1. Introduction
The use of force by non-state actors is an unfortunate but recurring phenomenon in international affairs. It is, to be sure, not a new phenomenon. One of the most famous episodes in the history of international law, the Caroline incident between the US and Great Britain in 1837, involved the use of cross-border force by insurgents.Footnote 1 In contemporary conditions, the threat posed by non-state actors’ uses of force is, however, greater than it has ever been. Technological and military advancements, as well as their availability to private actors, has exponentially increased non-state actors’ capacity for destruction.Footnote 2 It is only necessary to read the daily news to realize just how devastating the actions of these irregular groups can be. How to deal with this phenomenon (in all its aspects) is one of the most pressing issues in and one of the most important challenges for contemporary international law.
One of the central questions in this regard is that of the availability of a right of self-defence against non-state actors on the part of states that are victims of military attacks by these groups.Footnote 3 Since 1945 at least, international law has envisaged a right of self-defence applicable in inter-state relations only: by a victim state against the aggressor state.Footnote 4 In this framework, uses of force by armed bands and the like had to be attributed to a state to trigger the right of self-defence, which would then become the author of the attack and the target of self-defensive force.Footnote 5 But since the attacks on the World Trade Center on 9/11, states have increasingly called into question this understanding of the right of self-defence by invoking the right to respond to non-state actors’ uses of force.Footnote 6 To cite two well-known examples, it was widely recognized that the US acted in self-defence against al-Qaida in Afghanistan in 2001,Footnote 7 and Iraq, and its allies, including France, the UK and the US, have asserted a right of self-defence (be it individual or collective) against Islamic State of Iraq and Syria (ISIS) in Syria.Footnote 8 Whether this practice is sufficient (and sufficiently uniform)Footnote 9 to evidence a change in the traditional understanding of self-defence is still a contentious issue.Footnote 10 The most charitable view in this regard is that the use of force in self-defence against non-state actors is ‘not unambiguously illegal’;Footnote 11 but even then, it is unclear that this view is generally accepted.
In no small part, the difficulty with recognizing a right of self-defence against non-state actors is that of explaining why the state in whose territory those actors operate (the ‘host state’) is liable to the use of force within its territory. The ‘dilemma’, as Tom Ruys and Sten Verhoeven put it, is ‘clear’:
On the one hand, it can not be accepted that states must simply undergo private attacks without having the right to defend themselves by using force against the home bases of these groups and their accomplices . . . On the other hand, state sovereignty is and remains one of the basic pillars of international law and order and should not lightly be violated.Footnote 12
Why must the host state tolerate the infringement of its territorial sovereignty, or a (forcible) intervention in its affairs, or the impairment of, say, legally protected trade and other commercial relations by the victim's use of force? In considering this question, Kimberley Trapp has (rightly) observed that self-defence must ‘in some way excuse the violation of’ these rights of the host state if it is to be an effective mechanism.Footnote 13 It may be added that the legal justification of the impairment of host state rights is necessary to avoid triggering the host state's right of self-defence against the victim state; to uphold, in short, the principle that there is no self-defence against self-defence.Footnote 14 Indeed, if the victim's use of force is, from the standpoint of the host state, a forcible violation of its sovereignty, then why can the host state not react in self-defence against it?Footnote 15 Farfetched as this may sound, a right of self-defence in these circumstances was invoked by Syria in the wake of indications by the US and other European powers that they intended to attack ISIS targets within Syria's territory as an exercise of collective self-defence of Iraq.Footnote 16
There are many disagreements as to the conditions and the legal ground of justification for the impairment of host state's rights by self-defensive action – if one exists at all. At a bare minimum, states and scholars agree that any such justification must be grounded, in some measure, on the involvement of the host state with the non-state actors mounting the attack. This is evidenced by the references to, for example, the ‘harbouring’ by Afghanistan of al-QaidaFootnote 17 and, most recently, by the claim of Iraq and the US that Syria was unwilling or unable to deal with ISIS in its territory.Footnote 18 To be sure, host state involvement with the non-state actors warrants differential treatment depending on the kind and degree of the relationship between the two.Footnote 19 Where involvement is enough for attribution of the private conduct to the state, self-defence can be exercised in an inter-state manner.Footnote 20 Where attribution is not possible, two questions arise: first, what kind and degree of involvement is necessary; and, second, how to rationalize the fact of host state involvement into a ground of legal justification. As to the kind and degree of involvement, agreement seems to be coalescing around a standard of ‘unwillingness and inability’ of the host state to deal with the non-state actors.Footnote 21 But host state involvement is a factual question,Footnote 22 and does not, on its own, constitute a legal ground of justification. It is thus necessary to rationalize this fact into one such legal ground. The latter point is still controversial; many solutions have been proposed in the literature, but none of these has received general endorsement.Footnote 23
The majority of the debates on self-defence against non-state actors have centred on Article 51 of the Charter,Footnote 24 either to demonstrate that the restrictive inter-state understanding is not supported by the language of the provision or intended by the drafters,Footnote 25 or to identify a different standard of attribution of the non-state actors’ attack to the host state.Footnote 26 These works constitute a crucial step for the development of the law of self-defence, if this body of law is to adequately address the threat posed by non-state actors. Nevertheless, this focus has also produced the ‘dilemma’ that Ruys and Verhoeven mention: Article 51 (may) authorize the use of force against the non-state actors, but it fails to address the relations between victim state and host state. In contrast with this approach, this study queries whether Article 21 of the Articles on the Responsibility of States for Internationally Wrongful Acts (ARS)Footnote 27 may provide an answer to the ‘dilemma’.Footnote 28 Pursuant to Article 21: ‘The wrongfulness of an act of a State is precluded if the act constitutes a lawful measure of self-defence taken in conformity with the Charter of the United Nations.’
Curiously, the vast literature on self-defence against non-state actors has thus far overlooked Article 21. In fairness, this provision is mostly neglected in the scholarship on the use of force,Footnote 29 and when actually addressed it has been largely misunderstood.Footnote 30 Can Article 21 provide the legal justification for the impairment of host state rights by self-defensive force used within its territory?
The study answers this question in three steps. First, in Section 2, it will begin by explaining the intended role and scope of Article 21 of the ARS. The section will describe the context and historical developments which rendered this provision necessary, and will show how, even though unacknowledged in practice, self-defence operates to preclude the wrongfulness of the infringement of other rights of the target state by lawful measures of self-defence. Second, in Section 3, the study then considers the possible application of Article 21 in the context of self-defence against non-state actors within the host state. Finally, Section 4 will provide some concluding remarks.
Before entering into the substance of this study, some methodological and terminological observations are necessary. As to the methodology, this study is limited to the theoretical question of how to justify the encroachment of host state rights by the use self-defensive force in its territory. In accordance with the trend in state practice, this article will assume that host state involvement with the non-state actors launching the attack is necessary. In particular, given its recent momentum, this article will assume that the kind and degree of host state involvement necessary amounts to ‘unwillingness or inability’. The study will not, therefore, seek to prove or disprove this test by reference to state practice; equally, it will not elaborate on the precise content of the standard.Footnote 31 As to terminology, in order to avoid confusion between the two provisions in Article 51 of the Charter and Article 21 ARS, the rule in Article 51 will be referred throughout as the ‘right of self-defence’ or the ‘primary rule of self-defence’, whereas the rule in Article 21 will be referred to as the ‘circumstance precluding wrongfulness’, ‘justification’ or ‘secondary rule’ of self-defence. This use of terminology should not suggest, however, that these are two different norms. As Section 2 will clarify, the customary law recognizes a right of self-defence, and its functions and effects in the international legal order are codified in two provisions: Article 51 excepting self-defence from the prohibition of force, and Article 21 providing a justification for the collateral impairment of rights of the target state by self-defensive force.
2. Self-defence as a circumstance precluding wrongfulness
Until 1945, the institution of the state of war (also known as ‘formal’ or ‘technical’ war) governed all of the legal relations existing between belligerent states.Footnote 32 The state of war wholly excluded the law of peace, and replaced it with the law of war in the relations between belligerents.Footnote 33 Moreover, (peace-time) treaties were terminated or, at the very least, suspended in the relations between the parties.Footnote 34 As a result, the military measures adopted by either party to the conflict could not violate or even impair the rights and obligations which governed their relations during peace.
The state of war is now an institution of pure historical interest. Whether it is compatible with the Charter's collective security system remains debatable,Footnote 35 but it is certainly the case that states no longer claim to be in a ‘state of war’ when they are engaged in armed conflict.Footnote 36 In contemporary conditions, states engaged in armed conflict remain ‘formally at “peace”’,Footnote 37 which means that all their legal relations remain in force throughout the period of hostilities.Footnote 38 In these conditions, the use of force can impair a multiplicity of obligations existing between the two states engaged in conflict. For example, when the use of force occurs in the target state's territory, that force will constitute a breach of the prohibition of force, and it will also, among others, be an impairment of the target's territorial sovereigntyFootnote 39 and of its right to be free from intervention.Footnote 40
Ordinarily, these impairments constitute breaches of international law. But what if a state resorts to force in self-defence? Customary international law recognizes the right of states to use force in self-defence when they are the victims of an armed attack. This is a unanimously accepted entitlement,Footnote 41 codified in Article 51 of the UN Charter and regularly described as an ‘exception’ to the prohibition of force.Footnote 42 Resort to force in self-defence, being excepted from the prohibition of force, does not therefore constitute an infringement of that prohibition. That use of force is lawful by reference to the prohibition of force itself. But what about the other legal relations just mentioned; is defensive force lawful also by reference to these legal relations? As will be seen, Article 21 of the ARS is intended to justify the impairment caused by lawful force on these other legal relations.
The next two sections unpack Article 21 in an effort to elucidate its role in the international legal order. The analysis will be performed in two steps. First, the legal relations between the victim and the aggressor state will be considered. For analytical purposes, the legal relations between these two states will be divided into three categories and the interaction between self-defence and each of these categories will be considered. Second, this section will consider the question of third-party rights and whether self-defence can provide a justification for the forcible interference with the rights of third states. As will be seen later, this is an issue of special relevance where self-defence is exercised against a non-state actor within the territory of the host state.
2.1. Effects of self-defence on victim-aggressor state relations
The first step in understanding the function and scope of Article 21 is to separate the legal relations existing between the states involved in an armed conflict into three categories. First is the relation governed by the prohibition of force and its exception, the right of self-defence. Second are all the other legal relations (conventional or customary) existing between those states. These comprise an ‘infinite variety’Footnote 43 and include the rights of territorial sovereignty and non-intervention, trade and commercial rights, treaties for the exchange of technology, and so on. Third is the category of legal relations which impose ‘obligations of total restraint’,Footnote 44 including (some) rules of international humanitarian law and human rights law. As will be explained in what follows, the customary right of self-defence provides a legal ground of justification for the breach of the first and second sets of legal relations. These effects are codified in separate provisions: the legality of force under the first legal relation is addressed by Article 51, and the legality of the impairment by (defensive) force of the second set of legal relations is addressed by Article 21.Footnote 45 The third set of legal relations cannot lawfully be impaired even when acting in self-defence.
2.1.1. First legal relation: Legality of resort to force
States are bound, under international law, by an obligation not to use force against one another. The prohibition, codified in Article 2(4) of the UN Charter, is widely regarded as one of the jus cogens norms.Footnote 46 This obligation not to use force possesses (at least) one exception in the form of a right of self-defence.Footnote 47 What this means, essentially, is that states are bound to respect their obligation not to use force unless they are the victims of an armed attack, in which case they are authorized to respond to force with (defensive) force.
The structural relation between the prohibition of force and the right of self-defence is very complex, and it has been conceptualized in many different ways. The International Law Commission (ILC) construed the prohibition of force and its exception of self-defence as a single norm.Footnote 48 The prohibition of force and the right of self-defence, in other words, are two segments of an over-arching norm which can be restated, in simplified but relevant part, as follows: ‘the use of force is prohibited except in self-defence against an armed attack’. In conceptualizing the norm in this way, the ILC was accommodating the opinions of states according to which the use of force in self-defence is not inconsistent with the prohibition of force and that it is, rather, lawful ab initio.Footnote 49 In this understanding, the exception of self-defence constitutes a limit on the material scope of the prohibition. Thus, the prohibition of force would be applicable only if there were a non-defensive use of force and defensive force would fall wholly outside of the prohibition's scope of application.Footnote 50
The ILC's approach is, to be sure, not uncontested and has been criticized for its artificiality.Footnote 51 Indeed, the relation between the two could be construed differently, in a way that upholds the autonomy of the right of self-defence from the prohibition of force. In this construction there would be two rules, the prohibition and the right, both applicable to the same set of facts: the use of defensive force would be banned under the prohibition of force and permitted by the right of self-defence. The two rules would, therefore, be in a normative conflict. To conclude in favour of the legality of self-defence, the right of self-defence must be able to set aside the application of the prohibition. This construction of the relationship between the prohibition and the right runs into an obvious obstacle: the prohibition's peremptory status. As a jus cogens rule, the prohibition cannot be set aside,Footnote 52 and the conflict would be resolved by the lex superior principle; the prohibition would prevail over the right, with the result that defensive force is unlawful. This is a paradoxical conclusion, in utter contradiction with the practice of states. The paradox can be overcome in one of two ways. First, by ‘elevating’, so to speak, the right of self-defence to a peremptory rule. In this case, the conflict would arise between two peremptory rules and could potentially be resolved by application of the lex specialis principle.Footnote 53 Since self-defence is the most factually specific of the two rules, it would set aside the prohibition of force in the circumstances. Alternatively, it could be argued that only the prohibition of aggression possesses peremptory status. The resulting conflict between self-defence and the prohibition of force would be a conflict between ‘ordinary’ rules, in which self-defence would again prevail as the factually most specific rule. Both solutions, however, lack grounding in positive law: states do not accept the jus cogens status of the right of self-defence,Footnote 54 and they do uphold the jus cogens status of the prohibition as a whole.Footnote 55 The ILC's solution, artificial as it may be, appears to be the most descriptively accurate understanding of this relation.Footnote 56
In whatever way the relation is construed, the point is that the right of self-defence constitutes an authorization to use force where force would otherwise be banned. To put it in symbolic terms, Article 51 of the Charter authorizes conduct otherwise prohibited by Article 2(4). When a state resorts to force in self-defence, Article 51 in a sense ‘takes care’ of the inconsistency existing between that force and Article 2(4), either by limiting the prohibition's scope or by setting aside its application in the circumstances. Crucially, however, the right of self-defence is not an authorization to impair all other rights of the aggressor state.Footnote 57
2.1.2. Second set of legal relations: ‘Other’ rights of the aggressor state
The victim and aggressor states are also bound by an ‘infinite variety’ of other rights and obligations, be they customary or conventional. All of these other rights can potentially be impaired by military measures. The case-law of the ICJ and other international tribunals shows the great variety of rules of international law potentially impaired by the use of force, including: commercial obligations under bilateral treaties,Footnote 58 aviation agreements,Footnote 59 the obligation to settle disputes peacefully,Footnote 60 and so on. Indeed, as the ICJ made clear in the Oil Platforms case, international obligations may be breached by whatever means, including: by the decision of a court, by an Act of Parliament, or by the use of force.Footnote 61
The question thus emerges whether the impairment of these obligations through defensive force is unlawful, or whether these impairments may be justified if the military force was resorted to in self-defence. Of course, it is perfectly plausible that these interferences are unlawful even if the state using force was acting in self-defence. As the ICJ stated in Croatia v. Serbia, ‘[t]here can be no doubt that, as a general rule, a particular act may be perfectly lawful under one body of legal rules and unlawful under another’.Footnote 62 Thus, self-defence may be lawful in relation to the prohibition of force and be unlawful in relation to, for example, the right of territorial sovereignty. Namely, a use of force may be compatible with the first legal relation mentioned above, but incompatible with the second set of legal relations. International law, however, opts for the opposite conclusion: the use of force in self-defence is lawful by reference to both the first and second sets of legal relations.
Article 21 is concerned precisely with the effects of self-defensive force on the second set of legal relations. That is, it is concerned with the collateral effects, so to speak, of self-defensive force on other rights of the aggressor state comprised in the second set mentioned above. The provision reflects a recurrent, if not directly acknowledged, phenomenon in practice.Footnote 63 The ICJ, for example, has never directly acknowledged that self-defence may justify these collateral violations, though this effect can be deduced from its decisions in Nicaragua and DRC v. Uganda (Nicaragua). In Nicaragua, the Court found that the violation of Nicaragua's territorial sovereignty and non-intervention by US military operations was not justified by self-defence because, on the facts of the case, the US did not have a right of (collective) self-defence.Footnote 64 The same reasoning can be found in DRC v. Uganda, where the Court found that, since Uganda did not have a right of self-defence, its actions breached the DRC's rights of territorial sovereignty and non-intervention.Footnote 65 A contrario, it may be inferred that if the US and Uganda had a right of self-defence against Nicaragua and the DRC respectively, then there would have been no breach of those rights either. In state practice, the parties’ pleadings in Oil Platforms are especially instructive in this regard. On the question whether self-defence could justify the breach of the Iran-US 1955 Amity Treaty,Footnote 66 the US, directly invoking Article 21, maintained that ‘[a]ny actions of the US deemed incompatible with Article X of the Treaty would not be wrongful by the operation of this principle [self-defence] of customary international law’.Footnote 67 Iran also accepted this proposition, and said that:
[A]ction otherwise lawfully taken in self-defence could constitute a circumstance precluding wrongfulness in relation to Article X, paragraph 1, of the Treaty. In other words, it accepts the proposition contained in Article 21 of the ILC Articles on the Responsibility of states.Footnote 68
Iran later added that if made out, self-defence would ‘exonerate the United states entirely; it would provide a complete justification for their conduct, in accordance with Article 21 of the ILC's Articles’.Footnote 69
At customary law, then, it appears that self-defence can also justify the collateral impairment of these obligations by means of force. Article 21 simply reflects this additional function of the right of self-defence in the international legal order. It is worth pointing out that this is not an additional norm in the legal order, capable of going beyond the right of self-defence. Simply put, Article 51 and Article 21 codify different effects of the exercise of the customary right of self-defence in the legal order. Article 51 concerns its effect on the prohibition of force, and Article 21 the effects of that exercise on the second set of legal relations.
2.1.3. Third set of legal relations: Obligations of ‘total restraint’
Self-defence, however, cannot justify the collateral impairment brought about through force of all obligations of the victim state (or rights of the aggressor state). The Commentary to Article 21 expressly excludes from the scope of this provision what it termed, following the ICJ's advisory opinion in Nuclear Weapons,Footnote 70 the ‘obligations of total restraint’.Footnote 71 By these are meant those obligations ‘expressed or intended to apply as a definitive constraint even to states in armed conflict’, including non-derogable human rights, the ‘intransgressible’Footnote 72 principles of international humanitarian law and other obligations designed to place a limit on the conduct of hostilities.Footnote 73 This exclusion was accepted by Iran in Oil Platforms,Footnote 74 and by Uganda in DRC v. Uganda, whose memorial explicitly stated that self-defence may in no circumstances justify breaches of human rights and humanitarian law.Footnote 75 Since the impairment of these obligations may not be justified by self-defence, no more will be said about them in this study.
2.2. The effect of (defensive) force on third states?
The Commentary to Article 21 specifies that the ‘essential effect’ of this provision ‘is to preclude the wrongfulness of conduct of a state acting in self-defence vis-à-vis an attacking state’.Footnote 76 And yet, the use of force in international relations may, in addition, affect the rights of third (neutral) states. Can self-defence justify the impairment of these states’ rights?
During the first reading of the ARS, the Commission had entirely foreclosed the possibility that self-defence may justify the impairment of third states’ rights. The commentary to draft Article 34, the predecessor to Article 21, was clear in this regard:
The Commission [wished] to point out that the provision in article 34 is not intended to preclude the wrongfulness of, so to speak, indirect injury that might be suffered by a third state in connection with a measure of self-defence taken against a state which has committed an armed attack.Footnote 77
The interests of third parties were, therefore, ‘fully protected’.Footnote 78 But during the second reading of the ARS, the Commission changed its position. Prompted by Special Rapporteur Crawford,Footnote 79 the ILC observed in 1999 that ‘a state acting in self-defence might be entitled to take action against a third state’ and that ‘there was no need to make an explicit reference to that circumstance’ in the text of, or in the Commentary to, Article 21 since the issue ‘was adequately covered by the relevant primary rules’.Footnote 80 To reflect this, the Commentary to Article 21 currently indicates that this provision ‘leaves open all issues of the effect of action in self-defence vis-à-vis third states’.Footnote 81
2.3. Interim conclusions
A brief summary of the discussion so far may be useful before turning to the application of Article 21 to the exercise of self-defence against non-state actors. When a state resorts to force in self-defence, it does so against the background of the complex web of legal relations which bind it to the aggressor state (and third states). For analytical clarity, these legal relations were divided into three sets since the effect of the right of self-defence on each of them is codified in different provisions. First is the prohibition of force. When force is resorted to in self-defence it does not constitute a breach of the prohibition since self-defensive force is excluded from the scope of the prohibition. In conventional law, the legality of the use of force is grounded in Article 51 of the UN Charter. Second is the category of ‘other’ obligations of the victim state (or ‘other’ rights of the aggressor) which may be impaired by forcible measures. These include the rights of territorial sovereignty, non-intervention, commercial or other trade rights, and so on. The impairment of these rights is also justified by self-defence if the force that encroached upon them, in other words, constitutes a lawful exercise of the right of self-defence in international relations.Footnote 82 This justifying effect of self-defence, a sort of incidental effect of self-defence, is codified in Article 21 of the ARS. Third, is the category of legal relations which impose absolute restraints on states in times of armed conflict, even if the state is acting in self-defence. These include non-derogable human rights and ‘intransgressible’ humanitarian law rules. The impairment of these obligations may not be justified, even by reference to the right of self-defence. Finally, Article 21 and its Commentary leave open the question of whether the impairment of third state rights by self-defensive force may be justified.
3. The circumstance precluding wrongfulness of self-defence and the use of force against non-state actors
The ILC's work on Article 21 assumed the exercise of self-defence in an inter-state context.Footnote 83 So long as self-defence is exercised as between two states – including when the armed attack is private but attributable to that stateFootnote 84 – and the resort to force complies with the requirements of the jus ad bellum, then self-defence will also produce the incidental effect of justifying collateral violations of the aggressor state's rights belonging to the second set of legal relations described above. Could this provision be applied to situations involving unattributable armed attacks by non-state actors; namely, to situations where self-defence is exercised against a non-state actor within the territory of the host state, so as to justify the impairment of host state rights by the defensive force?Footnote 85
There is a crucial distinction between the ‘essential’ case covered by Article 21 and the present scenario in that in the present scenario there is an additional party: the non-state actor. This additional party intrudes at the level of the first legal relation. The prohibition of force is, of course, owed as between the two states, the victim and the host state. Nevertheless, the exception of self-defence accommodates (or can accommodate) a third party: the non-state actor. The victim state is under an obligation not to resort to force unless it is the victim of an attack, which could include attacks by a private party.Footnote 86 In positive terms, a state would be authorized to resort to force in international relations whenever it is the victim of an armed attack – regardless of its source. This authorization is codified in Article 51 of the Charter which, in colloquial terms, takes care of Article 2(4) whenever force is used in self-defence. Article 51 ensures that defensive force is not incompatible with the prohibition in Article 2(4), even if the defensive force is directed against a non-state actor. But this additional party is not accommodated by the second set of legal relations; these exist as between the victim state and the host state only. It is precisely for this reason that the ‘dilemma’ arises to begin with; while the victim may have a right of self-defence against the non-state actors, the defensive force encroaches upon the rights of a different entity, the host state.
Scholars have articulated different ways to resolve this dilemma. As noted in the Introduction, all these solutions have at their basis the (factual) premise of host state involvement with the private groups mounting the armed attack. The type and degree of involvement seems to be coalescing around a standard of ‘unwillingness or inability’, a standard which remains vague and has been abused. Be that as it may, of interest here is how the fact of state involvement (whatever the standard may be) has been and can be rationalized into a legal ground of justification that may explain why the host state's rights are not impaired by the use of defensive force against non-state actors within its territory. Section 3.1 will provide an overview and critique of the three main approaches offered by the scholarly literature to answer this question. Section 3.2 then considers whether Article 21 ARS may have a role to play in this regard.
3.1. Insufficiency of the mainstream approaches
Legal scholarship has appraised the fact of host state involvement with non-state actors in this context in at least three different ways: as an element of the requirement of necessity of self-defence; as a breach of international law by the host state through complicity with the non-state actors; and as a breach of a due diligence obligation to protect the rights of other states within its territory.Footnote 87 None of these approaches has received general support in either state practice or the scholarly literature arguably because, as will be explained in what follows, each of these approaches is unsatisfactory in some way and this diminishes their explanatory power.Footnote 88
According to some scholars, host state involvement with the non-state actors is relevant to the (customary) requirement of necessity of self-defence.Footnote 89 Pursuant to this view, the right of self-defence of the victim state is triggered by the armed attack of the non-state actors and its exercise is limited by the conditions of necessity and proportionality. The host state's involvement with the non-state actor, so this view goes, would fall to be analyzed under the requirement of necessity. Thus if the host state were willing and able to deal with the threat, then the use of force in self-defence would not be necessary. Contrariwise, if the host state is unwilling and/or unable to do so, then the necessity condition would be met and the victim state could lawfully resort to force against the non-state actors in the host state's territory.
This is a rather simple way to account for the host state's involvement with the non-state actors in situations of self-defence, but crucially it fails to explain why the impairment of host state rights caused by defensive force is justified. This can be elucidated by reference to the first two sets of legal relations described in Section 2. The requirement of necessity belongs to the first legal relation – the exceptional entitlement to use force in self-defence, codified in Article 51 of the Charter. The condition of necessity must be met to establish the legality of the resort to defensive force. When host state involvement is accounted for under the condition of necessity, it only serves to show that in the circumstances the victim state resorted to force against the source of the threat (the non-state actors) lawfully by reference to the norm prohibiting force in international relations. But the conclusion about the legality of the resort to force as such says nothing about the legality of the impairment of rules in the second set of legal relations (those between the victim and host state) affected by the exercise of self-defence – so precisely the ones the impairment of which is in need of legal justification. In addition, this solution distorts the ‘original purpose’, in Christine Gray's words, of the requirement of necessity insofar as it treats necessity as an enabler of, rather than as a limitation on, the use of force.Footnote 90
Other scholars have rationalized host state involvement under the notion of complicity. In this view, since the host state is complicit in the armed attack of the non-state actor then it becomes liable to the use of defensive force against the non-state actor within its territory.Footnote 91 This is not an argument about attribution – it is not the case, under this approach, that the complicity with the non-state actor determines the attribution of that armed attack to the host state. Rather, the position is that the host state's breach of its obligation not to aid or assist in the commission of a wrongful act by another renders it liable to the use of defensive force in its territory. This is a questionable conclusion. Under the law of state responsibility, complicity in the act of another state entails the responsibility of the accomplice. Crucially, the accomplice is responsible for its own conduct, for its own aid and assistance in the commission of a wrongful act by another state.Footnote 92 In the current scenario, a host state would be responsible for assisting and aiding the non-state actors to engage in an armed attack. The host state's own wrongdoing (the aid and assistance) would entail its responsibility, in the form of cessation (if the act were continuing) and reparation towards the victim state.Footnote 93 But the accomplice is not responsible for the wrongful act of the principal actor. The host state is thus not responsible for the private armed attack as such.Footnote 94 This being the case, the host state is not liable, by the mere fact of its complicity with the non-state actors, to the use of force in its territory. To be successful, this approach must explain why complicity in this context entails a consequence not normally entailed under the general law of responsibility. As it currently stands, however, the approach simply assumes this consequence to be the case; the argument, that is, assumes the conclusion it seeks to prove.
The same criticism can be levelled at the third and final approach mentioned. Several scholars have attempted to ground the justification of the impairment of host state rights by defensive force by reference to the host state's infringement of its due diligence obligations. It is well-established that states have an obligation to protect the rights of other states within their own territory.Footnote 95 When non-state actors operating in a state's territory successfully mount an armed attack against the victim state, the host state could be found to have breached its due diligence obligation.Footnote 96 The infringement of this obligation, the argument goes, ‘opens up’ the state to force being used within its territory.Footnote 97 As a result, the impairment of host state rights is lawful. But this too is an unconvincing argument.Footnote 98 Just as with the argument based on complicity, the due-diligence approach assumes the conclusion. A violation of due diligence obligations entails responsibility, involving the obligations of cessation and reparation, but this violation does not render the state liable to the use of force within its territory any more than the violation of any other rule (other than the prohibition of force) entails this consequence.Footnote 99 The argument fails to explain why a breach of due diligence in the circumstances would generate this consequence. As it is, it too simply assumes this to be the case.
Both of the last two approaches, based on the notion of complicity and due diligence obligations, are underpinned to some extent by countermeasures-like reasoning. In essence, both approaches posit that one violation of international law (the complicity in another's wrongful act, or the breach of due diligence duties) justifies another violation of international law (the encroachment of host state rights). But countermeasures may not themselves involve the use of force,Footnote 100 so it is not clear why the breach of obligations in these circumstances would attract a forcible response.
3.2. Self-defence as a circumstance precluding wrongfulness: An alternative solution?
None of the mainstream approaches provide a satisfactory explanation to the ‘dilemma’ of impairment of host state rights by the use of defensive force against non-state actors within that state's territory. Could Article 21 provide an alternative solution?
The scenario where a state exercises self-defence against a non-state actor in another state's territory is not the ‘essential’ situation falling within the ambit of Article 21 – that of inter-state self-defence – as noted earlier. Rather, the situation of the host state is better assimilated to one involving a third party affected by the exercise of self-defence. The action in self-defence exists as between the victim state and the non-state actor, and the host state is an (affected) third state. As was mentioned in Section 2.2, the ILC does not answer the question whether Article 21 can justify the impairment of third states’ rights by an action in self-defence. The matter, it says, is left ‘open’. In this way, the ILC does not foreclose the possibility that Article 21 may extend to the justification of the impairment of third party rights. Insofar as Article 21 codifies an effect of the right of self-defence at customary law,Footnote 101 the question then is whether as a matter of customary law the exercise of self-defence permits these intrusions on third parties’ rights.
There are two bodies of law relevant to this question. First, the law of neutrality which protects the rights of and imposes duties on third parties during armed conflict. A third party may, by its involvement with one of the parties in conflict, forfeit its neutral protection and render itself liable to attack by the opposing party. Where this is the case, there is no role for Article 21 to play; the law of neutrality would itself resolve the ‘dilemma’ of host state rights. Second is the right of self-defence itself. In certain circumstances – whether or not a third state is involved in any way in the conflict – a self-defending state may lawfully impair (some of) the third states’ rights. Does this extend to the impairment of any of its rights? If so, such an effect could be rationalized under Article 21 of the ARS.
3.2.1. What role for the law of neutrality?
Under the law of neutrality, neutral states have a right not to be affected by the conflict and they have duties of non-participation and impartiality.Footnote 102 A neutral state which renders assistance to one of the parties to an armed conflict thereby violates its neutral duties, and becomes liable to reprisals by the affected belligerent. Nevertheless, this is not a liability to forcible reprisals. Since international law prohibits armed reprisals,Footnote 103 the affected belligerent may not use force against the neutral state on this ground. According to Michael Bothe, the only circumstances in which the neutral state may be liable to forcible attack by the affected belligerent is when its violation of neutral duties amounts to an armed attack; ‘a reaction against violations of neutrality which would involve the use of force against another state is permissible only where the violation of the law triggering that reaction itself constitutes an illegal armed attack’.Footnote 104 The standard is high; not all support for an aggressor is equivalent to an armed attack.Footnote 105 Bothe notes that the non-neutral services rendered by the US, Saudi Arabia and Kuwait to Iraq, for example, ‘did not entitle Iran to adopt measures against those states involving the use of military force’.Footnote 106 Applying this to the present context,Footnote 107 the use of force would be permissible within the host state's territory whenever its assistance to the non-state actors amounts to an armed attack.
The difficulty is, of course, that of determining when assistance amounts to an armed attack – is attribution necessary, or will an indirect attack suffice? By requiring a degree of state involvement with the non-state actors (or any other state in conflict) amounting to ‘an armed attack’ against the other party, the law of neutrality reaches the same stalemate as any arguments based on Article 51; failing attribution of the non-state actors’ conduct to the host state, there will be no entitlement for the victim to use force against, or even within, the host state. A plausible avenue for development could be the lowering of the threshold of host state involvement with the non-state actors, such that, for example, unwillingness and inability to deal with the private groups might suffice. But there are two weaknesses in this argument. First, it is difficult to see how unwillingness and, especially, inability might constitute an involvement equivalent to an armed attack. Second, it would probably be simpler to redefine the notion as a matter of the right of self-defence itself. At any rate, given the reluctance to redefine the notion of armed attack in the context of the right of self-defence, it is highly unlikely that the law of neutrality might develop in this direction.
3.2.2. Self-defence and third parties
The other possibility is that the right of self-defence itself may allow a state to use force against third parties or, more likely, that it permits the impairment of third party rights by self-defensive force. Certain impairment of third-state rights during hostilities have been justified on the grounds of a state's exercise of self-defence. This is the case, for example, of the institution of war-zones or maritime exclusion zones on the high seas, where they affect the freedom of navigation of all states.Footnote 108 While in the past these would have been grounded on the exercise of belligerent rights, in recent decades recourse to self-defence as a legal basis of justification for these interferences is increasing.Footnote 109
States having justified interferences with neutral shipping on the basis of self-defence include France, during the Algerian emergency of the mid 1950s, and Pakistan, during the Indo-Pakistani conflict of 1965.Footnote 110 The UK justified the institution of a 200 nautical miles exclusion zone around the Falklands/Malvinas and the institution of ‘security bubbles’ around its warships traveling through the Atlantic Ocean, on the right of self-defence.Footnote 111 These zones were initially applied only against Argentine warships, though they were later extended to ‘any other ship, whether naval or merchant vessel, which . . . is operating in support of’ Argentina’.Footnote 112 During the Iran-Iraq war, at least two states, the UKFootnote 113 and the US,Footnote 114 accepted the principle that self-defence could justify interference with neutral shipping when there was reasonable suspicion of non-neutral service.Footnote 115 In particular, the UK held that:
a State . . . actively engaged in an armed conflict, is entitled in exercise of its inherent right of self-defence to stop and search a foreign merchant ship on the high seas if there are reasonable grounds for suspecting that the ship is taking arms to the other side for use in the conflict.Footnote 116
On the whole, the Iraqi and Iranian war-zones were widely criticized by other states and by the Security Council. But as emphasized by Ross Leckow, these critiques responded to the ‘little respect . . . shown by either side for [the] principle of restraint’ in their enforcement of these areas. It was ‘because of this complete disregard for “reasonableness” that attacks on neutral merchant vessels on both sides must be condemned as violations of international law’. Most recently, after the 9/11 attacks, the US engaged in the boarding of neutral vessels suspected of transporting terrorists. While the explicit legal basis for this was never articulated, the US generally justified its action in this regard on the right of self-defence.Footnote 117
The legal basis of interference with neutral shipping, to be sure, remains a matter of some debate; some uphold a wider right of interference as a matter of belligerent rights, and therefore available to both victim and aggressor state, whereas others – like the examples just mentioned – opt for a narrower right based on a state's exercise of self-defence.Footnote 118 As observed by Natalie Klein, the invocations of self-defence are of special significance given their context. In her view, ‘[t]his recent practice should be seen against previous practice where there was far less acceptance of reliance on the right of self-defence to justify intrusions against the freedom of navigation on the high seas’.Footnote 119 This practice, at the very least, shows states’ willingness to accept the impairment of the rights of third states in the exercise of self-defence in certain, if limited, circumstances.
The reasoning underpinning the legal justification of these interferences with third party rights could be extended, by analogy, to the ‘dilemma’ of host state rights as well. The analogy could be built on the UK's position on self-defence-based interferences with neutral shipping, quoted earlier, as this is the clearest articulation of a justification for impairment of third state rights in the circumstances. Under the UK position, a state acting in self-defence may impair the rights of third states to freedom of navigation only where there is suspicion of involvement between the third state and the aggressor. Involvement, in this regard, amounts to assistance in the form of arms-provision – so a far lower standard than what is necessary for that assistance to itself constitute an armed attack. Applying this same reasoning to the non-state actor scenario, the victim state exercises its right of self-defence against the non-state actor and, in so doing, it impairs the rights of a third party, the host state. Such interference may justified under self-defence so long as there exists some measure of host state involvement with the non-state actor, an involvement which need not amount to an armed attack and could, if accepted, be as limited as unwillingness and inability. Nevertheless, given the different extent to which third state rights are impaired in the two scenarios (in one case, the right of freedom of navigation, in the other, the rights to territorial sovereignty, non-intervention, and so on), it may be necessary to qualify the UK position above if it is to apply in respect of host-states. So, instead of requiring a reasonable suspicion of host state involvement, the justification would only be afforded where host state involvement actually existed and there was evidence to prove it.
To put this approach in terms of Articles 51 of the Charter and 21 of the ARS, the resort to force in self-defence against the aggressor (the non-state actor) would be grounded on Article 51 of the Charter – which, it may be recalled, is applicable only to the question of the legality of the use of force by reference to the general prohibition – whereas the interference with third state rights may be grounded on Article 21 ARS.Footnote 120 Such an approach, which relies on Article 21 in addition to Article 51, overcomes the explanatory difficulties of the mainstream theories reviewed earlier. To begin with, it relies on the concept of ‘circumstances precluding wrongfulness’, a category of circumstances in the law of responsibility whose function is precisely that of justifying the infringement of legal rules. Most importantly, the relevant circumstance precluding wrongfulness, Article 21, can justify the breach of obligations brought about through the use of force, which no other circumstance precluding wrongfulness, or any of the mainstream approaches reviewed earlier, are capable of achieving.Footnote 121
4. Conclusion
The use of force against non-state actors presents, for international law, a clear dilemma: it is not possible to expect the victim of non-state actors’ assaults to tolerate that conduct, but by the same token the rights of the host state must be upheld. State practice, albeit far from the generality, uniformity and consistency necessary for the modification of customary law, shows an increased acceptance of a right of self-defence against non-state actors – but the recognition of this right must overcome an important practical obstacle: how, legally, to justify the interference with the rights of the host state by the use of force in its territory?
Scholars working in this field have proposed different explanations for the justification of interferences with host states' rights. Building from the practice of states, they all rely, in some form or another, on the involvement of the host state with the non-state actors. Nevertheless, they disagree as to the rationalization of the fact of involvement into a ground of legal justification. Thus, some scholars view host state involvement as relevant to the condition of necessity of self-defence, others as a form of complicity, and others still as a breach of the host state's due diligence obligation to protect the rights of the victim state in its territory. But none of these approaches is sufficient; at the crucial point, that of explaining why the host state is liable to the use of force within its territory, they all contain leaps in the reasoning.
This study has proposed an alternative answer to this question, one based on Article 21 of the ARS, namely on self-defence in its role as a circumstance precluding wrongfulness. This provision, albeit much misunderstood in the literature, has an important role to play in inter-state invocations of self-defence. It is Article 21 which explains why force that is lawful under Article 51 of the Charter does not constitute an infringement of other rights of the target state, most importantly of its right of territorial sovereignty. While Article 21 assumes an inter-state use of force, in which there is identity of the parties to the relevant legal relations (self-defence, territorial sovereignty, and so on), its Commentary shows that Article 21 may extend to the justification of the interference with the rights of third states, namely situations in which there is a mismatch between the parties to the affected legal relations. Indeed, the right of self-defence may, in certain circumstances, impair the rights of third parties. Thus, states have justified the institution of maritime exclusion zones, affecting third party rights of free navigation, on the basis of self-defence. Perhaps host state involvement with the non-state actors may be reconceptualized as one of the situations in which the right of self-defence permits interference with a third party, the host state. This way, the legal relation between the victim state and the non-state actor would be governed by Article 51, and the legal relations between the victim state and the host state would be governed, and thereby justified, by Article 21.
To be sure, a solution based on Article 21 is still far from being entirely satisfactory. This is not so much as a result of the approach itself, but of the difficulties inherent in the claim that force can be lawfully used in self-defence against non-state actors. These difficulties are both legal and practical. Legally, any theory about the use of self-defence against non-state actors must overcome an important hurdle – that of grounding the illegality of the armed attack. An armed attack, to trigger the right of self-defence, must not be merely a ‘factual’ armed attackFootnote 122 – it is necessary that it also be unlawful. If this were not the case, then the continued validity of the principle ‘no self-defence against self-defence’Footnote 123 would be undercut. For if all that is necessary is a factual armed attack, regardless of its legal qualification, then a use of force in self-defence can, too, be a factual armed attack (if it attains the required gravity threshold) triggering the aggressor's own right of self-defence. The result is a logical spiral of violence, where self-defence can be invoked against self-defence. Yet, there is no identifiable ban on the use of force by non-state actors.Footnote 124 Practically, the desirability of this development remains questionable, as recently shown by Christine Gray, who has argued that the use of force against non-state actors is far from an effective instrument.Footnote 125
Article 21 can provide a sound framework to address the ‘dilemma’ of host state rights. Of course, this solution would still require the acceptance of states to become binding at positive law (possibly through their practice and opinio juris). While a conceptually more satisfying solution can be found to the ‘dilemma’, the availability of a right of self-defence against non-state actors still faces many theoretical and practical hurdles.
