered helpless, just because particular acts that threaten its security do not amount to an armed attack against it, ignores the primary role assigned to the UN Security Council. In such instances, the Security Council has both the right and responsibility to act. Indeed, even in cases that do amount to an armed attack as understood by the International Court of Justice, the right to use force in self-defence is meant to be temporary, in order to repel an attack until the Council has acted. The dangers of this expanded right to use force, which places the collective security architecture on the sidelines, is poignantly reflected on by the Secretary-General, responding to arguments for an attack against Iraq:

According to this argument, States are not obliged to wait until there is agreement in the Security Council. Instead, they reserve the right to act unilaterally, or in ad hoc coalitions. The logic represents a fundamental challenge to the principles on which, however imperfectly, world peace and stability have rested for the last fifty-eight years. My concern is that, if it were to be adopted, it could set precedents that resulted in a proliferation of the unilateral and lawless use of force, with or without justification.²⁸²

Expanding on the need to adopt a strict interpretation of Article 51, the International Court of Justice has made the following observation:

Article 51 of the Charter may justify a use of force in self-defence within the strict confines there laid down. It does not allow for the use of force by a State to protect perceived security interests beyond these parameters. Other means are available to a concerned State, including in particular, recourse to the Security Council.²⁸³

It is true that often, due to political constraints, including the exercise of the veto by its permanent members, the Council does not act when it should. Proponents for the interpretation of the right to self-defence in international law to include the use of force on the territory of innocent third States point out that it would be unacceptable to subject the right of a State to protect itself to depend on an ineffective system. In addition to the general policy consideration already referred to – i.e. this broad interpretation would lead to legitimation of the ‘might is right’ phenomenon and would undermine the collective security system – this argument suffers from at least two flaws. First, the UN Security Council has, in fact, adopted measures against international terrorism.²⁸⁴ The measures adopted by the Security Council need not necessarily be use of force measures. States might feel that the measures adopted are insufficient and that the Council may have, or even should have, done more, but that does not detract from the fact that the Council has acted. Second, it seems problematic, on the basis of any perceived ineffectiveness of the Security Council, to expand the rule relating to self-defence which, as suggested earlier, would significantly erode the prohibition on the use of force. The content of the law on self-defence should not be dependent on the effectiveness of the Security Council. If there is a problem with the collective security system, it is that problem that ought to be addressed.

Even if, as a policy requirement, current circumstances require some change to the law of self-defence as reflected in the jurisprudence of the International Court of Justice in order to allow affected States a greater possibility to respond to non-State actors, the question remains what those changes should be. As suggested above, the idea of removing the requirement for attribution is not only inconsistent with the law but it is also inappropriate from a policy perspective. One possibility, which would at least not offend the basic principles of fairness and equity, would be to lower the attribution standard of effective control stated in the Nicaragua case. From a policy perspective, permitting the use of force in self-defence in the territory of a State that harbours or aids non-State actors would seem acceptable. Importantly, this lowered standard retains the requirement of attribution and thus would not countenance the use of force in the territory of an innocent third State. This possibility has been considered and supported by, amongst others, Randelzhofer and Nolte and, it appears, Tams.²⁸⁵ This approach is also consistent with some individual opinions of judges of the International Court of Justice.²⁸⁶ At any rate, it appears the US attacks on Afghanistan, which did not meet with objection from third States, were based on this same standard. Whether that one single episode is sufficient to effect a change to the rules is less clear. Tams (p. 164) seems to suggest that events post-9/11 may have already resulted in a shift in the law to a point where the ‘harbouring’, ‘aiding or abetting’ or ‘provision of assistance to’ non-State actors involved in attacks in a State would allow the attacked State to use force in self-defence even in the absence of consent. As discussed above (section V.B.3), I have a different impression of the impact of that practice. Moreover, I am less certain that the law has already shifted, although I am prepared to accept that it may be shifting in that direction.

Murase, however, cautions that not any involvement by the territorial State can or should justify the use of force.²⁸⁷ To the extent that such involvement constitutes the breach of an obligation under international law, then such a State can be held responsible for that assistance under the law of State responsibility.²⁸⁸ He observes, correctly in my view, that not all cases of State responsibility permit recourse to the use of force in response.²⁸⁹ The point here would be to emphasise the exceptional nature of the right to use force in self-defence. It is probably for this reason that the Court in the Nicaragua case set such a high threshold for attribution. Nonetheless, the law is not static and it does change. If it were to be developed in the direction of lowering the threshold for attribution from the strict ‘sent by or on behalf of’ test of Nicarargua, the key question would be to determine the extent of the third State’s involvement necessary to attribute to the non-State actor’s activities to the third State. It would require an assessment – well beyond the scope of this contribution – of whether the involvement of the third State in the activities of the non-State actors was ‘substantial’.²⁹⁰

For the purposes of this chapter, only two points need to be made. First, such a shift, even if desirable, cannot be achieved by separate and dissenting opinions of the judges of the International Court of Justice or the writings of commentators based on acts that either do not receive the acceptance of other States or that may be explained in terms of the Nicaragua test of attribution. To lead to a different interpretation of Article 51, there has to be practice of States establishing the agreement of the UN member States as a whole or general and widespread practice accepted as law to establish a rule of customary international law. The second important point for the purposes of this chapter is that, even if that shift has occurred, or does in the future occur, the requirement of attribution has (or would) remain intact because the lowering of the threshold does not detract from the requirement for attribution.

VII. Summary and Conclusions

The rules relating to the use of force and, in particular, self-defence are sensitive and complex. Yet these rules have a definite content and are not ‘in the eye of the beholder’. Moreover, these rules have to be seen in the context of the broader collective security architecture underpinning the UN system. Under these rules, States are prohibited from using force against the territorial integrity or independence of another State. Any use of force in the territory of another State amounts to a use of force which is prohibited under international law. In the event of a threat to the State, including in the case of armed aggression, the Security Council has the primary responsibility to take measures on behalf of the international community as a whole. This mechanism for collective action is the primary tool through which threats to peace and security, including threats to individual States, are to be addressed. Recourse to self-defence, an important right recognised in the Charter, is meant to be exceptional, limited and temporary, ceasing to exist if and when the Council assumes its responsibility.

The exceptional right to self-defence applies in the case of an armed attack from a State. While the Charter does not stipulate that the armed attack must be by ‘a State’, this has, until recently, been the generally accepted interpretation of Article 51. Moreover, this interpretation is consistent with the context of the Charter provisions. Of course, where the acts of a non-State actor are attributable to a State, the victim State is entitled, in order to repel the attack, to use force in the territory of the third State.

The arguments, based on ‘practice’, that international law does permit the unilateral use of force in self-defence against non-State actors on the territory of innocent third States are, at best, unconvincing and, at worst, dangerous. Very often the practice that is advanced has been objected to by other States such that it cannot contribute to the interpretation of Article 51 of the Charter or form the basis of a rule of customary international law. Where the acts have not been objected to, such as the post-9/11-related acts, these have not supported the proposition that international law permits the use of force in self-defence absent some form of attribution to the territorial State. It is true that the events in Syria, and in particular the use of force by the US-led coalition in Syria without the latter’s consent, may suggest a turn of law in the direction of an expansive interpretation of the law on self-defence. However, even these acts by some States reveal sufficient inconsistencies to prevent an interpretative 180 degree turn of the law. At best they reveal the possibility that law could, with sufficient coordination by a group of States and the acquiescence of others, turn in that direction. This would be a dangerous turn, undermining the collective security architecture of the United Nations – a turn that may well lead to the unleashing of the ‘scourge of war’ on succeeding generations that the drafters of the Charter felt so inspired to prevent. Moreover, subsequent to the advent of operations in Syria, a large grouping of States, the NAM, reiterated its view that Article 51 does not permit the extraterritorial use of force against non-State actors.

The collective security system is imperfect. As a result, not all harm can be prevented. But the alternative, unilateralism, is much worse. It means that we are all in danger, at all times. Unilateralism means the militarily powerful are free to determine when and how to exercise force, and the weak are at the eternal mercy of the powerful.²⁹¹ However, any sense of security that comes with the power to unilaterally use force is a false sense of security. Unilateralism carries the risk of escalation that the authors of the Charter sought to avoid. States in a position to use their military power might not see the dangers now, but tomorrow other States will hold the military might to do the same and might point to the precedents of current actions to justify what might otherwise be generally understood as illegal acts under international law.²⁹²

The law has not changed. It remains that the right of self-defence is an exceptional right, limited in scope, available only in the most urgent of cases to repel an attack. There are those arguing vociferously for an expansion of this right to such an extent that States would be free to use force in the territory of that State, without that State’s consent or without attribution of the conduct of the non-State actors to that State. Terrorism, as heinous as it certainly is, does not offer sufficient reason to depart from the constraints placed by international law. What terrorism does do is offer an opportunity for us to enhance the cooperation necessary to make the collective security system more effective. The alternative, undermining the collective security system by promoting unilateralism and the law of the jungle, should be resisted. Indeed, this may be precisely what the terrorists want from us: to undermine the collective security system and the rule of law in favour of unilateralism, the law of the jungle and, ultimately, chaos and insecurity.