Ch.VII Action with Respect to Threats to the Peace, Breaches of the Peace, and Acts of Aggression, Article 51
Georg Nolte, Albrecht Randelzhofer
Edited By: Bruno Simma, Daniel-Erasmus Khan, Georg Nolte, Andreas Paulus, Nikolai Wessendorf (Assistant Editor)
- UN Charter — United Nations (UN) — Armed attack — Self-defence — International peace and security
Nothing in the present Charter shall impair the inherent right of individual or collective self-defence if an armed attack occurs against a Member of the United Nations, until the Security Council has taken the measures necessary to maintain international peace and security. Measures taken by Members in the exercise of this right of self-defence shall be immediately reported to the Security Council and shall not in any way affect the authority and responsibility of the Security Council under the present Charter to take at any time such action as it deems necessary in order to maintain or restore international peace and security.
See also the select bibliography on Art. 2 (4).
- Bothe M, ‘Die Erklärung der Generalversammlung der Vereinten Nationen über die Definition der Aggression’ (1975) 18 JIR 127.
- ——— ‘Terrorism and the Legality of Pre-emptive Force’ (2003) 13 EJIL 227.
- Cannizzaro E, ‘Contextualizing Proportionality: Jus Ad Bellum and Jus In Bello in the Lebanese War’ (2006) 864 IRRC 779.
- Cassin V, Debevoise W, Kailes H, and Thompson T, ‘The Definition of Aggression’ (1975) 16 Harv Intl LJ 589.
- (p. 1398) Chainoglou K, Reconceptualising the Law of Self-Defence (Ant N Sakkoulas, Bruylant 2008).
- Constantinou A, The Right of Self-Defence under Customary International Law and Article 51 of the UN Charter (Ant N Sakkoulas, Bruylant 2000).
- Corten O, The Law Against War: The Prohibition on the Use of Force in Contemporary International Law (Hart 2010).
- Dinstein Y, War, Aggression and Self-Defence (5th edn, CUP 2012).
- Fassbender B, ‘Selbstverteidigung und Staatengemeinschaftsinteresse in der Zeit des Völkerbundes’ in T Bruha, S Heselhaus, and T Marauhn (eds), Legalität, Legitimität und Moral (Mohr Siebeck 2008) 99.
- Franck TM, ‘On Proportionality of Countermeasures in International Law’ (2008) 102 AJIL 715.
- Gardam J, Necessity, Proportionality and the Use of Force by States (CUP 2004).
- Gazzini T, The Changing Rules on the Use of Force in International Law (Juris Publishing 2005).
- Gill TD, ‘The Law of Armed Attack in the Context of the Nicaragua Case’ (1988) 1 Hague YB Intl L 30.
- Gray C, International Law and the Use of Force (3rd edn, OUP 2008).
- Green J A, The International Court of Justice and Self-Defence in International Law (Hart 2009).
- Greenwood C, ‘Self-Defence’ MPEPIL (online edn).
- Greig DW, ‘Self-Defence and the Security Council: What Does Article 51 Require?’ (1991) 40 ICLQ 366.
- Kreß C, Gewaltverbot und Selbstverteidigungsrecht nach der Satzung der Vereinten Nationen bei staatlicher Verwicklung in Gewaltakte Privater (Duncker & Humblot 1995).
- Krisch N, Selbstverteidigung und kollektive Sicherheit (Springer 2001).
- Lagoni R, ‘Gewaltverbot, Seekriegsrecht und Schiffahrtsfreiheit im Golfkrieg’ in W Fürst, R Herzog, and DC Umbach (eds), Festschrift für Wolfgang Zeidler, vol 2 (de Gruyter 1987) 1833.
- Lubell N, Extraterritorial Use of Force Against Non-State Actors (OUP 2010).
- McCormack TLH, Self-Defense in International Law: The Israeli Raid on the Iraqi Reactor (Palgrave Macmillan 1996).
- Murphy SD, ‘Terrorism and the Concept of “Armed Attack” in Article 51 of the U.N. Charter’ (2002) 43 Harv Intl LJ 41.
- ——— ‘Self-Defense and the Israeli Wall Advisory Opinion: An Ipse Dixit from the ICJ?’ (2005) 99 AJIL 62.
- Reisman WM and Armstrong A, ‘The Past and Future of the Claim of Preemptive Self-Defense’ (2006) 100 AJIL 525.
- Ronzitti N, ‘The Expanding Law of Self-Defence’ (2006) 11(3) J Conflict & Security L 343.
- Roscini M, ‘World Wide Warfare—Jus Ad Bellum and the Use of Cyber Force’ (2010) 14 Max Planck YB UN L 85.
- Ruys T, Armed Attack and Article 51 of the UN Charter (CUP 2010).
- Ruys T and Verhoeven S, ‘Attacks by Private Actors and the Right of Self-Defence’ (2005) 10 J Conflict & Security L 289.
- Schachter O, ‘Self-Defense and the Rule of Law’ (1989) 83 AJIL 259.
- Tams C J, ‘Light Treatment of a Complex Problem: The Law of Self-Defence in the Wall Case’ (2006) 16 EJIL 963.
- ——— ‘The Use of Force Against Terrorists’ (2009) 20 EJIL 359.
- Tomuschat C, ‘Der 11. September 2001 und seine rechtlichen Konsequenzen’ (2001) 33 EuGRZ 535.
- Wilmshurst E, ‘The Chatham House Principles of International Law on the Use of Force in Self-Defence’ (2006) 55 ICLQ 963.
- Zimmermann A, ‘The Second Lebanon War: Jus Ad Bellum, Jus In Bello and the Issue of Proportionality’ (2007) 11 Max Planck YB UN L 99.
(p. 1399) A. Evolution and Significance of the Right of Self-Defence*
1 The development of the right of self-defence should be viewed against the background of the development of international law towards the prohibition of war, and, eventually, of the use of force. Until the beginning of the twentieth century the right of self-defence was of only modest significance.1 Public international law permitted States to wage war freely,2 so that no justification for doing so was needed. When, in spite of this, States explicitly invoked their right of self-defence, their purpose was merely political.3 Even so, the right of self-defence played a certain role as a legal justification with regard to hostilities not occurring in a state of war, since the use of force in times of peace was in fact restricted by international law.4 The content and scope of the right of self-defence were relatively unclear, however, and extended well into the sphere of self-help.5
2 At the beginning of the twentieth century, when the freedom to resort to war became more and more restricted,6 the right of self-defence gained in significance. This development first culminated in the conclusion of the Briand-Kellogg Pact in 1928. The general prohibition of war laid down in Art. I of the Pact was subject only to the reservation of the right of self-defence.7 Consequently, it was solely in the exercise of the right of self-defence that war could still be lawful. Various incidents and legal developments subsequent to the entry into force of the UN Charter led to a further increase in the importance of the right of self-defence. Today, the right of individual or collective self-defence is invoked with regard to almost every unilateral use of military force.8
3 The prohibition of the use of force embodied in Art. 2 (4) not only proscribes war, but any use or threat of force. Apart from the now obsolete clauses concerning the former enemy States,9 the UN Charter contains only two exceptions to the prohibition of force, namely SC enforcement actions pursuant to Chapter VII, and the right to individual and collective self-defence laid down in Art. 51. International legal practice since 1945, (p. 1400) contrary to the intentions of the authors of the Charter, has continued to see a significant amount of unilateral use of force by States. Yet in this respect the Charter provides in Art. 51 for a regulation which allows individual States the threat or use of force, in principle, only under the conditions stipulated therein.10 The right of self-defence laid down in Art. 51 of the UN Charter has therefore become the pivotal point upon which disputes concerning the lawfulness of the use of force in interstate relations usually concentrate. Such disputes not only raise questions of legal interpretation, but even more often questions of fact, as shown by several decisions of the ICJ.11
4 The interpretation and the development of Art. 51, like the law on the use of force in general, remain subject to the applicable rules of international law. The UN Charter is to be interpreted according to the general rules on treaty interpretation which are codified in Arts 31–33 VCLT, and which also reflect customary international law, taking duly into account the character of the Charter as the constitutive document for the organization of the international community of States.12 The general rule on treaty interpretation of Art. 31 (3) (b) VCLT requires in particular that ‘any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’ and any ‘established practice of the organisation’ (Art. 2 (f) VCLTIO as a reflection of customary international law) be taken into account. In principle, the rules on treaty interpretation and on the sources of international law do not exclude the possibility that Art. 51 is reinterpreted, including on the basis of subsequent practice.13 Theoretically it is even possible that an additional exception to the general prohibition of the use of force could develop alongside Art. 51 by way of superseding customary international law.14 In view, however, of the fundamental importance of the right to self-defence for the Charter system of collective security, the conditions for the recognition of any significant reinterpretation of, or superseding exception to, Art. 51 are strict. This is true regardless of whether the Charter is seen as possessing a special constitutional status, or as only having the quality of an ordinary treaty. It is therefore not sufficient that one or more ‘incidents’ lead a State or a group of States to assert a certain reinterpretation of Art. 51, or to postulate another exception to the general prohibition of the use of force (Art. 2 (4)), even if such assertions are not immediately, clearly, or broadly opposed by a number of other States. It would rather be necessary that such assertions (p. 1401) of a reinterpretation or a new exception are ‘generally accepted’ or that they constitute an ‘established practice’.
5 A significant amount of current disagreement over the proper interpretation of Art. 51, both among States and among commentators, can ultimately be traced to underlying differences of opinion over the interpretation and application of the rules on the sources of international law, as well as to different assessments of what the respective historical situation (with its characteristic threat scenarios) requires.15 If properly identified, such underlying differences of opinion should not prevent an appropriate interpretation of Art. 51.
6 Arts 51 and 2 (4) do not exactly correspond to one another in scope, ie not every use of force contrary to Art. 2 (4) may be responded to with armed self-defence. The UN Charter did not intend to exclude self-defence entirely, but has restricted its scope considerably. A comparison of the different terms used in the two provisions illustrates that, remaining uncertainties apart, ‘armed attack’ is a narrower notion than ‘threat or use of force’.16 If Art. 51 is thus read in connection with Art. 2(4), the—at first sight—counter-intuitive conclusion is that any State which is affected by another State’s unlawful use of force that does not reach the threshold of an ‘armed attack’, is bound, if not exactly to endure the violation, at least to respond only by means falling short of the use of cross-border force. One possible means of defence is, of course, the use of force by a State on its own territory. While such means may sometimes not be fully effective, this result is intended by the Charter, since the unilateral use of force is meant to be excluded as far as it is bearable for States, in view of the typical dangers of escalation which are connected with mutual uses of armed force. Until an armed attack occurs, States are expected to renounce forcible self-defence. Because of the pre-eminent position of the SC within the Charter system of collective security, the affected State can in that situation merely call upon the SC to qualify the violations of Art. 2 (4) as constituting a breach of the peace and to decide on measures under Chapter VII. Only if and when the prohibited use of force rises to an armed attack can the State concerned unilaterally resort to forcible measures for its defence. But even this authority is limited in two ways: first, the State acting in self-defence must observe the principle of ‘proportionality’;17 secondly, it has to report immediately to the SC the measures taken, and it has to discontinue (p. 1402) them as soon as the latter has itself taken the measures necessary for the maintenance of international peace.
7 The view of the present authors that there exists a gap between Arts 2 (4) and 51 not only corresponds to the prevailing view in international legal writings18 but has also been confirmed by the ICJ’s Nicaragua19 and Oil Platforms20 judgments, as well as, indirectly, in the Congo v Uganda decision.21 It clearly follows from the ICJ’s observations that not every use of force is necessarily to be considered an armed attack.22
8 This conclusion has, as a matter of policy, given rise to objections, for it means that there is not always effective protection against States violating the prohibition of the use of force, as long as they do not resort to an armed attack. Therefore, critics in international legal doctrine have made efforts to deny or to fill the lacuna existing between Arts 2 (4) and 51. Some do not regard Art. 51 as being more restrictive than the term ‘use of force’ in Art. 2 (4)23 and others would permit ‘proportionate defensive measures’ against any smaller-scale ‘unlawful use of force “short of” an armed attack (“agression armée”) within the meaning of Article 51’.24 Such approaches may at first sight appear more satisfactory policywise than the prevailing view. It is true that a State which has violated the prohibition of the use of force has at least provided the primary cause for the response. And it must be admitted that if States are bound to endure acts of force that do not reach the intensity of an armed attack, they may remain devoid of fully effective protection until the SC has taken remedial measures, which may not always come about. These considerations do not, however, override the concern which underlies the prevailing view that the danger of escalation which is inherent in most forms of transboundary uses of force justifies the rule that States deal with small-scale uses of force against them by using force on their own territory or by using non-violent means, and thus in a way which does not involve the use of cross-border force. The concern that an escalation, or even a full-scale war, could be the consequence of a State responding in self-defence to slight uses of armed force, such as a small frontier incident, cannot be dispelled by (p. 1403) responding that force used in self-defence must never be disproportionate25 since the principle of proportionality is not determinate enough to provide sufficiently clear guidance in such situations. The fact that the Charter has not explicitly banned uses of force on a smaller scale does not mean that self-defence against such acts is permissible. On the contrary, it can be claimed equally well that, by deliberately not pronouncing upon the small-scale use of force in the context of self-defence and by allowing self-defence only in response to an armed attack, the UN Charter has expressed the position that no small-scale forcible self-defence is permissible against small-scale uses of force. Thus the critique of the prevailing position merely leads back to the question of whether, and to what extent, there is a broader general right of self-defence apart from Art. 51, but it does not provide an independent foundation for a right to ‘small-scale self-defence’.
9 What has been shown so far is only that Arts 2 (4) and 51 do not entirely coincide in their scope. But whether the UN Charter really does allow for forcible self-defence only in the case of an armed attack also depends on whether and to what extent a general right of self-defence exists apart from Art. 51, and whether Art. 51 is intended to abolish that right, or at least to restrict it as far as possible. This point is the subject of a long-standing controversy.
10 The prevailing view considers Art. 51 to exclude any self-defence, other than that in response to an armed attack, referring, above all, to the purpose of the UN Charter, ie to restrict as far as possible the use of force by individual States.26 According to this view, the designation in Art. 51 of the right of self-defence as ‘inherent’ simply means that the right is also vested in States other than UN members, and that UN members may give assistance to a non-member falling victim to an armed attack.27 This position is supported, though by no means unambiguously, by most commentators28 in that it restricts the right of self-defence to cases in which ‘an armed attack occurs against (p. 1404) a Member of the United Nations’. The view that Art. 51 has superseded any previous customary right of self-defence does not, however, exclude that subsequent practice concerning the scope and the limits of the right of self-defence needs to be taken into account when interpreting the provision.29
11 Another approach regards the customary right to self-defence as not being affected by Art. 51, but rather as having only received a particular emphasis, in a declaratory manner, for the case of an armed attack.30 Above all, this approach is intended to serve as a justification for certain traditional forms of self-defence, even of self-help, in particular cases. For example, the protection of the lives and property of a State’s own nationals abroad,31 and even the forcible protection of certain economic interests in a foreign country,32 have been claimed to be justified on the grounds of self-defence not prohibited by Art. 51. The latter case aptly illustrates that this view of Art. 51 is highly questionable.
12 The content and scope of a customary right of self-defence are unclear and could extend far into the spheres of self-help in such a way that its continuing existence would, to a considerable extent, reintroduce the unilateral use of force by States, the far-reaching abolition of which is intended by the UN Charter.
In the case of individual self-defence, the exercise of this right is subject to the State concerned having been the victim of an armed attack. Reliance on collective self-defence of course does not remove the need for this.35…
The Court has recalled above (paragraphs 193 to 195) that for one State to use force against another, on the ground that that State has committed a wrongful act against a third State, is regarded as lawful, by way of exception, only when the wrongful act provoking the response was an armed attack. Thus the lawfulness of the use of force by a State in response to a wrongful act of which it has not itself been the victim is not admitted when this wrongful act is not an armed attack. In the view of the Court, under international law in force today—whether customary international law or that of the United Nations system—States do not have a right of ‘collective’ armed response to acts which do not constitute an ‘armed attack’.36
… there are two levels to be distinguished: there is, first, the level of ‘armed attacks’ in the substantial, massive sense of amounting to ‘une agression armée’, to quote the French authentic text of Article 51. Against such armed attacks, self-defence in its not infinite, but still considerable, variety would be justified. But we may encounter also a lower level of hostile military action, not reaching the threshold of an ‘armed attack’ within the meaning of Article 51 of the United Nations Charter. Against such hostile acts, a State may of course defend itself, but only within a more limited range and quality of responses (the main difference being that the possibility of collective self-defence does not arise, cf. Nicaragua) and bound to necessity, proportionality and immediacy in time in a particularly strict way.42
While Judge Simma’s position has found some support among commentators, the prevailing view, in particular among States, does not accept this position.43
14 Since the right of self-defence embodied in Art. 51 is restricted to the case of an armed attack, and in the absence of further exceptions to Art. 2 (4) allowing for the use of force by individual States,44 the exercise of force for the enforcement of a vested right (p. 1406) or for the purpose of ending another State’s unlawful behaviour is prohibited.45 Not even arbitral awards or judgments by the ICJ46 may be enforced by means of forcible self-help. It is of particular importance that countermeasures (formerly reprisals), once the most frequently used form of force, are today likewise only admissible in so far as they do not involve the use of armed force.47
15 For example, the firing of 23 cruise missiles, on 26 June 1993, by US warships in the Red Sea and the Persian Gulf upon Iraqi intelligence headquarters in Baghdad, in response to an unsuccessful Iraqi attempt to murder former US President Bush during his visit to Kuwait from 14 to 16 April 1993, could not be justified as an act of self-defence, as US President Clinton claimed on the evening of the raid.48 Even if the attempted assassination of former President Bush were to be qualified as an armed attack, it is quite clear that in this situation, such an armed attack no longer existed more than two months later, nor did the threat of such an attack. On 27 June 1993, Madeleine Albright, US Permanent Representative to the UN also referred to Art. 51 to justify the raid, but added that the raid ‘was designed to damage the terrorist infrastructure and deter further acts of aggression against the United States’.49 This suggests that the raid was intended to be a countermeasure but as such it was not compatible with the existing law governing the use of military force.50
16 The term ‘armed attack’ (French: ‘agression armée’) represents the key notion of the concept of self-defence pursuant to Art. 51. Its interpretation decides how far unilateral use of force is still admissible and is of utmost significance for the effectiveness of the rules of international law on war prevention. The more clear-cut and unambiguous the definition of an ‘armed attack’, the more plausible attempts by States to justify illicit uses (p. 1407) of force as self-defence can be successfully discredited. In addition, since the facts which give rise to an allegation that an armed attack has occurred are mostly disputed51 it is also important to note that the alleged victim State has the burden of proving the existence of an armed attack if it wishes to justify any use of force in self-defence.52
17 The UN Charter uses the terms ‘attack’ and ‘aggression’ in Arts 1 (1), 39, 51, and 53, albeit without defining them precisely. Following futile attempts during the League of Nations era, the UN has been unsuccessfully striving since 1950, first in the ILC, then in four subsequent Special Committees of the GA, for a definition of these terms.53 The adoption of GA Res 3314 (XXIX) (14 December 1974)54 constituted a significant indirect contribution to this effort. The ‘Definition of Aggression’ contained therein, however, apart from constituting a mere recommendation, formally only claims to specify the notion of ‘act of aggression’ as it is embodied in Art. 39 of the Charter, and not that of ‘armed attack’ as used in Art. 51.55 Pursuant to paras 2 and 4 of the Preamble, as well as according to Art. 6, the Definition does not contain an interpretation of the right of self-defence in response to an armed attack.56 The travaux préparatoires of the Definition also confirm that a definition of the notion of ‘armed attack’ was not intended.57 In the Special Committee which worked out the Definition, the United States, supported by other Western States, strongly emphasized that the task of the committee was to elaborate a definition of aggression pursuant to Art. 39 of the UN Charter, without prejudice to other provisions of the Charter, especially Art. 51.58 Like the Soviet Union,59 the US expressed the view that the notions of ‘act of aggression’ and ‘armed attack’ are not identical.60 While it must therefore be assumed that the notions of ‘armed attack’ (‘agression armée’) and ‘act of aggression’ (‘acte d’agression’) do not necessarily coincide fully and (p. 1408) that ‘armed attack’ is the narrower concept of the two,61 the difference between the two is so small that it is often overlooked.62
18 The jurisprudence of the ICJ has brought about only a modest amount of clarification in this respect. On the term ‘armed attack’,63 the ICJ simply remarked in the Nicaragua judgment that ‘[t]here appears now to be general agreement on the nature of the acts which can be treated as constituting armed attacks’.64 This statement, albeit fairly surprising in view of the unsuccessful previous efforts towards agreeing on a definition of ‘armed attack’, has not led the ICJ to provide a more specific definition of the term. Instead, the ICJ merely gave one example in order to illustrate the existence of an armed attack in a specific situation and stipulated that the notion not only comprises unspecified cross-border actions by regular forces, but also the participation of a State in the use of force by unofficial armed bands, as described in Art. 3 (g) of the Definition of Aggression.65 The Court reaffirmed this approach in 2005 in its Congo v Uganda judgment.66 Such use by the ICJ of GA Res 3314 demonstrates the (indirect) importance of the definition contained therein for the purpose of defining the concept of armed attack. In its Oil Platforms judgment, the ICJ avoided precisely determining whether several (established and alleged) instances of uses of force in a maritime context, individually or together, constituted an armed attack.67 The jurisprudence of the ICJ, while leaving many controversial questions unanswered, demonstrates that the ascertainment of the specific factual circumstances of each alleged instance of an exercise of the right of self-defence, and the corresponding attribution of the burden of proof, are often more decisive for the determination of a situation, in particular the outcome of a judicial decision, than the resolution of certain questions of legal interpretation.
19 GA Res 3314 has proved to be of continued value as an (indirect) indication of whether certain forms of uses of force constitute ‘armed attacks’ in the sense of Art. 51 and this (p. 1409) seems to have been accepted by the vast majority of States.68 The lasting significance of GA Res 3314 has been reaffirmed in 2010 when it served as the basis for the definition of the crime of aggression for the purposes of the Statute of the International Criminal Court.69
20 As has been pointed out earlier,70 the notion of ‘armed attack’ has a narrower meaning than the term ‘threat or use of force’ in the sense of Art. 2 (4). Whereas an ‘armed attack’ always includes a use of force in the sense of Art. 2 (4), not all such uses of force constitute an ‘armed attack’. The latter only exists when force is used on a relatively large scale, is of a sufficient gravity, and has a substantial effect.71 Thus, mere frontier incidents, such as the incursion of an armed border patrol into another State’s territory, may well be characterized as a use of force contrary to Art. 2 (4), but not necessarily as an ‘armed attack’. This exclusion of frontier incidents from the concept of armed attack has been confirmed by the ICJ.72 A fortiori, a mere ‘threat’ of the use of force is not, in principle, an ‘armed attack’, but would have to be significantly qualified to be so regarded.73
21 The criterion of gravity need not necessarily be applied in a temporally isolated manner or be confined to one particular act. The accumulation of several smaller attacks, each of which does not as such reach the necessary gravity, can under certain circumstances trigger the right of self-defence of Art. 51.74 The ICJ can, however, be read as suggesting that a mere addition of several small-scale attacks alone does not easily translate into one sufficiently large-scale cumulative attack which would give rise to the right of self-defence.75 Given the manifest danger of circumvention of the gravity and immediacy76 elements of an ‘armed attack’, international practice and many commentators have been reluctant to recognize specific assertions of accumulations of events as constituting an (p. 1410) armed attack.77 It should, on the other hand, be noted that the openness for the concept as such seems to have increased in recent years.78
22 Even though the GA’s Definition of Aggression does not, as such, claim to define the notion of ‘armed attack’,79 its Art. 3 does in fact give useful indications on how to interpret this term.80 The provision lists examples of ‘acts of aggression’, which can, subject to certain qualifications, be taken to characterize ‘armed attacks’ within the meaning of Art. 51 as well. The ICJ has, for example, referred to the case of Art. 3 (g) of the Definition of Aggression as being one possible form of ‘armed attack’.81
23 The invasion or attack by the armed forces of a State on the territory of another State, as well as the cross-border use of weapons or bombardment of foreign territory, as mentioned in Art. 3 (b) of the Definition of Aggression, represent the classic cases of ‘armed attacks’,82 provided, however, that the military actions are on a certain scale and have a major effect, and are thus not to be considered mere frontier incidents.83 This would usually seem to be the case when an invasion occurs, but ‘attacks’, ‘bombardments’, and the ‘use of weapons’ do not in every case reach an intensity that enables them to be classified as ‘armed attacks’.84 Furthermore, Art. 3 (a) describes military occupation resulting from an invasion or attack and the annexation of foreign territory by the use of force as ‘acts of aggression’. The concept of ‘permanent aggression’85 to which these formulations refer is not applicable to the notion of ‘armed attack’. Occupation and annexation do not as such constitute ‘armed attacks’, since they do not necessarily involve the use of military force, even though they will typically go hand in hand.86
24 Pursuant to Art. 3 (c) of the Definition of Aggression, the blocking of a State’s ports or coasts by the armed forces of another State is deemed an ‘act of aggression’. At least if (p. 1411) maintained effectively, a blockade is also to be considered an ‘armed attack’,87 regardless of whether the obstruction is carried out by land, air, or naval forces. Since Art. 3 (c) is confined to the access to coasts and ports, only a part of a State’s transport system is protected. The land-locked countries did not succeed in including the barring of passage to the open sea across another State’s territory into the Definition of Aggression.88 This indicates that the majority of States were not willing to regard the obstruction of transit across land to the open sea as an ‘armed attack’.89 But exceptions have to be admitted where the blockade is equivalent to a military invasion, eg by cutting off all communication routes. In such extreme cases an ‘armed attack’ can be taken to exist.90
25 Art. 3 (d) of the Definition of Aggression stipulates that attacks by a State’s armed forces on the land, sea, or air forces or marine and air fleets of another State belong to the category of ‘acts of aggression’. In each of these cases, an ‘armed attack’ is involved as well, provided that the use of force is not insignificant.91 Thus it is undisputed, for instance, that warships and combat aircraft, when assaulted by foreign forces on the high seas or in international airspace respectively, have the right to defend themselves by means of military force.92 The ICJ has carefully formulated in the Oil Platforms Case that it ‘does not exclude the possibility that the mining of a single military vessel might be sufficient to bring into play the “inherent right of self-defence”’.93 An armed attack may also be found when military units of a State abroad are assailed by forces of the territorial State or a third State. Being instruments for the safeguarding of its political independence, (p. 1412) the regular forces of a State, wherever they are, enjoy the protection of the prohibition of the use of force.94
26 In principle, this also applies where a State attacks military units of another State which are present on disputed territory.95 If, however, the presence and actions of foreign troops are manifestly illegal, as in the case of a valid revocation of a stationing agreement and the lapse of an appropriate withdrawal period,96 the conduct of the foreign troops will in many cases constitute an ‘armed attack’, so that a forcible response by the territorial State would in turn be justified according to Art. 51.97 Insofar as the conduct of the foreign forces does not reach the scale and effects of an ‘armed attack’ proper, however, armed countermeasures would be contrary to the prohibition of force laid down in Art. 2 (4) and, given the required scale and intensity, would themselves be classified as an ‘armed attack’.
27 According to one view, coercive military measures against commercial vessels and aircraft outside the territory of their home State cannot be equated with attacks on that State itself, and are therefore not to be regarded as ‘armed attacks’.98 In the Oil Platforms Case, the ICJ carefully assessed the evidence regarding who fired a missile on one particular commercial vessel.99 This seems to give a certain support for another view which holds that the right of self-defence can also be triggered by an attack on a single commercial vessel.100 However, the safer interpretation is that the Court evaded addressing this issue and instead decided to deal with the specific allegation under review on evidentiary grounds. On the other hand, as distinct from individual commercial vessels or aircraft, assaults on the whole of the civilian marine or air fleet,101 as purportedly referred to in Art. 3 (d) of the Definition of Aggression,102 are said to threaten the affected State as such and thus, if applicable,103 to constitute ‘armed attacks’.104 The same should be true for continuous assaults on essential parts of them.
(p. 1413) 28 Diplomatic missions105 and individual citizens are not considered by most States to be ‘external positions’ of a State that could be the objects of an armed attack.106 The list in Art. 3 (d) of the Definition of Aggression does not include attacks on nationals abroad as an example of an act of aggression. In order to justify military rescue operations to help nationals who find themselves in difficulties in another country, there has been no shortage of attempts in the literature to declare the use of force by a State on its territory against foreign nationals to be an armed attack against the latter’s home State.107 It is argued that an assault on its nationals abroad constitutes an attack against a State, because nationals form part of a State’s population and are thus one of its essential constituents.108 This position is rejected by most commentators on the grounds that the security or existence of a State is not under threat if its nationals are assaulted in another country.109 Such an extensive interpretation of ‘armed attack’ would result in the blurring of any contours of the right of self-defence.110 Indeed, States which have actually carried out such rescue operations have refrained from accusing the territorial State of an ‘armed attack’.111 It is, however, not excluded that a specific justification for certain forms of evacuation operations may have evolved under customary international law.
29 Article 3 (e) of the Definition of Aggression classifies as an ‘act of aggression’ the use of one State’s armed forces which are within the territory of another State with the agreement of the receiving State, in contravention of the conditions provided for in the agreement, or any extension of their presence in the host State beyond the termination of the agreement. This form of aggression, a concession by the Great Powers to the anxieties of smaller States,112 can only be an ‘armed attack’ subject to the condition that minor violations of a stationing agreement may not be considered an armed attack. Only if the breach of the terms of the agreement has the effect of an actual invasion or occupation can an ‘armed attack’ triggering the right of self-defence pursuant to Art. 51 be held (p. 1414) to exist.113 In addition, the principle of bona fides would seem to require that the foreign troops must be given appropriate time to leave the country after the valid revocation of the consent by the territorial State.114
30 According to Art. 3 (e) of the Definition of Aggression, the action of a State in allowing its territory, which it has placed at the disposal of another State, to be used by that other State for perpetrating an act of aggression against a third State is itself deemed an act of aggression. This provision envisages only the voluntary ‘placing’ of territory at another State’s disposal and does not cover the case where a State has done what is required to prevent acts of aggression from being carried out from its territory by another State.115 Provided that the ‘act of aggression’ by the acting State qualifies as an ‘armed attack’, the complicity of the two States may be taken to comprise an ‘armed attack’, including by the State which remains passive from the military point of view.116
31 It is generally recognized that the participation by a State in the use of force by irregular organized armed groups is covered by the prohibition of the use of force. However, the specific pre-conditions required for ‘private’ uses of force to be a breach by a State of Art. 2 (4) are disputed.117 The same is true of the question as to whether and to what extent such indirect use of force may be classified as constituting an ‘armed attack’. In its Art. 3 (g), the Definition of Aggression characterizes certain forms of assistance to the ‘private’ use of force as ‘acts of aggression’. Although aware of the provision’s impact on the right of self-defence laid down in Art. 51, even those States which had previously denied the existence of a right of self-defence against indirect aggression, such as the Soviet Union118 and the non-aligned States,119 tolerated the inclusion of this example in the list of ‘acts of aggression’ and concentrated their efforts on keeping the scope of Art. 3 (g) as narrow as possible.120
In particular, it may be considered to be agreed that an armed attack must be understood as including not merely action by regular armed forces across an international border, but also the ‘sending by or on behalf of a State of armed bands, groups, irregulars or mercenaries, which carry out acts of armed force against another State of such gravity as to amount to’ (inter alia) an actual armed attack conducted by regular forces, ‘or its substantial involvement therein’.121
(p. 1415) This description, as it is contained in Art. 3 (2) of the Definition of Aggression, has been essentially confirmed by the ICJ in its Congo v Uganda judgment122 and may therefore be taken to reflect both Charter law and customary international law. The ICJ saw no reason to deny that, in customary law, the concept of the prohibition of armed attacks may apply to the sending by a State of armed bands to the territory of another State, if such an operation, because of its scale and effects, would have been classified as an armed attack rather than as a mere frontier incident had it been carried out by regular armed forces.123
33 The sending of armed bands by a State is an armed attack provided that they are committing acts of armed force against another State, the gravity of which equals that of the other acts listed in Art. 3 (g). In the case of ‘sending’, a sufficiently close link exists between the State and the private groups, so that the latter’s position is coming close to that of de facto State organs.124 If the aggressive acts carried out by such organized armed groups are of the required gravity,125 it seems perfectly justified to hold the sending State responsible for an armed attack. As far as the ICJ, in its interpretation of the term ‘armed attack’, has referred to the element of ‘substantial involvement’ in the sending of irregular armed groups which is contained in Art. 3 (g) of the Definition and thereby implies that this conduct also comes within the notion of an ‘armed attack’, it is necessary to bear in mind that application of the rather vague term ‘substantial involvement’, which allows for a variety of value-oriented assessments,126 requires a particularly careful consideration of the rules on the burden of proof, as the ICJ has demonstrated in its Congo v Uganda judgment.127
34 In its Nicaragua judgment, the ICJ did not consider assistance for rebels in the form of the provision of weapons or logistical support to suffice for an ‘armed attack’.128 Yet the ICJ did not suggest under what circumstances an involvement must be taken to be substantial enough as to amount to an armed attack. Taking current forms of state disintegration and international terrorism into consideration, the original formulation of the ICJ in the Nicaragua judgment today appears to be too narrow and in need of further differentiation. Otherwise, States could not sufficiently protect themselves against force committed by other States in an indirect manner.129 If, for example, a State knows that an irregular organized group is willing to commit significant acts of armed force against another State and it places its territory at the disposal of this group to train its members and to offer them a safe haven after they have committed these acts, and additionally provides them (p. 1416) with weapons and logistical support, it would be difficult to understand why this should be a lesser participation in the acts of the group than the mere sending of such a group. It is not appropriate to exclude certain types of support of irregular organized armed groups a limine from being qualified as ‘substantial involvement’ and consequently also as an ‘armed attack’. Instead it should be determinative to what extent State involvement has made it possible for irregular groups to commit acts of armed force of a certain scale which, if committed by a State, would have to be qualified as an ‘armed attack’.
35 The terrorist acts committed on 11 September 2001, against the United States of America, causing the death of thousands of victims and hitting at the centres of defence and trade, amounted to an armed attack as far as the scale and intensity of these acts of terrorism are concerned. These terrorist acts have caused an intense debate on whether the concept of ‘armed attack’ still requires any involvement of a State in the use of armed force by irregular armed groups as a precondition for an exercise of the right of self-defence.130
Such a restrictive reading of Article 51 might well have reflected the state, or rather the prevailing interpretation, of the international law on self-defence for a long time. However, in the light of more recent developments not only in State practice but also with regard to accompanying opinio juris, it ought urgently to be reconsidered, also by the Court. As is well known, these developments were triggered by the terrorist attacks of September 11, in the wake of which claims that Article 51 also covers defensive measures against terrorist groups have been received far more favourably by the international community than other extensive re-readings of the relevant Charter provisions….134
The attacks did not emanate from armed bands or irregulars sent by the DRC or on behalf of the DRC, within the sense of Article 3 (g) of General Assembly resolution 3314 (XXIX) on the (p. 1417) definition of aggression, adopted on 14 December 1974. The Court is of the view that, on the evidence before it, even if this series of deplorable attacks could be regarded as cumulative in character, they still remained non-attributable to the DRC.135
The Court nevertheless noted that it had ‘no need’ in this case ‘to respond to the contentions of the Parties as to whether and under what conditions contemporary international law provides for a right of self-defence against large-scale attacks by irregular forces’.136 It thereby left open the possibility that such large-scale attacks need not be attributable to a State in order to trigger the right of self-defence.
37 Although individual judges and many commentators have expressed a different view,137 the preferable view still seems to be that attacks by organized armed groups need to be attributed to a State in order to enable the affected State to exercise its right of self-defence, albeit under special primary rules of attribution.138 It is certainly true that recent developments have confirmed that organized armed groups are able to launch serious forms of transboundary force even without active support from a State. However, this consideration should only give rise to a reassessment of the pertinent rules on attribution in the light of the applicable primary rules.139 Recent practice suggests that any form of substantial assistance by a State to an organized armed group which conducts armed cross-border attacks leads to an attribution of such attacks to the assisting State for the purpose of the exercise of the right of self-defence. Such attribution flows from a proper interpretation of Art. 51 as a primary rule and not necessarily from the subsidiary (secondary) rules on State responsibility.140 It would go too far, however, to potentially expose any State, from the territory of which organized armed groups operate, to forcible measures of self-defence even if this State has either been unaware of the relevant activities of such groups or if it has done what can reasonably be expected of it under international law to remove the threat for other States which emanates from its territory. It is significant, for example, that the OAS and the Rio Group have in 2008 condemned the use of transboundary force by Colombia into Ecuador to pursue armed groups ‘who (p. 1418) were clandestinely encamped on the Ecuadorian side of the border’ and which had conducted raids into Colombia.141
38 Thus, acts of force, including terrorist acts, committed by irregular armed groups, as such, are not armed attacks in the sense of Art. 51 of the UN Charter. But if acts of force of a sufficient gravity, including terrorist acts, by irregular armed groups are attributable to a State, they are an armed attack in the sense of Art. 51. They are also attributable to a State if they have been committed by private persons and the State has encouraged these acts, has given its direct support to them, planned or prepared them at least partly within its territory, or was unwilling to take steps which can reasonably be expected of it to prevent these acts after having received substantiated information. The same is true if a State demonstrably gives shelter to terrorists after they have committed an act of terrorism within another State in a situation in which the attack can still be regarded as ongoing.
39 It is noteworthy in this regard that the SC, in its Res 1368 (12 September 2001)142 in which it unequivocally condemned the terrorist attacks which took place in New York, Washington DC, and Pennsylvania and regarded them as a threat to international peace and security,143 also recognized in the Preamble ‘the inherent right of individual or collective self-defence in accordance with the Charter’. On 8 October 2001, after the beginning of their military action against Afghanistan, the United States and the United Kingdom reported to the SC that this action was taken in accordance with their inherent right of individual and collective self-defence. In a press statement of the same day, the President of the SC declared that ‘the members of the Council were appreciative of the presentation made by the United States and the United Kingdom’.144
40 As far as measures of self-defence against acts of organized armed groups, including terrorists, are carried out on the territory of another State, the principles of proportionality and necessity145 must be strictly respected.
41 A special situation arises if a State is not unwilling but simply unable to impede acts of organized armed groups committed by making use of its territory. The ICJ touched upon this issue in its Congo v Uganda judgment in which it did not accept that the right of self-defence could be exercised in such a situation against smaller-scale attacks by irregular forces. The Court did, however, leave open the possibility that ‘contemporary international law provides for a right of self-defence against large-scale attacks of irregular forces’.146 It may indeed be appropriate that a victim State should be able to react by military means against large-scale attacks by armed groups which operate from the territory of a State which is manifestly incapable of taking serious steps to prevent such (p. 1419) attacks from originating in its territory.147 In such cases the special rule on attribution would be exceptionally wide (‘manifestly unable to prevent large-scale attacks’) and the respect of the principles of necessity and proportionality would have to be particularly insisted upon.148
42 Since electronic means of communication are today also used for the purpose of causing damage, including as means of warfare, the question is increasingly discussed whether so-called ‘cyber attacks’ can trigger the right of self-defence.149 Certain incidents, such as the causation of the self-destruction of Iranian uranium centrifuges by the ‘Stuxnet’ worm in 2010,150 or the overwhelming of government servers in Estonia in 2007,151 have fuelled the debate.
43 It seems that, despite their novelty and their specific character, cyber attacks can be satisfactorily assessed by properly taking into account the general considerations which apply to other forms of aggression. Since the ICJ has held that the exercise of the right of self-defence does not depend on the type of weapon which is used for an attack,152 an ‘armed attack’does not require the use of kinetic weapons, but may, in principle, also be conducted by electronic weapons.153 It is another question whether the use of an electronic weapon must produce physical effects which are comparable to those which kinetic, atomic, chemical, or biological weapons typically produce.154 In order to reach the level of an ‘armed attack’, every kind of weapon must produce substantial and immediate destructive effects.155 In order to reach the threshold of an ‘armed attack’, such effects must be comparable to those which are required for a conventional use of force to be recognized.156 Whether the effect of a ‘cyber attack’ is actually comparable to such a conventional use of force must take into account the specific character and importance of the protected object for the affected State and its population. Thus, the immediate disabling of vital infrastructure with inhibitive (and not sufficiently quickly reparable) effects on the ability of the State to act or on the elementary living conditions of the (p. 1420) population can, in principle, produce the necessary destructive effect which would justify the designation ‘armed attack’.157
44 The possibility that ‘cyber attacks’ can, in certain extreme cases, amount to ‘armed attacks’ which may trigger the right of self-defence, does not mean, however, that the possibility of exercising this right can be easily established. This is true even if it were clear that a ‘cyber attack’ with the required immediate destructive effect has actually taken place. So far, the main difficulty seems to lie in the fact that it is normally not possible to quickly identify the attacker and thus to attribute the attack to a particular actor within the necessary time frame.158 Even if the requirement of attribution of the attack to a State were relaxed considerably in view of the fact that it is even more difficult to distinguish between State and private activity in the virtual sphere, the difficulty would mostly remain in identifying the source of the electronic attack.
45 Furthermore, even if the source of an electronic attack could be located, additional conditions must be fulfilled before the right of self-defence can actually be exercised. In particular, the attack must not be over but still be ongoing. This will often be difficult to determine, and the burden of establishing that the attack is still ongoing lies on the State that purports to exercise its right of self-defence.159
46 Finally, even if all other conditions are fulfilled, the principle of proportionality may in certain circumstances require that those States which have the capacity to do so, exercise the right of self-defence solely by electronic means, and not by way of kinetic or other weapons which produce direct physical effects. The answer to the question of whether this is actually required will usually be quite fact-specific. Again, the burden to demonstrate the proportionality of the response lies on the State which purports to exercise its right of self-defence.160
47 Art. 51 of the Charter allows not only individual, but also collective self-defence. It is generally accepted that the right of collective self-defence also authorizes a non-attacked State to lend its assistance to an attacked State.161 Thus, the right to collective self-defence is not, as the terms of Art. 51 might suggest, restricted to a common, coordinated exercise of the right to individual self-defence by a number of States which have all been attacked.162 Such a restrictive interpretation of the right of collective self-defence corresponds neither to the history of Art. 51163 nor to State practice since 1945.164 (p. 1421) Furthermore, this interpretation would diminish the effectiveness of the prohibition of the use of force, since the occasional lack of reliability of the UN system of collective sanctions may leave weaker States unprotected and at the mercy of militarily superior States.
48 The ICJ has interpreted the right of collective self-defence accordingly.165 In its Nicaragua judgment, the ICJ emphasized that any collective exercise of the right of self-defence requires the attacked State for whose benefit collective self-defence would be used ‘to form and declare the view that is has been so attacked’.166 It is not required for the exercise of the right of collective self-defence that the State invoking the right be under an obligation resulting from a treaty of assistance.167 Rather, it is sufficient, but also necessary, that the support be given with the consent of the attacked State.168 Such consent needs to be declared, as the ICJ has s