1. Introduction
The laissez-fair attitude of the nineteenth century, with the liberty to go to war seen as an attribute of the sovereignty of states,Footnote 2 slowly gave way to new thinking with the 1899 and 1907 Hague Peace Conferences.Footnote 3 Some change of heart was indeed noticeable, and the beginning of the twentieth century brought increasingly strict rules limiting the use of force or war by states.Footnote 4 But that century nevertheless saw two World Wars disrupt international relations and the lives of countless individuals. The UN Charter (the Charter),Footnote 5 which was negotiated to set up the new post-war organization to replace the discredited League of Nations, broke radically with the past and attempted to install a collective security system with a force monopoly for the Security Council and a narrow exception for self-defence.
Those educated in the international legal tradition of the post-war period may be surprised to see the tidal change of thinking in the post 9/11 world. Although one could certainly not speak of a consensus on various legal problems that arose in relation to the use of armed force by states, mainstream scholarship before the 9/11 attacks tended to favour restrictive rules over permissive norms.Footnote 6 Instead, what we see now is the suggestion that self-defence is permissible in anticipation of imminent attacks or threats, an expansive interpretation of the notion of armed attack, increasing acceptance of the ‘pinprick’ or ‘accumulation of events’ theory, and an understanding that armed attacks may not only originate from states but also from non-state actors.
This last topic forms the subject-matter of the current contribution. According to conventional wisdom, the exercise of the right of self-defence was predicated upon an armed attack by another state. This implied that armed activities of non-state actors needed to be attributable to a state or, at the very least, that the state was substantially involved in these. Since the 9/11 attacks, these rules have been (re)interpreted to allow for self-defence against a state that supports non-state actors carrying out armed activities in other states, harbours them, or is unwilling or unable to repress such armed activities within its territory. Some even contend that self-defence may be invoked against non-state actors based in a particular state independently of that state's conduct.
It should be understood that this contribution does not intend to provide substantive analysis on the merits of this or that interpretation of Article 51 of the Charter or comparable rules of customary international law. Nor is it intended to engage in substantive discussion of the rules of treaty interpretation or the method(ology) of customary international law creation. Rather, the purpose is to identify and trace restrictivist reasoning by authors upon and following the 9/11 attacks, and to critically assess the accountability provided, or lack thereof, with respect to the rules of interpretation and method(ology) of customary international law creation. In using the term ‘restrictivist’, this contribution does not aim to qualify specific authors, their arguments and positions (per se) as such, and nor is it intended to suggest that those are the result of a deliberate choice rather than the result of considered reasoning.
This contribution will deal with the interpretation of Article 51 of the Charter and state practice and its acceptance (as law) evidenced by the 9/11 attacks and later practice and responses thereto. To this purpose, Section 2 will address the silence of Article 51 as to the origin of an armed attack primarily in terms of contextual and teleological interpretation. Section 3 will look at the attribution of armed activities of non-state actors to a state, or its substantial involvement therein, to allow for the exercise of the right to self-defence. Finally, Section 4 will investigate certain issues of method concerning subsequent practice establishing the agreement of the parties and practice accepted as law.
2. The interpretation of the silence of article 51: an armed attack by a state
This section will mainly discuss restrictivist reasoning in terms of contextual and teleological interpretation. Although Article 51 of the UN Charter requires the existence of an armed attack as a conditio sine qua non for the exercise of the right of self-defence, it fails to specify from whom or which entity such an attack should originate. Drawing upon this silence, many authors have claimed that measures of self-defence may be taken by a state in response to attacks by non-state actors operating from the territory of another state.Footnote 7 In this regard, Judge Higgins's observation that there is nothing in Article 51 that indicates that self-defence is available only in case of attacks by states is frequently quoted.Footnote 8 As a matter of deductive reasoning, various authors make the argument that this silence signifies that Article 51 not only covers armed attacks by states but also those by non-state actors, and sometimes this argument is framed in reverse – that armed attacks by non-state actors are not excluded from the scope of Article 51.Footnote 9 Responses of more restrictivist inclined authors have been to draw attention to the context and telos of the Charter.
What can be deduced from the context of the Charter as a whole in order to answer the question whether attacks by non-state actors are included or excluded under Article 51? Generally, Article 2(4) of the Charter is taken as a starting point: This prohibits UN members from having recourse, in their international relations, to the threat or use of force against other states or which is inconsistent with the purposes of the UN.Footnote 10 It follows that the obligation imposed by Article 2(4) does not address non-state actors, and their violent actions consequently do not violate what could be called the jus contra bellum.Footnote 11 In the same vein, the reference to ‘international relations’ is interpreted to mean that armed action will only violate Article 2(4) if directed against a state.Footnote 12
This is relevant to the construction of Article 51 because the right of self-defence is available only in response to unlawful armed attacks.Footnote 13 The reason for this is the necessity to prevent regression into a sequence of equally valid self-defence claims, and in this way to break an otherwise inevitable vicious circle.Footnote 14 If non-state actors cannot violate Article 2(4) and there is no rule of customary international law to this effect addressed to them, then the exercise of the right of self-defence must, generally, be directed against the unlawful conduct of a state, and, specifically, against an illegal use of force (of a certain gravity), in order to be justified.Footnote 15 In other words, there must be an identity between the unlawful armed attack and the use of force prohibited to member states under Article 2(4) of the Charter.Footnote 16
Another contextual aspect frequently referred to is the general rule–exception relationship that exists between Articles 2(4) and 51.Footnote 17 If the right of self-defence could be freely exercised in response to armed activities of non-state actors and would allow their targeting by military force, whatever territory they might be found in, this implicates Article 2(4).Footnote 18 However, the use of force in self-defence needs to be invocable and justifiable as against the state on whose territory it takes place and not merely against the non-state actors concerned.Footnote 19A fortiori, as Kammerhofer quite astutely points out, military action by a state against non-state actors does not require any justification in se by way of self-defence under the jus contra bellum, since Article 2(4) prohibits the use of force against states.Footnote 20
Moving now from context to object and purpose, the relationship between self-defence and collective security may be raised. Wolfrum's commentary witnesses that the Charter's object and purpose is to prevent and suppress the use of armed force in international relations through the institution of a collective security system.Footnote 21 Otherwise, the maintenance of international peace and security is claimed as the Charter's fundamental purpose, and a restrictive interpretation of armed attack or Article 51 as an exception to the prohibition of the use of force is called for.Footnote 22
In assessing the reasoning of authors in relation to the silence of Article 51, the following observations may be made with respect to the methods of interpretation. First, it may be observed that the Vienna Convention on the Law of Treaties (Vienna Convention or VCLT), although covering ‘constituent instruments of an international organization’ (Article 5), is not applicable to the Charter due to its non-retroactivity (Article 4).Footnote 23 In consequence, only rules of treaty interpretation established under customary international law apply, and one would have expected more frequent mention of this circumstance.Footnote 24 Second, it may be noted that only some authors making use of this or that element of interpretation make any actual reference to the rules of interpretation and to Articles 31 and 32 of the Vienna Convention.Footnote 25 Indeed, hardly any provide an account of the elements of interpretation or their weight, and little attempt is made to qualify arguments in terms of elements of interpretation.
The particular assertions regarding the silence of Article 51 as to the origin of an armed attack appear at times to function as a knock down argument, making redundant any subsequent inquiry as to the proper construction of Article 51. Certainly, its silence makes a determination of the ordinary meaning of its terms impossible, but this does not prejudge the use of other elements of interpretation. In assessing this silence, authors do not problematize the general rule of interpretation as to ordinary meaning of terms, even when the interpretation of silence in a treaty has been addressed in different contexts.Footnote 26
However, authors do generally contextualize silence by making reference to Article 2(4) of the Charter, and indicate that the use of force in self-defence on the territory of another state in response to armed activities of non-state actors requires that such use of force is invocable and justifiable against the territorial state. The suggestion that non-state actors may be targeted in self-defence independently from any linkage to the territorial state – whether evidenced by attribution of their activities, the territorial state's (substantial) involvement therein, or the territorial state's unwillingness or inability to repress such activitiesFootnote 27 – is not generally admitted by authors as a matter of contextual interpretation.
Although the telos of the Charter is mentioned at times, relatively little weight appears to be given to this element of interpretation, and expansionist authors do not engage with it much. The commentary of the International Law Commission (ILC) suggests by reference to a treaty's object and purpose, and good faith, that an interpretation should be preferred that allows the treaty to have appropriate effects.Footnote 28 When discussing collective security or the maintenance of international peace and security, authors mostly do not argue these to make a choice between one or another interpretation of the silence in Article 51 of the Charter. Moreover, at times the interpretation of that silence is rather linked to the telos of self-defence rather than the Charter,Footnote 29 without however accounting for the fact that Article 31(1) of the VCLT speaks of the object and purpose of the treaty rather than its individual clauses.Footnote 30
The preparatory works of the Charter are referred to with some frequency, though authors come to contradictory conclusions as to the inter-state character of self-defence.Footnote 31 No attempts are made to explain or justify recourse to the preparatory works, whether to confirm the meaning found by application of Article 31 or to determine a meaning when this is left ambiguous or obscure or leads to an unreasonable result.Footnote 32 Furthermore, very few of the contributions surveyed make reference to the reports of the United Nations Conference on International Organization,Footnote 33 instead basing their claims in this respect on secondary sources.
3. The standard to determine an armed attack by a state: attribution and/or (substantial) involvement
For restrictivist authors, a contextual and teleological interpretation of Article 51 supports the necessity of some linkage of the armed activities of non-state actors to a state in order to qualify as an armed attack by a state, but this raises the question as to the substance of that connection. Many different options have been put on the table, but this section will explore especially the reasoning relevant to the question of attribution of the armed activities of non-state actors to a state or that state's (substantial) involvement therein. That (conceptual) division has formed the basis of discussion, following the court's seminal treatment of (the definition of) armed attack in the Nicaragua case.Footnote 34
In respect of the armed activities of non-state actors, the argument has been made that the lack of protection by a territorial state establishes the necessity of self-defence by another state, irrespective of the legality or illegality of the former state's conduct. As such, even the breach of due diligence obligations, which require a state to prevent its territory from being used to the detriment of another state, would not be required.Footnote 35 The link between the territorial state and the non-state actors consists, so it is claimed, in the inability or unwillingness of that state to stop their activities,Footnote 36 which does not necessarily constitute a wrongful act.Footnote 37 Since the law of state responsibility and the law of self-defence are different branches, the linkage evidenced by this inability or unwillingness founds the necessity of self-defence, i.e., that no alternative ways are available to defend against the non-state actors.Footnote 38 For some, the non-state actors concerned may be targeted, but self-defence against the state would require attribution of their armed activities.Footnote 39
This may be contrasted to slightly different reasoning, in which the inability or unwillingness of the territorial state to prevent the armed activities of non-state actors is linked to the breach of positive or due diligence obligations.Footnote 40 With respect to positive obligations, reference is made to the Declaration on Principles of International Law, which embodies the obligation not to engage in ‘organizing, instigating, assisting or participating in acts of civil strife or terrorist acts in another State or acquiescing in organized activities within its territory directed towards the commission of such acts, (. . .).’Footnote 41 As to due diligence obligations, these are based on the general duty for a state not to allow the use of its territory for acts by private persons contrary to the rights of other states, and to use reasonable means to prevent such acts.Footnote 42 Although not always clearly articulated, it is then the breach of such positive or due diligence obligations that would excuse or justify the violation of the territorial integrity of a state.Footnote 43
The first scenario, that the inability or unwillingness of a state to protect another state against armed activities of non-state actors founds the necessity of self-defence irrespective of any link to a state, is rejected as a matter of restrictivist reasoning, because to consider private attacks as an armed attack would lead to the absurd result that a state could be subjected to military force without having violated international law.Footnote 44
As to the second scenario, linking the inability or unwillingness of a state to breach of due diligence (and positive) obligations, it has been observed that these involve duties to protect rather than duties to abstain. Negligence in dealing with, or tolerance of, the armed activities of non-state actors would violate the former duties but not the latter, and a breach of obligation in this respect cannot give rise to self-defence against the territorial state.Footnote 45 In addition, a due diligence obligation is one of means and not of result, and it requires a State to employ all reasonable measures in the circumstances, requiring awareness and the means to act. As such, a failure to exercise due diligence with respect to armed activities of non-state actors ‘does not automatically qualify as an “indirect use of force” ’.Footnote 46
Discarding self-defence against non-state actors in the absence of a link to the territorial state, other than presence or to the breach of positive or due diligence obligations, the debate has mainly focused on the role of the attribution of armed activities to the territorial state or its substantial involvement therein.
The general rules relating to the attribution of conduct to a state are formulated in Articles 4 to 11 of the Articles on State Responsibility (ASR).Footnote 47 However, these also envisage in Article 55, which is titled lex specialis, that the general rules may not apply to the extent that special rules govern.Footnote 48 The Court made reference to the latter in the Genocide case, when it formulated an evidentiary standard:Footnote 49 ‘[t]he rules for attributing alleged internationally wrongful conduct to a State do not vary with the nature of the wrongful act in question in the absence of a clearly expressed lex specialis.’
The 9/11 attacks have served as a catalyst to the debate, as the question was raised as to whether these attacks – planned and executed by al-Qaeda – could be attributed to Afghanistan so as to allow the invocation and exercise of the right of self-defence by the United States. Various bases of attribution under the law of state responsibility have been put forward, in particular Articles 4, 5, and 8–11 of the ASR. Most commonly attribution under Article 8 is discussed, which allows for the attribution of conduct of ‘private’ individuals or groups acting under the instructions, direction or control of a state.Footnote 50 As instructions or direction tend to be difficult to prove, the debate has focused on the standard of control required to attribute the armed activities of a non-state actor to a state.
In this context authors discuss the open conflict between the International Court of Justice (ICJ) and the International Criminal Tribunal for the former Yugoslavia (ICTY) over the appropriate standard of either effective control or overall control. The ICJ first determined ‘effective control’ to be the proper standard in the Nicaragua case in 1986.Footnote 51 The ICTY in part disagreed and argued for ‘overall control’ in relation to organized armed groups in the Tadić case in 1999,Footnote 52 but the ICJ maintained its earlier position in the Genocide case in 2007.Footnote 53 The effective control standard requires operational control over the specific activities of non-state actors,Footnote 54 whereas the overall control standard requires involvement of the state in the organization, planning and coordination of an operation.Footnote 55 This difference is important in order to determine whether specific acts committed within a larger operation are attributable to a state.Footnote 56
Contributions of authors tend to deal with attribution issues, and may be divided into: those affirming the necessity of attributing the armed activities of non-state actors to a state; those who discuss the role of attribution but with an independent role for substantial involvement; those claiming that the standard of control to attribute such activities under Article 8 of ASR is too strict and/or has been modified; and those who argue that special rules of attribution exist in relation to an armed attack for purposes of deciding whether self-defence may be invoked.
Within the first category, certain authors have affirmed the necessity of attribution of armed activities of non-state actors.Footnote 57 Corten and Dubuisson affirm this necessity and deny that the right of self-defence may be exercised against another state in its absence. However, it is to be noted that they discuss two distinct acts of attribution – that of the armed attack and that of the armed activities. With respect to the former, they investigate the extent of ‘substantial’ involvement, whereas with regard to the latter, they examine the rules of attribution under the law of state responsibility.Footnote 58
Regarding the attribution of the 9/11 attacks, Murphy suggested, contingent upon the facts, that attribution could take place based on Articles 2, 4 and 5 of the ASR for toleration by Afghanistan of a terrorist group engaged in earlier attacks, on Article 9 for allowing al-Qaeda to exercise governmental functions in projecting force abroad, or on Article 11 because Afghanistan's de facto government adopted al-Qaeda's conduct by not extraditing its operatives.Footnote 59 However, many authors have denied attribution based on Articles 8–11 ASR in view of the available facts regarding the connection between al-Qaeda and the Taleban government both before and after the attacks.Footnote 60
Within a second category of reasoning, attribution of the armed activities of a non-state actor to a territorial state constitutes a sufficient but not a necessary condition for the invocation of self-defence, since self-defence is also available if a territorial state is substantially involved in those armed activities (infra this section).Footnote 61 Authors adhering to this kind of reasoning invoke the Definition of Aggression and/or the ICJ's observations in this respect in the Nicaragua case.Footnote 62 As such, an armed attack by a state may be said to have taken place even in the absence of attribution of such activities.
The third category of reasoning expounds that the 9/11 attacks and responses thereto contributed to a change of the standard of attribution under Article 8 of the ASR. Thus it has been argued that the effective control test of the Nicaragua case was overturned, and that the overall control standard propagated by the ICTY in the Tadić case sufficed to attribute the 9/11 attacks to Afghanistan.Footnote 63 This reasoning is supplemented by a variety of other legal constructions, such as attribution taking place under an active support standard,Footnote 64 a refined harbouring thesis,Footnote 65 or a low(er) degree of involvement.Footnote 66
But perhaps these last positions should more properly be ranged among the last category of authors positing the existence of special rules of attribution. As a basis for such special rules of attribution applicable to determine the existence of an armed attack, it is argued that the support of terrorists below the level of direction or control, or providing them with safe haven, suffices.Footnote 67 Adding to this, it is suggested that where a state is unwilling or unable to prevent armed activities of non-state actors the breach of the positive obligation to prevent such activities would allow for their attribution.Footnote 68 Similarly, Randelzhofer and Nolte assert that any form of substantial involvement not only leads to attribution, but also allows for this when a state is unwilling or unable to deal with (large-scale) armed activities of non-state actors.Footnote 69
However, in view of the incoherent practice since the 9/11 attacks, Ruys has come to the conclusion that there is no clearly expressed lex specialis.Footnote 70 Although admitting the possibility of an evolution of the rules of attribution embodied in Articles 8, 11 and 9 of the ASR,Footnote 71 he highlights that attempts at revision in the context of self-defence appear artificial and counter-intuitive in view of the basic premise underlying the rules on attribution – that a state is only responsible for its own conduct, i.e., for the conduct of persons acting on its behalf.Footnote 72 Rather, changes in the law ought to be seen from the perspective of the primary rule(s) of self-defence in terms of substantial involvement of a state in the armed activities of non-state actors.Footnote 73
This brings us to substantial involvement. While some claim that the armed activities of non-state actors need to be attributable to a state, others argue that a state's substantial involvement in such armed activities constitutes an armed attack attributable to the state.Footnote 74 Nevertheless, as Verhoeven observes, the sending of armed bands or substantial involvement in their activities presupposes that the state is engaged, and he points to Article 1 of the Definition of Aggression which stipulates that ‘[a]ggression is the use of armed force by a State’.Footnote 75 In his view, substantial involvement of a state is not a requirement of attribution under the law of state responsibility, and he considers that lesser forms of involvement might be sufficient to speak of an act of aggression.Footnote 76 It has been stressed in a similar vein that state involvement is not necessary to show attribution, but is required to show: whether an armed attack has taken place;Footnote 77 with support of a state that must be a conditio sine qua non without which armed activities of a non-state actor could not have taken place;Footnote 78 or that substantial involvement constitutes a principle of attribution under the primary rules of self-defence.Footnote 79 Such reasoning then suggests a special rule of attribution of the armed activities or the armed attack, with substantial involvement as the standard of attribution.
Irrespective of the precise characterization of substantial involvement, the battleground has shifted to the level of involvement required to qualify the armed activities of non-state actors as an armed attack. As already recounted above, arguments have been made that a state may invoke the right to self-defence under lower standards of involvement,Footnote 80 or when a state aids and abets the armed activities of non-state actors to which they substantially contribute.Footnote 81 In the Nicaragua case, the ICJ stipulated that an armed attack did not include assistance to rebels taking the form of weapons deliveries or logistical or other support, although it added that these might be a threat or use of force, or amount to intervention.Footnote 82 Those discussing the ICJ's (restrictive) interpretation have pointed to the dissents of Judges Jennings and Schwebel,Footnote 83 with the former indicating that the combination of different forms of assistance might well be crucial in an armed attack and the latter arguing that such forms of involvement are tantamount to an armed attack.Footnote 84
However, the latter's reading of substantial involvement in the Definition of Aggression and its history has been refuted.Footnote 85 Involvement of a state requires knowledge and participation, and the latter needs to be substantial – excluding accessory or incidental involvement and toleration.Footnote 86 Additionally, restrictivist argument holds that a state must engage in a combined attack with non-state actors and its involvement would require preparation and execution of armed activities, without controlling them, by making available its infrastructure, equipment and services.Footnote 87
In assessing reasoning on attribution, it may first of all be noted that all authors reference the rules on state responsibility as formulated by the International Law Commission in 2001. As such, those rules are taken to reflect customary international law, although little attempt is made to establish this independently from earlier studies. Of course, in 2007 the ICJ affirmed in the Genocide case that the rules embodied in Articles 4 and 8 of the ASR reflect customary international law.Footnote 88 Exceptionally, van Steenberghe dispenses with attribution as a requirement altogether, noting that states invoking self-defence do not claim that the territorial state has, itself, committed an armed attack.Footnote 89
When considering the standard of control required under Article 8 of the ASR, most authors proceed on the understanding that this is effective control as maintained by the ICJ and ILC rather than overall control for organized armed groups as asserted by the ICTY.Footnote 90 Although van Steenberghe claims that a large number of scholars have attempted to attribute the 9/11 attacks, orchestrated and executed by al-Qaeda, to Afghanistan,Footnote 91 the survey above paints a somewhat different picture. Rather, their attribution to Afghanistan under Articles 4 and 8–11 of the ASR is, overall, denied by authors with the exception of a few.Footnote 92 Moreover, a potential ground for attribution such as de facto status of organ under Article 4(2) of the ASR is not raised at all by any author, and most likely will not be raised due to the ICJ's even more demanding ‘complete dependence and control’ test set out in the Genocide case.Footnote 93
However, of those arguing a change in the law, only a few have asserted a change in the general rules of attribution as established in the ASR. Instead, many have made the case for a special rule allowing for the attribution of the armed activities of non-state actors to a state in certain circumstances. Among the latter, relatively few seem to argue in favour of lesser forms of support, and most have used legal constructions that would attribute such activities to a state in case of sanctuary and support, complicity, aiding and abetting, or a state being unwilling or unable to repress such activities. With respect to this last, a connection is regularly made with a state's breach of its positive or due diligence obligations to not tolerate and to repress the armed activities of non-state actors within its territory. Ruys has denied the existence of special rules of attribution, pointing to incoherent practice.Footnote 94
If a change in the law on self-defence cannot be established by reference to the (general or special) rules of attribution, this does not exclude the possibility that it could lie with the concept of armed attack and the involvement of a state in the armed activities of a non-state actor. When considering the required level of involvement, substantial or otherwise, similar kinds of movements may be detected as with attribution. Although some have insisted on a restrictive reading of substantial involvement and endorse the ICJ's observations in the Nicaragua case,Footnote 95 others have suggested that less than substantial involvement by a state may suffice to invoke self-defence against that state. Otherwise, a certain trend can be seen to the effect that awareness of the armed activities of non-state actors is required, and that more than mere presence on the territory of a state, or even their toleration, is required to invoke self-defence against that state.
Scholarship seems to have moved away from restrictivist positions with respect to the attribution of armed activities of non-state actors or a state's substantial involvement therein, although some pockets of resistance may be noticed. However, it would be a mistake to qualify all those arguing changes in the law to be expansionists, since quite often the legal constructions put forward attest to a concern to narrow the opportunities for states to invoke self-defence in relation to armed activities of non-state actors. Moreover, to the extent that authors argue changes of the law, these are contingent on their reading of examples from state practice, the justifications put forward by victim states and responses by other states. This brings us to the next section.
4. (Subsequent) practice, agreement of the parties, and acceptance as law
In deploying their reasoning, authors tend to make use of arguments derived from examples from practice and responses thereto. As such, they assert, expressly or implicitly, either that such practice and responses are relevant to determine a particular interpretation of Article 51 of the UN Charter using Articles 31(3)(b) or (3)(c) of the VCLT, or that these evidence the content of the customary rule of self-defence in international law (Article 38(1)(b) of the Statute of the ICJ).Footnote 96 This section will consider the way in which authors have referenced practice and responses in terms of method of interpretation or as evidence of a rule of customary international law.
In the context of the 9/11 attacks,Footnote 97 Security Council Resolutions 1368 and 1373 have been commented upon for their recognition of the right of self-defence.Footnote 98 A core conclusion, not as such assailed by anybody, is that the Council recognized the right of self-defence in relation to the 9/11 attacks,Footnote 99 but serious disagreement exists as to the (legal) foundation on which this rested. Subsequent to the 9/11 attacks, the US and the UK invoked the right to self-defence and reported their actions to the Security Council. The former claimed that al-Qaeda had a central role in the attacks and was supported by the Taliban regime in Afghanistan, whereas the latter invoked also individual self-defence.Footnote 100 Immediate responses to the 9/11 attacks, and the actual assistance given to the US, have been invoked to support the idea that the right to self-defence could not only be invoked, but this right could also be exercised with respect to Afghanistan for its harbouring and/or support for al-Qaeda, or in relation to al-Qaeda independently from the involvement of Afghanistan in the attacks themselves.Footnote 101
A thorough study by Kreß came out in 1995, discussing in detail the practice of states and their involvement in the armed activities of non-state actors.Footnote 102 Since the 9/11 attacks, various other authors have followed suit. Studies investigating later incidents and situations include: Russian incursions into Georgia in 2001 and 2002 to combat Chechen terrorists;Footnote 103 an Israeli bombardment of a Hezbollah camp in Syria in 2003 in response to an attack on Haifa;Footnote 104 the Rwandan military activities in 2004 in the Democratic Republic of Congo (DRC) in response to attacks by former militia involved in the genocide;Footnote 105 the Ugandan threat of cross-border action in the DRC in response to activities of the Lord's Resistance Army in 2005;Footnote 106 the Ethiopian intervention in Somalia in response to infiltrations of the Union of Islamic Courts in 2006;Footnote 107 the Israeli intervention in Lebanon in response to military acts of Hezbollah in 2006;Footnote 108 Israeli armed interventions in the Gaza strip in 2004 and 2008;Footnote 109 the Turkish military interventions in north Iraq in 2007 and 2008 against the Kurdistan Workers’ Party (PKK);Footnote 110 the Colombian military crossing the border with Ecuador to combat terrorists belonging to the Revolutionary Armed Forces of Columbia (FARC) in 2008;Footnote 111 and the US cross-border drone strikes in Pakistan in response to Taleban and al-Qaeda attacks in 2008–2009.Footnote 112
Regarding the 9/11 attacks and examples from later practice, the legal justifications claimed and the responses by other states, the following topics will be raised consecutively: the interpretation of Resolutions 1368 and 1373; the scope and relevance of practice; the legal justifications put forward and their connection to attribution and substantial involvement; and the assessment of responses.
Restrictivist reasoning as to the interpretation of Resolutions 1368 and 1373 has focused on language, location and competence. As to language, the generality of the references to self-defence have been noted,Footnote 113 along with a certain ambiguity, because the 9/11 attacks were qualified as a threat to peace rather than an armed attack.Footnote 114 Corten asserts that the Resolutions do not challenge or change the law, since self-defence is only recognized ‘in accordance with the Charter’.Footnote 115 Regarding location, it has been observed that the references to self-defence are contained in the preambles rather than the operative paragraphs.Footnote 116 More fundamentally, the Security Council's competence to provide some kind of authorized interpretation has been questioned.Footnote 117 Moreover, as the Security Council does not purport to lay down rules of customary international law, its resolutions must be subjected to the same kind of scrutiny as General Assembly resolutions, and the legal opinions of all states must be carefully analysed.Footnote 118
With respect to the interpretation of Security Council resolutions, few references are made to the observations of the ICJ in the Namibia opinion or to the ICJ's elaboration in the Kosovo opinion.Footnote 119 Questions regarding the relevance of Security Council resolutions to the determination of a rule of customary international law have been raised especially in view of the Council's limited composition, and the need for confirmation by the wider international community is stressed.Footnote 120 The ICJ's cautionary tale in the Nicaragua case and Nuclear Weapons opinion,Footnote 121 regarding the way in which the contribution of General Assembly resolutions to customary international law must be assessed, is infrequently recounted.
Moving to state practice, authors frequently reference the Caroline incident of 1837 invoking it as an example of self-defence against armed activities of non-state actors.Footnote 122 In this respect, only a few authors suggest its limited relevance, with Antonopoulos decrying the near ‘theological reverence’ for Webster's formula and noting that self-defence in those days was not ‘an exception in law to a prohibition established by law’.Footnote 123 As such, his view is that this deprives pre-Charter practice of any significance.Footnote 124
As to the extent of practice required to establish a rule, the claim has been made that the Bush doctrine, asserting the right of self-defence against states that harbour terrorists, led to a rule of instant customary law.Footnote 125 Opposed to this is the view that one particular incident cannot create a rule.Footnote 126 The ICJ's avowal in the North Sea Continental Shelf cases that a short period of time is not necessarily a bar to the creation of a rule is referenced,Footnote 127 but some add the Court's restrictive qualification that within such a period of time practice must have been ‘extensive and virtually uniform’.Footnote 128 The number of incidents or situations in which self-defence is invoked in relation to armed activities of non-state actors has been said to be on the rise when compared to practice before the 9/11 attacks.Footnote 129
Authors generally do not reference relevant cases, such as the Asylum and Nicaragua cases, with the former requiring a ‘constant and uniform usage’ and the latter indicating that only a general practice is required, which does not need to be in ‘absolutely rigorous conformity with the rule’.Footnote 130 As to the breadth of practice, it may be noted that overall, authors do not attempt to quantify the extent of state practice required to establish a rule of customary international law, instead preferring to note trends. Another question, whether the practice required to establish an exception could be less extensive than that needed to found the general rule, appears not to have been raised.
Regarding the relevance of practice, the legal justification put forward by the state invoking self-defence is considered crucial. van Steenberghe is singularly explicit in this respect, by positing that state practice must constitute an application of the law on self-defence, and he investigates legal justifications (not) invoked to show their relevance to assess an evolution of the law.Footnote 131 Others discount certain examples for examining whether attribution is no longer relevant, since the state invoking self-defence precisely premised its claim on attribution of the armed activities of the non-state actor concerned.Footnote 132
Speaking in relation to non-intervention, the ICJ accepted that customary international law might change as a result of new claims, and it held:Footnote 133
[t]he significance for the Court of cases of State conduct prima facie inconsistent with the principle of non-intervention lies in the nature of the ground offered as justification. Reliance by a State on a novel right or an unprecedented exception to the principle might, if shared in principle by other States, tend towards a modification of customary international law.
Clearly then, the broadening of the right to self-defence is possible too, whether through interpretation or customary international law. However, the ICJ also observed, although rarely referenced, that declarations made by states at times constitute ‘statements of policy, and not an assertion of existing rules of international law’, and spoke of intervention justified ‘on the political level’ but not ‘on the legal level’.Footnote 134
Attribution is generally discussed in relation to examples from practice, because it is necessary to determine whether a precedent that supports self-defence beyond the stricter confines of armed attacks by states exists.Footnote 135 With respect to the 9/11 attacks and substantial involvement, the close links that existed between the Taleban and al-Qaeda have been noted, with Corten qualifying this as more than ‘mere tolerance and diffuse support’.Footnote 136 Ruys contemplates that the required nexus between non-state actors and states might not have become redundant, considering the emphasis on the Taleban's harbouring and support of al-Qaeda.Footnote 137 Similarly, regarding the 2006 Israeli actions against Lebanon, the former pointed to the close ties of Hezbollah to, and forming part of, the Lebanese government and considering self-defence to be applicable for that reason.Footnote 138 The latter generally denied attribution, but noted that Hezbollah constitutes somewhat of a state within a state, with practically exclusive control over the south of Lebanon, which might allow attribution under Article 9 of ASR.Footnote 139
In assessing the responses of other states in terms of agreement to any particular interpretation or opinio juris, restrictivist reasoning challenges the extent of support for self-defence actions, the qualification and implications of support or condemnations of action allegedly taken in self-defence, and the assessment of silence either in support or against a claim of self-defence. To start with the extent of support, the statement of the Organization of the Islamic Conference (OIC), consisting of 57 member states, could not be seen to show support for the legal doctrine propounded by the US in response to the 9/11 attacks.Footnote 140 With respect to the 2003 Israeli action against Hezbollah in Syria, it has been noted that general support for a wide right to use force against terrorist camps in another state was lacking.Footnote 141
The nature of support or condemnation has drawn diverse restrictivist comments. First, the argument has been made that some support voiced for the US action in self-defence after the 9/11 attacks was political in nature and could not be (easily) qualified in legal terms.Footnote 142 The emotionally charged atmosphere surrounding the 9/11 attacks may have led states to respond out of emotion rather than legal conviction.Footnote 143 Second, condemnations are sometimes said to be based on grounds other than the petitio principii of self-defence against an armed attack by a non-state actor, such as lack of evidence, lack of necessity, or disproportionality of action in self-defence.Footnote 144 Implicit in this argument is the inference that states which condemn on other grounds support the principled claim of the state invoking self-defence.Footnote 145 In relation to the 2006 Israeli actions against Lebanon, restrictivists retorted that it was unclear whether states based their condemnations exclusively on disproportionality, or whether they also rejected the Israeli claim as a matter of principle.Footnote 146
Silence comes into the debate in different voices. Thus, lack of condemnation is at times argued to support a purported rule.Footnote 147 Restrictivist authors respond by pointing to factors that may explain lack of condemnation. In the case of the 9/11 attacks, for instance, the circumstance, condemned by the Security Council before the attacks,Footnote 148 that Afghanistan had allowed its territory to be used for the export of terrorism by al-Qaeda.Footnote 149 With respect to the Turkish incursions into northern Iraq, Ruys noted that states generally followed the lead of the US and the EU not to take any clear legal position, with muted criticism due to the limited objectives and proportionality of the operations concerned.Footnote 150 Although he considers that this may support an interpretation that a nexus between a non-state actor and the territorial state is not required in situations where the latter is unwilling or unable to prevent attacks, he warns that, in view of the lack of legal scrutiny by the international community, ‘acquiescence is a fickle barometer of opinio iuris’.Footnote 151 With respect to other incidents or situations, it has been noted that at times, when the facts remained unclear, intervening states refrained from expressing their legal claim, and international legal scrutiny was minimal or absent.Footnote 152 More generally, the interpretation of precedents may prove difficult because of divisions among states and imprecise discourse,Footnote 153 that states ‘frequently invoke only political or moral considerations’ or mix them up with legal ones, and at times remain silent on the issue of legal justification.Footnote 154
In assessing the use by authors of practice and responses in relation to the armed activities of non-state actors and the right of self-defence, a preliminary remark is that at times no clear indication is given as to the source investigated or the purpose of discussion of practice and responses.Footnote 155 Nevertheless, it does appear that most authors set forth their analysis or make arguments with a view to establishing the content of a rule of customary international law. In doing so they generally do not reference such cases or opinions of the ICJ as are pertinent to assess the weight to be given to examples from practice, the legal justification offered and responses thereto,Footnote 156 and many do not attempt to establish their method of investigation.Footnote 157
Some authors use practice and responses to interpret the UN Charter from the perspective of ‘relevant rules of international law applicable in the relations between the parties’ pursuant to Article 31(3)(c) of the VCLT.Footnote 158 Others refer to ‘subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’ pursuant to Article 31(3)(b) of the VCLT.Footnote 159 However, as with customary international law, little attempt is made to set out their understanding of the limitations on the use or weight of such elements of interpretation.Footnote 160
The choice of element of interpretation and source is important because different standards apply for the assessment of practice and responses to establish an interpretation or to establish a rule of customary international law.Footnote 161 For purposes of interpretation, Article 31(3)(b) of the VCLT may require practice by only some of the parties, but it does require agreement by all of them.Footnote 162 The use of Article 31(3)(c) of the VCLT calls for a renvoi to customary international law and might require more extensive practice but probably not (explicit) acceptance by all States. Furthermore, Article 31(3)(b) VCLT only requires agreement, otherwise unqualified, whereas under Article 38(1)(b) of the ICJ Statute customary international law demands that the practice be accepted as law.
Authors tend to take three steps in assessing the contribution of practice and its acceptance as law to change customary international law. The first step is to identify and investigate examples from practice. The second step is to examine the claim or justification offered by the state invoking self-defence. The (ir)relevance of practice is wound up with that claim or justification and constitutes opinio juris of the state(s) invoking self-defence.Footnote 163 This then ties up with the third step, namely to assess the responses of both the target state and other states in response to the claim or justification offered.Footnote 164
Sometimes the discussion is structured into timeframes, most commonly the Cold War, post-Cold War and post-9/11 periods.Footnote 165 In many contributions, responses of states are assessed and quantified by reference to support or acceptance of self-defence claims, condemnation on other grounds than the petitio principii of action in self-defence, and lack of condemnation or criticism. Such arguments amount, quite often, to a claim of ‘silence implies consent’ (qui tacet consentire videtur).
In many of the contributions surveyed, restrictivist arguments have tended to be snippets of wisdom divorced, more often than not, from any broader discussion of sources and method. This may be contrasted with the monographs by Corten, Ruys and van Steenberghe whose main contribution does not lie with their substantive analysis of practice and responses per se, but with the thorough and extensive accounting of their approach. Thus, all three have sizable chapters on sources and method, in which they set out the standard(s) by which they intend to assess the weight of practice and responses.Footnote 166 Although one cannot expect similar extensive treatment in contributions such as articles in journals or chapters in books, occasionally one does find an exemplary combination of an account of method and analysis.Footnote 167
5. Concluding observations: looking ahead and the road ahead
Corten, the restrictivist pur sang, comes to the conclusion that self-defence may only be invoked in response to armed attacks by states.Footnote 168 Attribution of armed activities of non-state actors is possible only if they are completely dependent on the state or when acting under the instructions, directions or control of the state.Footnote 169 The main conclusion by Ruys, rather more cautiously, is that de lege lata there is uncertainty as to the state of the law and he concludes, on that basis, that self-defence against attacks by non-state actors is ‘not unambiguously illegal’ in situations falling below the Nicaragua threshold.Footnote 170
Be that as it may, the next stage in the conflict between restrictivists and expansionists will no doubt be the current crisis, with the US and others invoking the right to self-defence as a response to the armed activities of the Islamic State in Iraq and the Levant (ISIL). In its letter to the Security Council, the US specifically asserts that states must be able to defend themselves when the government of a state where the threat is located is unwilling or unable to prevent the use of its territory for attacks.Footnote 171 One would hope that international legal scholars would at least note the discrepancy between the US claim that Iraq requested it to strike ISIL in Syria and the request by Iraq to the US ‘to strike ISIL sites and military strongholds, with our express consent’.Footnote 172 Quite clearly, Iraqi consent only operates in relation to Iraqi territory.
But all will be in vain unless international legal scholars can come to some kind of understanding on appropriate rules of engagement. The current projects underway at the ILC – on subsequent agreements and subsequent practice in relation to treaty interpretation and on the formation of customary international law –Footnote 173 are therefore not merely opportune but indispensable: if only to avoid an argumentative freedom on the dancefloor, or battlefield, of treaty interpretation and customary international law where anything goes.Footnote 174
