If the question whether only states can commit an armed attack is answered in the affirmative, the issue becomes what connection non-state entitites ought to have to a state to call their actions an armed attack. This question has been discussed under the heading of ‘attribution’ or ‘imputability’ and the law of state responsibility has been used as a fount of ‘normative ideas’ to discover what norm governs this specific case. And here we have an issue that does not simply divide legal cultures, but one of immense importance, great reach and great uncertainty. It also is an incredibly complex issue, made such through what the International Law Commission has called ‘the fragmentation of international
186 Dinstein (2005) supra note 9 at 245.
187 In the advisory opinion of 9 June 2004 the International Court of Justice held that the right of self-defence under Article 51 UN Charter is confined to inter-state relations. Wall (2004) supra note 53 at 194 (para 139): ‘Article 51 of the Charter thus recognises the existence of an inherent right of self-defence in the case of armed attack by one State against another State. However, Israel does not claim that the attacks against it are imputable to a foreign State. . . . Consequently, the Court concludes that Article 51 of the Charter has no relevance in this case’ (emphasis added).
Several judges did not share the Court’s view of the inter-state character of the right of self-defence in international law (Wall (2004) supra note 53, Separate Opinion Higgins at 215 ( paras 33–34), Separate Opinion Kooijmans at 229–230 (paras 35–36), Declaration Buergenthal at 241–243 (paras 4–6). In Armed Activities the Court again seemed to require attributability of private acts to the state against which self-defence was claimed to have been exercised:
‘[t]he attacks did not emanate from armed bands or irregulars sent by the DRC or on behalf of the DRC’; Armed Activities (2005) supra note 26 at 223 (para 146).
2.4.2 Self-defence 43
law’.188 This complex of problems will be approached by way of a dissection, starting with a look at the two groups of approaches on their own merit and proceeding to a discussion of how and why the two ‘worlds’ conflict and what, if any, synthesis might be drawn.
(a) The ‘traditionalist’ debate is concerned with the specific question of under what circumstances a state is allowed to resort to forcible means in self-defence – when an armed attack can be said to have occurred – against ‘action by armed bands’.189 The fountainhead of traditional views on the topic can be seen in the Caroline incident.190 However, for the present purposes scholars are not primarily interested in the formulation of the doctrine of self-defence by Webster and Ashburton, but the fact that they classified the incident as an instance of self- defence.191 Because the incident involved armed bands, and because self-defence was claimed and not disputed, and because the attacker was not the United States itself, but private individuals, the inference is drawn that self-defence properly so called may involve action against private individuals.
This traditional debate is also remarkable in other respects. First, while a state is not normally responsible for either its citizens’ actions or for other persons staying on its territory,192 a consistent application of the facts of the Caroline incident as if they were an example of lawful self-defence would mean that a state may use force in self-defence even when the ‘base state’ is powerless to act against the individuals concerned, thus imposing a sort of absolute duty to tolerate other states’ actions against it. Second, it may be noted that since the International Law Commission (ILC) first started laying down rules for the attribution of private behaviour to states,193 and since the International Court of Justice’s judgment in the Nicaragua case,194 the terms of the debate have shifted markedly towards a differing standard (described below) connected to the law on state responsibility.195
188 Fragmentation of international law: difficulties arising from the diversification and expansion of international law, in: International Law Commission, Report on the work of its fifty-eighth session (1 May to 9 June and 3 July to 11 August 2006) (A/61/10) (2006) 400–423.
189 For the origins of the concept of ‘armed bands’, encompassing a whole host of phenomena, from uncontrolled tribal groups to terrorists, see Ian Brownlie, International law and the activities of armed bands, 7 International and Comparative Law Quarterly (1958) 712–735 at 713–719.
190 See generally: Robert Y. Jennings, The Caroline and McLeod cases, 32 American Journal of Inter- national Law (1938) 82–99.
191 Dinstein (2005) supra note 9 at 248–249; Contra: Ago (1982) supra note 9 at 39–40 (para 57), 65–66 (para 113).
192 That is, a state may violate a duty to prevent individuals from acting, but an individual’s acts them- selves are not an ‘act of state’. Jennings and Watts (1992) supra note 22 at 549 (para 166). See below.
193 Roberto Ago’s Third Report on State Responsibility, proposing what later became Article 8(a) DASR 1996 and Article 8 ARS 2001, was issued in 1971. Roberto Ago, Third report on state responsibility [A.CN.4/246, A.CN.4/246/Add.1–3], 23 Yearbook of the International Law Commission 1971 (1973) Volume II, Part One, 262–267.
194 Nicaragua (1986) supra note 5.
195 Francis A. Boyle, Determining U.S. responsibility for Contra operations under international law.
Appraisals of the ICJ’s decision: Nicaragua v. United States (Merits), 81 American Journal of International Law (1987) 86–93 at 87; Farer (1987) supra note 100 at 113; Higginbotham (1987) supra note 60 at 548–549, 550; Schindler (1986) supra note 138 at 35.
Uncertainty in International Law 44
Third, those seeing the facts of the Caroline incident as an example of self-defence and positing a low standard of connection between state and individual are more likely to be earlier authors,196 although the events of 11 September 2001 have created a revival of the traditional standard.197
What were the proposed standards under the traditional debate and was there a consensus view? It is to be noted that views that do not require a link between state and private entity, whether ‘traditional’ or not, have been discussed else- where (Section 2.4.1) and thus will not be mentioned again here. In contrast, views considering some sort of state ‘complicity’198 or involvement to be necessary will be addressed here. Ian Brownlie’s words, written in 1958, may be taken as typical:
However, it is conceivable that a co-ordinated and general campaign by powerful bands of irregulars, with obvious or easily proven complicity of the government of the State from which they operate, would constitute an ‘armed attack’, more especially if the object was the forcible settlement of a dispute or the acquisition of territory.199
There is also a diffuse, if later, trend to completely exclude certain categories of state action. ‘Assistance’, it is said, merely providing supplies to an armed group, cannot conceivably be sufficient to fulfil the link-criterion of armed attack in the case of armed bands. The Court’s pronouncements in Nicaragua can be read independently of questions of attribution under state responsibility and as merely being concerned with defining what an ‘armed attack’ under Article 51 is.200 Higginbotham, for example, draws a clear distinction between assistance, on the one hand, and control, on the other hand.201 The words of Robert Jennings, in his dissenting opinion to the above-cited judgment, make one aware, however, that there is a problem of confounding a connection with the proof of its existence:
It may readily be agreed that the mere provision of arms cannot be said to amount to an armed attack. But the provision of arms may, nevertheless, be a very important
196 Brownlie (1958) supra note 189 at 731; Kaikobad (1993) supra note 28 at 311–314; Miller (1985) supra note 29 at 58.
197 Randelzhofer (2002b) supra note 28 at 801–802 (MN 33–36); Ratner (2002) supra note 80 at 908.
The views of those supporting the proposition discussed in Section 2.4.1 often come close to these voices and scholars’ writings are often difficult to pinpoint, both with respect to what precise
‘dimension’ of armed attack they discuss and what precise position they take. Vide Jonathan I.
Charney, The use of force against terrorism and international law, 95 American Journal of International Law (2001) 835–839 at 836 as an example.
198 The term ‘complicity’ is used here in a very wide sense, not merely as a ‘complicity doctrine’.
Cf. Kreò (1995) supra note 41 at 150–151.
199 Brownlie (1958) supra note 189 at 731 (emphasis added).
200 Contrast Nicaragua (1986) supra note 5 at 103–104 (para 195), 119 (para 230), 126–127 (para 247) with the discussion, in the same judgment, of questions of attributability: Nicaragua (1986) supra note 5 at 62 (para 109).
201 Higginbotham (1987) supra note 60 at 549–550.
2.4.2 Self-defence 45
element in what might be thought to amount to armed attack, where it is coupled with other kinds of involvement.202
Judge Jennings is not contradicting himself, for while the majority of judges saw assistance as necessarily excluded (this kind of activity cannot constitute the link required), he considered ‘assistance’ as one kind of proof of such link. The link would be established if proof of supply of arms were supplemented with proof of other involvement.
(b) The starting point of the newer line of argument to connect private attacks to a state is the law of state responsibility. This, the ‘international law of torts’, has had to deal with linking the state to human behaviour, because a state as a juridical person acts through human behaviour. The juridical link between real behaviour and legal person is commonly called ‘attribution’ or
‘imputability’. Within the International Law Commission’s project elaborate rules were developed, inter alia for our kind of link, the attribution of acts of persons or groups which are not formally organs of that state. While it is not necessary and impossible for this chapter to go into detail concerning the law of state responsibility, the most relevant elements of the debate will be summarised here.
An individual’s or a group’s actions can be attributed to a state – a state may be held responsible for these actions just as if its own organs had committed it – if these private entities were ‘in fact acting on behalf of’203 that state (‘agents’ or
‘agencies’).204 The abstract standard that the International Law Commission has codified in Article 8 ARS 2001 and that is generally accepted by international lawyers and judges (nota bene: as concerns state responsibility) is one of ‘control’.
Article 8 reads:
The conduct of a person or group of persons shall be considered an act of a State under international law if the person or group of persons is in fact acting on the instructions of, or under the direction or control of, that State in carrying out the conduct.205
Two arguments make the ‘Article 8 approach’ relevant to the debate of the present subsection. First, many scholars expressly import the language of control or even Article 8 itself into the debate on armed attack against armed bands.206 Second, the notion of ‘control’ has been adopted by the Court in Nicaragua207 as
202 Nicaragua (1986) supra note 5, Dissenting Opinion Jennings at 543.
203 Article 8(a) DASR 1996.
204 André J.J. de Hoogh, Articles 4 and 8 of the 2001 ILC Articles on State Responsibility, the Tadic´
case and attribution of acts of Bosnian Serb authorities to the Federal Republic of Yugoslavia, 72 British Year Book of International Law 2001 (2002) 255–292 at 268.
205 Article 8 ARS 2001 (emphasis added).
206 Boyle (1987) supra note 195 at 87; Farer (1987) supra note 100 at 113; Higginbotham (1987) supra note 60 at 548–549, 550; Schindler (1986) supra note 138 at 35. Also: Kreò (1995) supra note 41 at 149–150.
207 Nicaragua (1986) supra note 5 at 62 (para 109).
Uncertainty in International Law 46
well as by other tribunals208 as the generally applicable standard.209 But the real question, the most relevant source of uncertainty, arises not from the construction of either the traditional or the ‘state responsibility’ approaches, but from their synthesis or confrontation. Are the two standards materially the same, or are they incompatible? Are they two different norms and how do they conflict? This situ- ation seems to constitute a paradigm case of lex specialis and lex generalis; it seems that here two different regimes govern the same set of circumstances and that the norms of the first regime proscribe behaviour different from that of the second regime.210 We seem to have a conflict of norms.
An ‘armed attack’ is not as such illegal: Article 51 does not prohibit it. It is merely a condition for a threat or use of force to be justified (allowed) as self- defence. Therefore, an armed attack itself does not entail state responsibility.
Thus, since no ‘breach of an international obligation’ has occurred, there is no need to attribute the actions of natural persons to a state211 and the standard developed in Article 8 ARS 2001 is not applicable as such.
However, the majority opinion requires a threat or use of force in order for an armed attack to occur. It needs the acts constituting an ‘armed attack’ to fulfil the actus reus condition of Article 2(4), which, in turn, is a breach of an international obligation and requires attribution. However, the question remains whether ‘threat or use of force’ is equivalent with ‘armed attack’ or whether a breach of Article 2(4) is merely one precondition. Thus the violation of the prohibition of force is seen as a necessary condition, but not an integral part of an armed attack.212 Furthermore, the person who advocates making ‘armed attack’
illegal in this roundabout way will have to contradict the Court’s reasoning in Nicaragua:
208 The substantive standard employed varies slightly between different tribunals: ‘on the occasion . . . the militants acted on behalf of the State, having been charged . . . to carry out a specific operation’; United States Diplomatic and Consular Staff in Tehran (United States of America v. Iran), Judgment of 24 May 1980, ICJ Reports (1980) 3 at 29 (para 58) (emphasis added); ‘actually exercised such a degree of control in all fields as to justify treating the contras as acting on its behalf ’; Nicaragua (1986) supra note 5 at 52 (para 109) (emphasis added); effective control Nicaragua (1986) supra note 5 at 53–55 (paras 113–116); effective overall control (ECHR, Loizidou v. Turkey, Merits, Judgment of 18 December 1996, ECHR Reports (1996-VI) 2216 at 2235–2236 (para 56)); overall control (ICTY, Prosecutor v. Dusˇko Tadic´, Appeals Chamber, Judgment of 15 July 1999 AC (1999), at: www.icty.org (para 145)). For an excellent article comparing differing judicial views on the Article 8 standard see de Hoogh (2002) supra note 204.
209 Contrast de Hoogh (2002) supra note 204, with Kreò (1995) supra note 41.
210 It seems somewhat artificial to construe the difference here as different ‘prescribed behaviour’ of subjects of law. However, in this case there would be a difference in prescribed behaviour: if action against certain armed bands were classified as self-defence, then the ‘victim state’ could legally threaten or use force (within the limits set by self-defence law), whereas if such action were not so classified, the ‘victim state’ ought neither to threaten nor use force. Therefore, the difference is visible in theory and practice.
211 Article 2(b) and Article 2(a) ARS 2001, respectively.
212 In logical terms: force⇒armed attack, but not: force≡armed attack.
2.4.2 Self-defence 47
But the Court does not believe that the concept of ‘armed attack’ includes not only acts by armed bands where such acts occur on a significant scale but also assistance to rebels in the form of the provision of weapons or logistical or other support. Such assistance may be regarded as a threat or use of force, or amount to intervention in the internal or external affairs of other States.213
The Court’s argument is that whereas a state act of assisting rebels cannot be an
‘armed attack’, but may well constitute ‘use of force’, the two concepts cannot be identical. If material assistance to rebels as such cannot be considered a breach of the prohibition of force (but might well constitute intervention), the majority’s thesis could at least remain internally consistent. This would come at the price of having to contradict the Court.
One could speak of the question of indirect attack along the lines of a two-step process. First, the act is attributed to a state; second, the gravity of the act amounts to an armed attack ratione materiae. This would mean that – once the ‘one- size-fits-all’ criterion of attribution is fulfilled – it depends entirely on the private entity’s behaviour what the act will be classified as. Against this the Court seems to hold that there should be a stronger connection vis-à-vis ‘armed attack’ than vis-à- vis ‘intervention’ or ‘threat or use of force’. According to the Court it seems the state must do less to have a private act considered as its own intervention than as its own aggression or armed attack. In Nicaragua there obviously is a connection between the degree of involvement of the state and the acts it is held responsible for. If the connection is of a certain kind and scale, the insurgents’ acts may be armed attacks, but if it is not, the attribution may only result in a threat or use of force or intervention. However, attribution is not gradual: an act either is an act of a state or it is not and it seems impossible to prescribe differing degrees of connection for every single type of wrong in international law.
There are two arguments that might show that the ‘conflict’ of rules, as the situation could tentatively be described, might not exist at all. What if the differ- ence between the two regimes is neither a definition of a lex specialis standard on the attribution of private actions to a state for the purposes of Article 51, nor a further standard in addition to Article 8 ARS? What if the ‘specialis approach’
merely defines the Article 8 standard within the factual context of aid to insur- gents or guerrillas as a ‘concrete standard’? Maybe one could speak of the same standard manifested in different acts within different situations or legal regimes (e.g. ius ad bellum, international humanitarian law, human rights law or classical
‘international tort law’) without it being a lex specialis? The standard of ‘direction or control’ would be the same in all areas, but the acts required to establish either direction or control by a state would be different.
On the other hand, we might follow the hint dropped by Judge Ruda in his separate opinion to Nicaragua:
213 Nicaragua (1986) supra note 5 at 103–104 (para 195) (emphasis added). Higginbotham (1987) supra note 60 at 546.
Uncertainty in International Law 48
From my point of view it would have been sufficient to say, just as the Court does in its conclusions, that even if there was such assistance and flow of arms, that is not a sufficient excuse for invoking self-defence, because, juridically, the concept of ‘armed attack’ does not include assistance to rebels.214
The implication is that the question of whether assistance to rebels can be con- sidered an armed attack is one of means (Section 2.2.2) and that ‘armed attack’
lacks a ratione personae dimension. This would mean a renvoi of the whole matter to state responsibility. Whether an act can be attributed to a state seems irrelevant; the only valid question is whether the act can be called an armed attack.215 This argument leads to the minority view portrayed in Section 2.4.1 and is vulnerable to the counter-arguments presented there. Here we have yet another manifestation of uncertainty. Two bodies of norms happen to be applicable to the same situation where the two sets of do not match exactly. Even if the content of the norms were the same, it would be two different norms stipulating the same ideal (see Chapter 5).