4 Self-defence: the framework
The law on self-defence is the subject of the most fundamental disagreement between states and between writers. Some of the divisions over the scope of the right of self-defence, especially as to whether anticipatory self-defence against an imminent armed attack and the protection of nationals abroad are lawful, date back to the creation of the UN.1 These issues acquired new prominence in the light of the later development of a ‘Bush doctrine’ of pre-emptive self-defence.2 Other divisions centre on the right to use force in self-defence in response to colonial occupation, to terrorist attacks, and to other attacks by non-state actors. The attacks of 9/11 and their aftermath have brought a fundamental reappraisal of the law on the use of force against terrorism. Some states have been attempting to widen the scope of self-defence in radical ways. Chapter 5 will examine the impact of the ‘war against terror’ on international law; this chapter focuses on the general framework of the law of self-defence.
Differences over the scope of self-defence prevented any substantive provision on this being included in General Assembly resolutions designed to codify the law on the use of force. States negotiating the 1970 Declaration on Friendly Relations and the 1974 Definition of Aggression did not include any provision on self-defence; in the 1987 Declaration on the Non-Use of Force they could not go beyond the statement that ‘States have the inherent right of individual or (p. 121) collective self-defence if an armed attack occurs, as set forth in the Charter of the United Nations’.3
However, in practice these fundamental doctrinal differences were not (until recently) of decisive significance as to the legality of the use of force except in a few isolated, though much discussed, instances. States using force against another state almost invariably invoke self-defence; in the vast majority of such claims this has not given rise to any doctrinal issues or to any divisions between states as to the applicable law. Whether the use of force is a one-off minor incident (either involving an attack on a state’s territory or on its land, sea, or air forces outside its territorial limits4) or an ongoing conflict, typically one or both states involved asserts that it has been the victim of an armed attack and claims the right to self-defence; the controversy centres on the questions of fact as to whether there has actually been an armed attack of the type claimed and, if so, which state was the victim. In theory it should always be possible to determine whether there was an armed attack and who is acting in self-defence. But in practice the situation is more complex.5 The difficulties in establishing the facts in cases involving self-defence and the legality of use of force were very obvious in the ICJ cases, Cameroon v Nigeria (2002),6 Iranian References(p. 122) Oil Platforms (2003),7 and Armed Activities on the Territory of the Congo (DRC v Uganda) (2005).8 The pleadings of the parties in the Georgia v Russia case, Application of the International Convention on the Elimination of All Forms of Racial Discrimination, reveal an almost total lack of agreement on the fundamental questions of fact.9 The issue is left unresolved in the vast majority of conflicts; certainly the Security Council does not generally make express pronouncements determining this crucial legal issue. The parties may register their positions with the Security Council, but often there may be no debate and no resolution or statement. Even if there is a resolution or statement, it is far more common for this to take the form of a call for a ceasefire rather than any attribution of responsibility. This can be seen in the 1998–2000 conflict between Ethiopia and Eritrea; the Security Council did not condemn one or the other of the two states involved in the conflict, but repeatedly called for an end to the hostilities and peaceful settlement of the territorial dispute which was at the root of what the Secretary-General called an ‘incomprehensible war’.10 In contrast, in 2008 the Security Council did condemn Eritrea’s military action against Djibouti over an insignificant piece of territory, but it did not use the language of Article 51, and it simply called on the parties to commit to a ceasefire and to resolve their border dispute peacefully.11
References(p. 123) The 1980–8 Iran/Iraq conflict was unusual in that the Security Council asked the UN Secretary-General to investigate responsibility for the conflict and the latter did make an express finding on the facts of the case after the conflict had ended. Iran persistently claimed that Iraq bore responsibility for initiation of the conflict and eventually secured an inquiry into the origin of the conflict by the UN Secretary-General and vindication of its position. The Secretary-General reported that the conflict was begun in contravention of international law through the illegal use of force and disregard for a state’s territorial integrity; Iraq was responsible for the conflict because of its armed attack against Iran on 22 September 1980.12 This willingness to identify the outbreak of a conflict and to determine responsibility was more common in the early days of the UN.13 It is rare for the Security Council today to enter into this question; members clearly see its role as the promotion of the restoration of peace rather than as the assignment of responsibility. The 1990 Iraq/Kuwait conflict was another exceptional case, seen by many as marking a new role for the Security Council and the start of a new legal order; in this case the Security Council did explicitly uphold the right of Kuwait to self-defence.14 But more typical have been the many, relatively minor, limited conflicts where the Security Council did not involve itself in any pronouncements on self-defence. There is a striking contrast between the hundreds of communications to the Security Council in which states claim to be the victims of armed attacks and the few conflicts discussed by the Council. The vast mass of use of force passes unmarked by any debate or resolution, let alone by any formal finding as to who was the victim. And in the vast mass of cases—both before and after 9/11—there is no controversy as to the applicable law.
Thus the natural focus of writers on controversial cases where states invoke self-defence in protection of nationals, anticipatory or pre-emptive self-defence, and response to terrorism inevitably gives an unbalanced picture and distorts our perception of state practice; it helps to give the impression that the far-reaching claims of states like the USA and Israel are normal rather than exceptional.
The disagreement as to the scope of self-defence generally turns on the interpretation of Article 51. This provides:
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 measures necessary to maintain international peace and security. Measures taken by Members in the exercise of the 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 to maintain or restore international peace and security.
There is no need here to do more than set out the basic arguments of the two main groups of writers whose opposing positions have become well entrenched in the last sixty years. Those who support a wide right of self-defence going beyond the right to respond to an armed attack on a state’s territory argue, first, that Article 51 of the UN Charter, through its reference to ‘inherent’ right of self-defence, preserves the earlier customary international law right to self-defence. The Charter does not take away pre-existing rights of states without express provision. Second, they argue that at the time of the conclusion of the Charter there was a wide customary international law right of self-defence, allowing the protection of nationals and anticipatory self-defence.15 The opposing side argues that the meaning of Article 51 is clear; the right of self-defence arises only if an armed attack (French: agression armée) occurs. This right is an exception to the prohibition of the use of force in Article 2(4) and therefore should be narrowly construed. Article 51 cannot be read to preserve customary international law that is inconsistent with its terms.16 The limits imposed on self-defence in Article 51 would be meaningless if a wider customary law right to self-defence survives unfettered by these restrictions. Moreover, they claim that by the time of the Charter customary law allowed only a narrow right of self-defence.17 These early arguments turned, first, on treaty interpretation and, second, on an assessment of the state of customary international law in 1945. Policy considerations as to the realism of taking a wide or narrow view also played a crucial role.
References(p. 125) Those who supported the wide right of self-defence discounted the rejection of their position by the large majority of states in practice since 1945; for these writers it seems that the Charter preserved customary law as it allegedly was in 1945. Thus for them the term ‘inherent right of self-defence’ in Article 51 was apparently not susceptible of restriction in the light of subsequent state practice. It did, however, apparently allow the expansion of the right of self-defence.18 An alternative approach invoked the breakdown of the UN collective security system during the Cold War in order to justify a wide right to self-defence in the same way that some argue for a narrow interpretation of the prohibition of the use of force in Article 2(4).19 Again this argument was at variance with the mass of state practice and had to discount the views of the vast majority of states.
In practice, states making their claims to self-defence usually try to put forward arguments that will avoid doctrinal controversy and appeal to the widest possible range of states. Especially since the Nicaragua case, states have taken care to invoke Article 51 to justify their use of force. They do so even when this seems entirely implausible and to involve the stretching of Article 51 beyond all measure. Even when relying on a wide right of self-defence, in the absence of an armed attack on their territory or on their armed forces outside their territory, states invoke Article 51. Either this is just ritual incantation of a magic formula, not expected to be taken seriously, or their case is implicitly that Article 51 allows a wider customary right, including anticipatory self-defence or forcible response to terrorism.
States, in making their own justification or in responding to the claims of others, on the whole and not surprisingly do not enter into extended doctrinal debate in their communications to the Security Council. And even in Security Council debates or in negotiation of General Assembly ‘law-making’ resolutions on the use of force, they tend simply to assert a wide or narrow view of self-defence without going into the theoretical justifications for their view. Generally, more time is devoted to expounding their own version of the facts and their political justifications. It is only in the most controversial cases where there is a doctrinal division that states do enter into protracted legal justification. Israel’s arguments in defence of its 1976 rescue operation at Entebbe and of its attack on the Iraqi nuclear reactor in 1981 are unusual in that they are protracted.20 There was a References(p. 126) similarly detailed discussion of the US 1983 intervention in Grenada and its 1986 bombing of Tripoli.21
The Role of the Security Council
Article 51 assigns a central role to the Security Council: states are under a duty to report measures taken in the exercise of the right of self-defence to the Security Council and the right to self-defence is temporary until the Security Council ‘takes measures necessary to maintain international peace and security’. The USA in the Nicaragua case argued that the ICJ should not pronounce on claims of self-defence because Article 51 provides a role in such matters only for the Security Council.22
Although Article 51 envisages a crucial role for the Security Council, it does not necessarily require the Council to pronounce on the legality of any claim to self-defence. In practice the Security Council has generally not made such express pronouncements. Some French writers have therefore claimed that it has not done enough to give self-defence a clear content or indeed any real meaning: the right of self-defence is ‘indeterminate’ or even obsolescent.23 Thus Combacau took a rather formalistic approach. He argued that the Security Council can only contribute to the crystallization of the law in this area when a state expressly makes a claim to be acting in self-defence to the Council and the Council makes an express response; on the basis of the Repertoire of the Practice of the Security Council up to 1974 Combacau claimed that states rarely made such claims. First, this approach seems too rigid. Security Council resolutions and statements may be of significance in the development of the law if in substance they deal with state behaviour and implicitly or expressly accept or reject claims of self-defence. Second, his argument was based on the Repertoire of the Practice of the Security Council up to 1974. This does not give a complete picture and, moreover, is based References(p. 127) on practice before the decision in the Nicaragua case which led to a clear change in state behaviour.
It is true that only a very few Security Council resolutions have made express reference to Article 51. Typically these assert in general terms the right of a particular state to take action in self-defence. Such resolutions have generally not been passed in recent years. They were passed in response to South Africa’s attacks on the front-line states during the apartheid era, and in response to the use of force by Portugal and Israel. For example, Angola’s right to take measures in accordance with Article 51 when it had been subject to attacks by South Africa was affirmed by the Security Council; these resolutions also condemned South Africa’s use of force.24 Subsequently, and exceptionally, Kuwait’s right to self-defence was affirmed by the Security Council after the 1990 Iraqi invasion.25 And Resolution 1234 (1999) on the conflict in the DRC affirmed in general terms the right of individual or collective self-defence in accordance with Article 51.26 Resolutions 1368 (2001) and 1373 (2001) passed after the 9/11 terrorist attacks are thus unusual in their express invocation of the inherent right of self-defence in their preambles.27
Other resolutions respond to the use of force by states; Insofar as they condemn particular actions they may be taken as rejections of a state’s claim to self-defence even if this is not express in the resolution. Thus the attempt to deny any clear content to the right of self-defence because of the nature of the decision-making of the Security Council underestimates the significance of the vast mass of state practice, and especially of the many state communications to the Security Council. The core content of self-defence is universally accepted.28
However, the approach of Combacau and Delivanis, although formalistic, has some justification. The Security Council resolutions and statements, although they may be authoritative as to the legality of particular uses of force, cannot do much to resolve the doctrinal controversies as to the scope of the right of self-defence. Any condemnation of controversial use of force such as protection of nationals, anticipatory self-defence, and action against irregulars and terrorists may be limited to the particular facts. Rather than condemn protection of nationals or anticipatory self-defence in general, the Security Council condemns the particular use of force. The Security Council debates will usually reveal the doctrinal divisions between states; it is clear that in order to secure agreement on a resolution the Security Council may have to avoid any pronouncement on the (p. 128) underlying doctrine. Therefore, the resolutions may provide only indirect evidence as to the state of the law. They do not contain general statements of the law. Pronouncements on individual breaches may do no more than make it possible to argue, for example, that the fact that almost all uses of anticipatory self-defence before 9/11 were condemned suggests the weakness of such a doctrine.
The duty to report to the Security Council
Since the judgment in the Nicaragua case it is noticeable that states on the whole do comply with the Article 51 requirement that ‘measures taken by Members in the exercise of the right of self-defence shall be immediately reported to the Security Council’; it is clear that states have taken seriously the Court’s message that failure to do this will weaken any claim to be acting in self-defence. The Court held that ‘the absence of a report may be one of the factors indicating whether the State in question was itself convinced that it was acting in self-defence’.29 Judge Schwebel, in his Dissenting Opinion, strongly criticized this as unacceptable in the case of covert self-defence.30 But any attempt to attack this finding by the Court as an objectionable innovation is fundamentally misconceived.
The argument that failure to report was evidence against a claim to self-defence had been made many times even before the case. For example, the UK during the Vietnam conflict said that the fact that the USA had reported to the Security Council in 1964 its actions in response to alleged attacks by North Vietnamese naval vessels in the Gulf of Tonkin was an indication that it was actually acting in self-defence.31 And after the USSR intervention in Afghanistan in 1979 the UK asked in the General Assembly debate why, if there had really been attacks on Afghanistan, it had not raised the matter before the Security Council.32 Failure to report was also used as a sign of bad faith by the USA itself. After the clashes between the USA and Libya in the Gulf of Sirte in March and April 1986 (that is, during the Nicaragua case proceedings) the USA used the argument that Libya had not reported its actions to the Security Council as evidence that it was not acting in self-defence. Conversely, the UK said that the US report of these episodes to the Security Council under Article 51 was a sign of good faith.33 However, it is clear that the reporting requirement is merely procedural; failure to comply does not of itself invalidate a claim to self-defence.34
(p. 129) After its decision in the Nicaragua case the Court in Armed Activities on the Territory of the Congo (DRC v Uganda) simply ‘noted’ Uganda’s failure to report the use of force it claimed as self-defence.35 The Court did not discuss this further, but the clear implication was that this was another factor indicating that Uganda had not been acting lawfully. And in the Eritrea/Ethiopia Claims Commission award on Ethiopia’s Ius ad Bellum Claims 1–8 the tribunal in making its controversial finding that Eritrea was not acting in self-defence (even though it was using force against Ethiopian troops on Eritrean territory) took account of the fact that Eritrea, unlike Ethiopia, had not reported its actions to the Security Council under Article 51.36 It is therefore surprising that Ethiopia did not report its 2006 intervention in Somalia to the Security Council under Article 51, even though it claimed that it was acting in self-defence.37
Before the Nicaragua case the reporting requirement was not always strictly observed in cases of individual self-defence (in marked contrast to the practice with regard to collective self-defence). But, even before the Nicaragua decision, reporting by states was more common than the Repertoire of the Practice of the Security Council indicates; a study of the communications of states to the Security Council gives a fuller picture of state practice in this regard.38 After Nicaragua it can no longer be maintained that the reporting requirement is rarely observed.39
Indeed, there was for a while a tendency to over-report claims to individual self-defence, if anything. It seems clear that a state involved in a one-off episode should report if relying on self-defence. Also states parties to a prolonged conflict should, References(p. 130) if relying on self-defence, go to the Security Council at the start of that conflict.40 However, when there is a prolonged conflict the states parties have sometimes chosen not simply to make their claims to self-defence at the start of the conflict, but often to report each episode separately. That is, they apparently interpreted the reporting requirement in Article 51 that ‘Measures taken by Members in the exercise of this right of self-defence shall be immediately reported to the Security Council’ as requiring continuing reports. This significantly increased the burden on the state claiming self-defence in that it had to show that each episode in isolation constitutes necessary and proportionate self-defence, rather than simply the campaign taken as a whole. This repeated reporting was marked in the practice of Iran and Iraq during their 1980–8 conflict41 and in the practice of the UK and Argentina in the Falklands conflict.42 Similarly in the 1998–2000 conflict between Eritrea and Ethiopia both parties repeatedly invoked self-defence.43
It was also the practice of the USA with regard to its involvement in the 1980–8 Iran/Iraq conflict, when the US navy was providing convoys for US-flagged ships through the Gulf to protect them against attack by the belligerent parties. Instead of making a blanket statement at the start of its involvement, the USA sought to justify each episode of the use of force against Iran.44 Here we see self-defence being invoked with reference to the protection of US ships and aircraft. This choice to report individual episodes led the USA into some difficulties when it had to justify its actions against Iranian oil platforms and its shooting down of the Iran Airbus in 1988 as self-defence.45
Such repeated reporting may seem to play partly a propaganda role.46 Given that the Security Council does not usually pronounce on the legality of a claim to self-defence at the start of a conflict, it may be understandable that the states refer each individual episode to the Security Council in an attempt to portray themselves as victims, as in the Iran/Iraq conflict when Iran and Iraq repeatedly reported particular incidents to the Security Council. Because the Security Council made References(p. 131) no initial determination that Iraq was the aggressor, it could seek to portray itself as the victim, especially when Iran later refused to accept the 1987 mandatory ceasefire resolution. Also controversially, such reporting of individual episodes as self-defence may represent an attempt to rely on Article 51 rather than the laws of war where an action’s legality is doubtful as a matter of international humanitarian law. Thus in the Vietnam War the USA justified its use of force generally as collective self-defence of South Vietnam. It also subsequently reported individual episodes such as its mining of the ports of North Vietnam and its bombing of neutral Cambodia as constituting self-defence.47 Again in the Falklands conflict the UK reporting of individual episodes as self-defence may reflect its doubts as to the adequacy of the laws of war at sea.48
Self-defence as a temporary right
The Security Council also has a role in the control of the right of self-defence through the stipulation in Article 51 that the right of self-defence continues ‘until the Security Council has taken measures necessary to maintain international peace and security’. Given that the UN Charter aims not only to limit, but also to centralize, the use of force under UN control, it seems clear that the intention was to give the Security Council itself the right to decide whether such measures terminating the right to self-defence had been taken. But, in the absence of express determination of the existence or continuation of the right to self-defence, this provision has in the past given rise to some controversy.49 The Falklands (Malvinas) conflict is a famous example; after the Argentine invasion of the UK colonial territory in 1982 the Security Council, in Resolution 502 (10-1-4), determined that there had been a breach of the peace, demanded an immediate cessation of hostilities, demanded an immediate withdrawal of all Argentine forces, and called on the governments of Argentina and the UK to seek a diplomatic solution to their difficulties. Did this amount to ‘necessary measures to maintain international peace and security’ which terminated any UK right to use force in defence of the Falklands? The UK argued that it did not, because Argentina, the aggressor, remained in occupation of the islands.50 The question came up (p. 132) again in the 1980–8 Iran/Iraq conflict. After the mandatory Security Council Resolution 598 (1987) calling for a ceasefire, was Iran subsequently exceeding its right to self-defence in its refusal to accept the ceasefire, given that it had already by mid-1982 recovered the territory earlier occupied by Iraq? Although the USA and the UK did not expressly make this argument in the Security Council, they came close to it.51
The UK apparently learned its lesson from the controversy over the Falklands. When the Security Council responded to Iraq’s 1990 invasion of Kuwait it imposed sanctions on Iraq; in the same resolution it included an affirmation of ‘the inherent right of individual or collective self-defence, in response to the armed attack by Iraq against Kuwait, in accordance with Article 51 of the Charter’. Thus no problem could arise as to whether the imposition of economic sanctions by the Security Council had terminated any right of states to use collective self-defence to help Kuwait. The USA and the UK could act in collective self-defence of Kuwait even before specific authorization for the interception of ships and aircraft bound for Iraq and Kuwait was given by the Security Council.52
Security Council measures and self-defence
The question has also arisen of the relationship between the state’s right to self-defence and the powers of the Security Council: are the powers of the Security Council under Chapter VII of the Charter limited by the requirement that such measures do not undermine the right of self-defence under Article 51? This question came up first in 1977 when France argued that an arms embargo on South Africa might violate its right to self-defence. However, France said, the intention in the aftermath of the recent crackdown by the South African government was to protest against the stockpiling of weapons intended for purposes of internal repression; therefore it had decided to vote in favour of a mandatory arms embargo on South Africa.53 This issue arose again in the debate over the compatibility of the arms embargo on the whole of the former Yugoslavia with the right of self-defence under Article 51.54 At the References(p. 133) outbreak of the conflict in Yugoslavia in 1991 the Security Council imposed an arms embargo on the whole of Yugoslavia. Resolution 713 (1991) was passed unanimously and the arms embargo was imposed with the consent of the federal government of Yugoslavia. When Yugoslavia split up and Bosnia-Herzegovina became a member state of the UN in May 1992 it argued that the arms embargo should not be applied to it. It sought the lifting of the embargo by the Security Council from September 1992. It claimed that its inherent right to self-defence under Article 51 took priority over the embargo, and that in order to exercise this right against Yugoslavia (Serbia and Montenegro) the embargo must be lifted. In the Security Council debates those in favour of lifting the embargo argued either that Resolution 713 (1991) had been superseded when Bosnia became a member of the UN or that the resolution should be interpreted as not applying to Bosnia or, more radically, that if the resolution did impose an embargo on it, then the resolution was invalid as outside the powers of the Security Council because it violated Bosnia’s inherent right to self-defence. The Security Council refused to accept this argument and did not lift the embargo even though the General Assembly repeatedly urged it to consider this.55
It is clear that there are strong arguments against a claim that an arms embargo violates Article 51 of the UN Charter. If every arms embargo is automatically inconsistent with Article 51 this would restrict the Security Council’s discretion to take measures under Article 41 and deprive it of a useful tool to put pressure on a wrongdoing state or to try to limit the escalation of a conflict. All states subject to an arms embargo could claim that their rights under the Charter prevailed over the arms embargo. It seems unlikely that Bosnia-Herzegovina, in putting its claim for the lifting of the embargo, was really making the argument that every arms embargo violated Article 51.
Even if Bosnia-Herzegovina was putting forward a less fundamental argument and was claiming merely that in the particular circumstances the arms embargo in Resolution 713 (1991) violated its right to self-defence, this seems a dangerous precedent and one that would undermine the freedom of the Security Council to maintain an arms embargo. States suffering civil wars and subject to arms embargos could make plausible cases that they were under outside threat and needed to exercise their rights to self-defence. The better position is that an arms embargo may affect the right to self-defence but does not actually deny that right.
References(p. 134) This question came up again with regard to Rwanda.56 The Security Council imposed an arms embargo in 1994, against the wishes of the government then in power, to try to prevent the escalation of violence. Following Bosnia’s claims, Rwanda pursued a similar line of argument, that the arms embargo imposed on it after large-scale massacres in 1994 should be lifted because there was a threat to it from outside. This time the Security Council did respond, noting with concern the reports of military preparations and incursions into Rwanda by supporters of the former government. It recalled that the original prohibition on the delivery of arms was aimed at preventing their use in the massacre of innocent citizens. The embargo was lifted as far as arms destined for the government were concerned, but otherwise remained in place.57 This precedent may have made it more difficult for the Security Council to keep in place against the wishes of the government of the state concerned an arms embargo imposed during a civil war. The modification of the total arms embargo on Sierra Leone to allow arms to be supplied to the government and those supporting it reinforces this view.58 But the argument for the lifting of an arms embargo is less attractive where the embargo has been imposed as a sanction, as in the case of Liberia and, arguably, in the case of the Ethiopia/Eritrea conflict. In the former Liberia was subjected to an arms embargo for its intervention in the conflict in Sierra Leone. In the latter the Security Council simply urged member states not to supply arms to either side: the arms embargo was designed to help to bring an end to a ‘senseless war’.59 The claims by Liberia and Ethiopia that the embargos were unlawful because they denied the states concerned the right to defend themselves were not successful.60
The Scope of Self-Defence
All states agree that if there is an armed attack the right to self-defence arises, but there are controversies as to what constitutes an armed attack.61 The paradigm case is obviously an invasion by the regular armed forces of one state into the (p. 135) territory of another state; it is also accepted that an attack on the land, sea, and air forces of a state outside its territory constitutes an armed attack.62 However, questions concerning the definition of the concept and the identification of the start of an armed attack may arise out of the special characteristics of particular weapons.63 Thus the concept of armed attack by modern missiles64 and naval mines65 has given rise to special questions. Questions as to the regulation of cyber-attacks have also been much debated in recent years, but states have not yet evolved detailed rules on this area.66 Writers are divided as to whether physical References(p. 136) harm to people or property is necessary for a cyber-attack to constitute and armed attack.67 There are also obvious problems with attribution, the identification of those responsible for an attack. To date no cyber-attack has been classified as an armed attack by states.68 There are also further disagreements as to the degree of gravity necessary for an armed attack, as to whether it is possible for a cumulative series of minor attacks to constitute an armed attack, and as to whether any specific intent on the part of the attacking state must be shown. Other questions centre on cross-border activity by irregular forces: what degree of state involvement, if any, is necessary for the existence of an armed attack?
The ICJ has considered the concept of armed attack in a series of cases, starting with the Nicaragua case.69 In that case, Oil Platforms,70 Armed Activities on the Territory of the Congo (DRC v Uganda)71 and in the brief and obscure passage in its Advisory Opinion on The Legal Consequences of the Construction of a Wall on the Occupied Palestinian Territory72 it generally took a cautious approach to the right of self-defence and was careful to avoid pronouncing on the most contentious issues where this was not necessary for its decision.73
The concept of armed attack was central to the ICJ’s judgment on collective self-defence in the Nicaragua case; the USA claimed that its use of force against Nicaragua was justified as collective self-defence of Costa Rica, Honduras, and El Salvador in response to armed attacks on those states by Nicaragua. It contended that Nicaragua had intervened in El Salvador and other neighbouring states in order to foment and sustain armed attacks upon the governments of those states, and that its subversive intervention in the governing circumstances was tantamount to an armed attack. But the Court rejected this as it found that there was no armed attack by Nicaragua. The Court’s view of armed attack has been severely attacked, especially by US writers.74 However, the Court’s description of References(p. 137) the scope of armed attack was consistent with state practice and with the practice of the Security Council.
The Court first considered whether an armed attack had to be by a regular army. It used the Definition of Aggression to support its view that ‘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’ could be an armed attack.75 This limited reliance on the Definition of Aggression (stopping short of a complete identification of the two concepts) to elucidate the meaning of armed attack seems justified in the light of state practice.76 States do not today challenge the view that actions by irregulars can constitute armed attack; the controversy centres on the degree of state involvement that is necessary to make the actions attributable to the state and to justify action in self-defence in particular cases. This question has attracted a large amount of academic discussion since the terrorist attacks of 9/11.77
The Court then held that assistance to rebels in the form of the provision of weapons or logistical or other support did not amount to an armed attack, although it could be illegal intervention.78 The Court’s position on the meaning of armed attack was strongly criticized by Judges Schwebel (USA) and Jennings (UK) in their Dissenting Opinions. Judge Schwebel said that the reference in the Definition of Aggression to ‘substantial involvement’ in the sending of armed bands meant that an armed attack could include financial and logistical support for armed bands. However, the drafting history of the resolution does not support this construction and it is not consistent with Schwebel’s own earlier recognition of a distinction between the wider conception of aggression and the narrower conception of armed attack.79 Schwebel argued that the Court’s narrow definition of armed attack and consequent limit of the right of self-defence offered a prescription for overthrow of weaker governments by predatory governments (p. 138) while denying potential victims what in some cases may be their only hope of survival.80
Judge Jennings similarly argued that the Court’s approach was not realistic, given that power struggles are in every continent carried on by destabilization, interference in civil strife, comfort, aid, and encouragement to rebels, and the like. Because Chapter VII of the UN Charter was not working it was dangerous to define unnecessarily strictly the conditions for lawful self-defence.81 The converse argument could equally well be made; because Chapter VII was not working it was important not to allow the abuse of the right of self-defence. Jennings said that ‘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 an important element in what might be thought to amount to an armed attack where it is coupled with other kinds of involvement.’82
The focus for both dissenting judges was on the question of fact: did the particular actions of Nicaragua taken as a whole amount to an armed attack? They were also making policy arguments as to what the law ought to be. Neither Schwebel nor Jennings adduced any evidence that in state practice mere provision of weapons and logistical support in isolation had been treated as armed attack (as opposed to unlawful intervention) in cases of self-defence.
A few commentators accepted the arguments of the dissenting judges on the facts, but also went further and made strong criticisms of the Court’s conception of armed attack. That is, they did not just reject the Court’s interpretation of the facts, they also said that it was mistaken on the law.83 They did not go so far as to say that a mere supply of arms could alone amount to an armed attack, but they argued that arms supply combined with financial and logistical support could in principle be an armed attack.84 However, their criticisms were based on policy considerations; they did not give any examples of state practice or Security Council practice to support their arguments. Nor do they apply such principles to US interventions. In contrast, the Court’s judgment is consistent with state practice. The Security Council, in its many calls for an end to the supply of arms or other outside support to opposition forces in situations such as those in References(p. 139) Afghanistan, Yugoslavia, and Rwanda, has never identified such interventions as an armed attack. Nor do states make such claims.
In Armed Activities on the Territory of the Congo (DRC v Uganda) the DRC brought an action against Uganda for unlawful use of force. Uganda sought to justify its use of force partly on the basis of self-defence. In order to do so it took a wide view of armed attack to support its claim that the DRC was responsible for attacks by the irregular forces of the Allied Democratic Forces (ADF) operating from the DRC against Uganda. Uganda ‘recalled the existence of a powerfully expressed alternative view according to which the formulation of the majority of the Court in the Nicaragua case was excessively narrow in its approach to the interpretation of the phrase “armed attack” ’. Uganda said that the alternative view could be expressed: the giving of logistical support to armed bands with knowledge of their objectives may constitute an armed attack.85 But the support Uganda offered for this argument in its Pleadings was extremely weak and the Court did not change its view.86
Cross-border action by irregular forces
The issue of cross-border action by irregular forces has given rise to much difficulty. If these forces are acting on behalf of the state from whose territory they are operating and their actions are of such gravity as to amount to an armed attack, the situation is clear.87 However, the question of what degree of state involvement is necessary to allow the use of force against the territory of the host state in self-defence has proved an intractable issue. In the Nicaragua case the Court treated the Definition of Aggression with its provision ‘sending by or on behalf of a state … or its substantial involvement therein’ as definitive as to what amounted to an armed attack. It did not expressly go into the issue of whether a lesser degree of state involvement, such as acquiescence or even inability to control armed bands operating on its territory, could ever be enough to constitute an armed attack, but it seems implicit in its judgment that armed attack is narrower than this.88
References(p. 140) The Court followed the same approach in Armed Activities on the Territory of the Congo (DRC v Uganda) in considering Uganda’s claim that it was using force in self-defence against armed attacks by non-state actors from the territory of the DRC in the period from August 1998 till June 2003.89 The Court first examined the nature of the Ugandan operations in the DRC in order to determine whether they could qualify as self-defence. Uganda had claimed that its Operation Safe Haven was conducted on this basis. However, the Ugandan forces in a very short space of time moved rapidly beyond the border area. The Court examined whether, throughout the period when its forces were rapidly advancing across the DRC, Uganda was entitled to engage in military action in self-defence against the DRC. It held, first, that the objectives of Operation Safe Haven were not consonant with the concept of self-defence as understood in international law. Second, Uganda did not argue that the regular army of the DRC was making attacks; it claimed that the ADF, a rebel group operating against Uganda from Congolese territory, was being supplied and equipped by the Sudan and the DRC government. But the Court found no evidence of a tripartite conspiracy between the DRC, the ADF, and the Sudan. The Court accepted that there was evidence of a series of cross-border attacks by the ADF from May 1998. However, Uganda had not shown any involvement of the DRC in these attacks. Moreover, third, Operation Safe Haven seemed to be essentially preventative in nature—to secure Uganda’s legitimate security interests—but Uganda was not claiming any right to anticipatory self-defence; Uganda was relying on self-defence against attacks that had occurred. However, it had not reported its use of force to the Security Council under Article 51.
It was after its recital of all these factors (which would in themselves cumulatively go a long way towards the undermining of Uganda’s claim to self-defence) that the Court then considered the question as to whether the DRC was responsible for the cross-border attacks by the ADF.90 Uganda did not claim that it had been subjected to an armed attack by the armed forces of the DRC. The armed attacks to which reference was made came from the ADF. The Court found that there was no satisfactory proof of the involvement in these attacks, direct or indirect, of the government of the DRC. It invoked the Definition of Aggression Article 3(g) to conclude that on the evidence before it the attacks were not attributable to the DRC. In its discussion of the prohibition of intervention, the Court rejected the argument that absence of action against rebel groups operating against another state amounted to tolerating or acquiescing in their activities. Inability to act against rebels or ineffective action did not make a state guilty of intervention. Accordingly it would seem that mere inaction cannot amount to an armed attack. References(p. 141) For all these reasons it found that the legal and factual circumstances for the exercise of a right of self-defence by Uganda against the DRC were not present. This is a clear reaffirmation of the Court’s position in the Nicaragua case.
Then follows what has proved to be the most controversial part of the judgment: ‘Accordingly, the Court has no need 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.’91 That is, the Court avoided the questions whether there may be an armed attack by non-state actors in the absence of state involvement, and what measures a state may take against such an attack.92
The Court was criticized by some judges and commentators for not taking a more radical approach.93 Judge Simma and Judge Kooijmans both said that the Court should have taken the opportunity presented by the case to clarify the state of the law on a matter which is marked by great controversy and confusion: self-defence against armed attacks by non-state actors. In their Separate Opinions they argued that if armed attacks are carried out by irregular forces from a state which does not exercise effective authority over its territory these activities are still armed attacks, even if they cannot be attributed to the territorial state, and that it would be unreasonable to deny the attacked state the right to self-defence.94 In References(p. 142) support of their argument they claimed that the events of 9/11 had brought about a change in the law. However, the significance of 9/11 and its aftermath for the general law of self-defence remains an extremely controversial question, and it is not surprising that the Court chose to avoid it.95
It is interesting that Judges Buergenthal and Higgins did not give Separate Opinions in this case, and did not here raise the question of armed attacks by non-state actors, as they had done earlier in the Court’s Advisory Opinion on Legal Consequences of the Construction of a Wall on Occupied Palestinian Territory.96 The Court in that case avoided any pronouncement on the possibility of self-defence against an armed attack by non-state actors. However, its discussion of self-defence was so brief and opaque that many have interpreted it as an express rejection of this doctrine.97 In the relevant paragraph on self-defence the Court said: ‘Article 51 of the Charter thus recognizes the existence of an inherent right of self-defence in the case of armed attack by one State against another State.’98 This is simply a statement of the basic central right of self-defence. The Court does not say there is a right of self-defence only in the case of an armed attack by one state against another.99 However, in their Separate Opinions Judges Higgins and Buergenthal seem to interpret it in this way.100 In contrast Judge Koojmans said that the Court had bypassed the issue.101 Israel was not claiming a right of self-defence against another state in this case. The Court then went on, ‘The Court also notes that Israel exercises control in the Occupied Palestinian Territory and that, as Israel itself states, the threat which it regards as justifying the construction of the wall originates within, and not outside, that territory. The situation is thus different from that contemplated by Security Council Resolutions 1368 (2001) and 1373 (2001), and therefore Israel could not in any event invoke those resolution in support of its claim to be exercising a right of self-defence. Consequently, the Court concludes that Article 51 of the Charter has no relevance in this case.’ This could be interpreted as not only rejecting Israel’s claim to self-defence, on the ground that it is an occupying power, but also as leaving open the possibility References(p. 143) of self-defence against non-state actors in situations like those contemplated in Security Council Resolutions 1368 and 1373 and as avoiding taking a position on the issue of principle.102 This question will be discussed further in Chapter 5.
‘Victim’ states have tended to blame the host state for incursions by armed bands operating from its territory and to hold it responsible in order to justify their invocation of self-defence; that is, they seem implicitly to take the view that if there is no state involvement in the actions of the irregular forces there can be no self-defence against that state but only lesser action not going beyond the territory of the victim state. The best-known practice is that of Israel, South Africa, and Portugal: they all took extensive action against irregular forces in neighbouring states.
Portugal’s reluctance to give up its colonial possessions in Africa led it into conflict with national liberation movements and newly independent African states. In the 1960s and 1970s Guinea, Senegal, and Zambia repeatedly complained of armed invasions by Portugal from its colonies. Portugal argued in response that it was acting in self-defence because these states were responsible for the acts of terrorists operating from their territories against its colonies.103 Similarly South Africa’s apartheid regime and illegal occupation of Namibia led to conflict; also on the independence of Angola and Mozambique, South Africa intervened in the Cold War-fuelled conflicts in those states. It said in justification of its invasions of the frontline states, Angola, Botswana, Mozambique, and Zambia, that these states had been supporting terrorist operations by the African National Council (ANC) and the South West Africa People’s Organization (SWAPO), or acquiescing in their operations, or allowing their territory to be used by them.104 Israel had been involved in cross-border actions against irregular forces operating from neighbouring states since 1948; in particular, it undertook operations against forces in Lebanon from 1967. It held Lebanon responsible for not preventing armed action against Israel and claimed the right to take action in self-defence.105 (p. 144) The precise degree of host state involvement alleged has varied, but Portugal and Israel seemed to feel the need to assert some degree of state involvement in the cross-border activities of the armed bands. At the widest, failure to prevent, or mere acquiescence in, the activities of armed bands was claimed not only to cause state responsibility but also to justify self-defence.
In contrast, South Africa sometimes did not allege state complicity to justify its use of force, but rather relied on a novel doctrine of ‘hot pursuit’. This is a law of the sea doctrine whereby coastal states have the right to pursue ships guilty of offences in territorial waters into areas of the sea beyond national jurisdiction; by analogy with this South Africa claimed the right to pursue alleged terrorists into neighbouring states.106 But this doctrine was not well received; in Resolution 568 (1985) the Security Council said that it ‘denounces and rejects racist South Africa’s practice of “hot pursuit” to terrorize and destabilize Botswana and other countries in southern Africa’. South Africa later abandoned this argument and returned to asserting the responsibility of the state from whose territory the guerrillas were operating.107
Although, as the Court recognized in the Nicaragua case, in principle self-defence is permissible against attacks by irregular forces where there is substantial state involvement in the sending of those forces, in practice the claims by Portugal, South Africa, and Israel to be acting in self-defence were generally not accepted by the Security Council. These claims to self-defence were undermined by the fact that the states invoking self-defence were regarded as being in illegal occupation of the territory they were purporting to defend. Portugal’s defence of its colonial possessions in Africa led it into conflict with forces fighting for decolonization; its attacks on states such as Guinea, Zambia, and Senegal were condemned by the Security Council. Many of the states arguing for condemnation did so because Portugal was using force to maintain its illegal colonial power. The right of self-defence could not be invoked to perpetuate colonialism and to flout the right to self-determination and independence.108 South Africa was in illegal occupation of Namibia and therefore many states were not willing to accept that it could use force in self-defence to protect the regime in Namibia.109 South Africa’s claims References(p. 145) to be acting in self-defence against incursions by SWAPO fighters seeking the liberation of Namibia were not valid. South African territory was not in danger; the cause of the dangerous situation was the illegal presence of South Africa in Namibia. Even France and the USA sometimes took this line; they continued to make this argument even when they later vetoed condemnations of South Africa in pursuit of their policy of constructive engagement or seeking a negotiated solution.110 Very unusually, in its resolutions rejecting South Africa’s justifications for its use of force against Angola and condemning this use of force the Security Council expressly asserted the right of Angola to self-defence under Article 51 of the UN Charter in Resolutions 546 (1984) and 574 (1985). Again the mere fact that many states regarded Israel’s occupation of the West Bank and Gaza, the Golan, and (until 2000) areas of South Lebanon as illegal was enough for them to condemn Israel’s use of force against cross-border attacks by irregulars. They say that Israel has no right to be in these territories and so no right to invoke self-defence against attacks on their forces in these territories or against attacks on Israel designed to secure its withdrawal from the territories it occupied illegally.111
The use of force against neighbouring states by Portugal, South Africa, and Israel was condemned on many different grounds. States in the Security Council debates on the use of force by these states mentioned various factors as contributing to the illegality of their actions in different cases: the neighbouring states were not responsible for any armed attack; the response to cross-border incursions was disproportionate; the use of force was not necessary. All three grounds for condemnation were invoked in the responses to the massive invasions and lengthy occupations of Lebanon by Israel in 1978 and 1982, and the South African operations in Angola from 1981.112 Also in some cases actions by South Africa and Israel were seen as unlawful reprisals rather than self-defence; the states using force regularly said that their aim was to prevent future attacks.113 And for many states the use of force by Portugal, South Africa, and Israel was illegal because it was directed against the legitimate struggle of a people with the right to self-determination.
Even when Western states on the Security Council abstained or vetoed a resolution condemning the use of force by Portugal, South Africa, and Israel, they References(p. 146) did not necessarily do so because they defended the legality of the actions of those states. In the early days of the decolonization struggle against Portugal, the USA and the UK sometimes abstained on the grounds that the facts of the particular case had not been properly established or there were extenuating circumstances.114 As regards South Africa, the USA and UK sometimes abstained as part of their policy of seeking a negotiated solution.115 Occasionally they said that they regarded the resolution as one-sided because they wanted an end to all foreign intervention in Angola.116 They used the veto to prevent the imposition of mandatory economic sanctions. As regards Israel, they sometimes said that they would abstain or veto a resolution because it was one-sided in that it did not condemn terrorist attacks against Israel.117 But for the most part, Portugal, South Africa, and Israel were regarded as not able to invoke self-defence because of their illegal occupation of territory. Issues of the precise involvement of states in the actions of armed bands thus did not have to be determined.
Other more straightforward claims to self-defence against irregular forces operating from neighbouring states have also been made. These have generally been more limited operations than the long-term and extensive action by the three states discussed above. For example, Thailand pursued guerrillas into Burma in 1995 after warning Burma to control the cross-border attacks by the guerrillas.118 Senegal similarly went into Guinea-Bissau in operations against opposition forces based in Guinea-Bissau in 1992 and 1995. In the latter case it is interesting that in 1992, when Guinea-Bissau protested that it had not supported the rebel incursions into Senegal, Senegal apologized for its action.119 Tajikistan was involved in more extensive actions against irregular forces operating from Afghanistan. On attaining independence in 1991 Tajikistan became involved in a civil war which continued until the 1997 General Agreement on the Establishment of Peace and National Accord in Tajikistan; forces opposing the government made cross-border attacks from Afghanistan. Tajikistan blamed Afghanistan for supporting the opposition forces and claimed the right to act in self-defence against the References(p. 147) armed bands in Afghanistan.120 More recent practice after the 9/11 attacks will be discussed in Chapter 5.
Turkey, Iraq, and the Kurds
Special problems over responses to cross-border attacks by irregular forces arose with regard to Turkey’s actions against the Kurds in Iraq.121 As part of its domestic campaign against the Kurdish Workers Party (PKK), a Kurdish separatist organization involved in terrorism since the 1970s,122 Turkey has undertaken cross-border operations against Kurdish bases in northern Iraq. These operations escalated after the Iraqi invasion of Kuwait in 1990, the imposition of the cease-fire on Iraq under Resolution 687, and the creation of ‘safe havens’ for the Iraqi Kurds in northern Iraq patrolled by US and British aircraft operating from Turkey. As long as Iraq acquiesced or even formally consented to Turkey’s operations on its territory, and in the absence of any international support for the right of the Kurds to independent statehood, Turkey at first offered little in the way of legal justification for its cross-border operations against Kurds.123
Since 1991 Iraq repeatedly protested at Turkey’s incursions. It complained of the penetration by Turkish armed forces inside Iraqi territory on the pretext that they were in pursuit of separatist terrorists; this was a violation of the UN Charter and of international law.124 Like Turkey, Iran also occasionally pursued Kurds over the border into Iraq; when it did so, it did not directly accuse Iraq of supporting the ‘bands of armed and organised terrorist mercenaries’ engaged in trans-border military attacks against and sabotage in Iranian border provinces. However, Iran did expressly invoke self-defence as a justification for its operations. It said that ‘in response to these armed attacks from inside Iraq and in accordance with Article 51 of the Charter of the United Nations, the fighter jets of the Islamic Republic Air Force carried out a brief, necessary and proportionate operation against the military bases of the terrorist group where the recent armed attacks had originated’.125
In contrast, Turkey has not expressly invoked Article 51; it did not itself report its operations in Iraq to the Security Council. It normally only responded (usually (p. 148) belatedly) to Iraq’s allegations; even then it did not clearly rely on self-defence. Thus Turkey avoided the issue as to how far Iraq was responsible for the actions of the Kurds and whether it was guilty of an armed attack, but left the legal basis for Turkey’s actions unclear. After a major operation in 1995 Turkey said:
As Iraq has not been able to exercise its authority over the northern part of its country since 1991 for reasons well known, Turkey cannot ask the Government of Iraq to fulfil its obligation, under international law, to prevent the use of its territory for the staging of terrorist acts against Turkey. Under these circumstances, Turkey’s resorting to legitimate measures which are imperative to its own security cannot be regarded as a violation of Iraq’s sovereignty. No country could be expected to stand idle when its own territorial integrity is incessantly threatened by blatant cross-border attacks of a terrorist organization based and operating from a neighbouring country, if that country is unable to put an end to such attacks. The recent operations of limited time and scope were carried out within this framework.126
It is very striking that the USA, in defending the Turkish action, apparently took the view that it was acting in self-defence.127 Turkey itself did not make this claim, although the language it used was similar to that in more recent claims to a wide right to self-defence put forward by the USA to justify its targeted killing programme and its intervention in Syria.128 And in other letters to the Security Council, in response to Iraqi protests about its cross-border actions in 1996 and 1997, Turkey again did not mention Article 51 or self-defence; it referred to the duty in the Friendly Relations Resolution to refrain from acquiescing in organized activities within its territory directed towards the commission of terrorist acts in another state. It relied on the principles of necessity and self-preservation. It also referred to its determination to take measures to safeguard its legitimate security interests, defending its borders and protecting its people against terrorism.129
Iraq claimed, with some plausibility, that there was a double standard: while the UN claimed that it was protecting the Kurds in Iraq against the Iraqi government, it closed its eyes to persecution by Turkey of its own Kurds. Iraq said that it could not be held responsible for the incursions by Kurds from its territory into Turkey because of the abnormal situation in northern Iraq, created particularly by the USA. It said that the US policy of interference and the deployment of US and British forces in Turkey in order to intervene militarily in northern Iraq prevented Iraq from exercising its sovereignty there. Iraq complained of the inaction of the Security Council:130 in spite of the absence of a clear legal justification for its use of force, Turkey avoided condemnation by the Security Council. The support of the USA apparently helped it to escape discussion of its actions. But References(p. 149) condemnation was expressed by the Arab League, the Gulf Cooperation Council, and the Non-Aligned Movement.131
In 2003 Operation Iraqi Freedom brought about the overthrow of Saddam Hussein and the installation of a new regime in Iraq which provided for the establishment of an autonomous Kurdish region in northern Iraq. Turkey was concerned that this would strengthen claims by the PKK to autonomy for the Turkish Kurds; it alleged that the Kurdish regional government in Iraq provided aid to the PKK and turned a blind eye to its cross-border incursions into Turkey. The Turkish parliament passed a resolution on 17 October 2007 authorizing cross-border operations against the PKK.132 And at the end of 2007 Turkey mounted a series of cross-border air raids into Iraq in pursuit of the PKK, with the apparent acquiescence of the USA.133 There were newspaper reports that the USA had agreed to provide intelligence to Turkey as to the movements of the PKK in Iraq in the hope that by allowing limited cross-border operations it might avoid a full-scale invasion.134 At a meeting with the Prime Minister of Turkey, President Bush said ‘PKK is a terrorist organization. They’re an enemy of Turkey, they’re an enemy of Iraq, and they’re an enemy of the United States.’135 In February 2008, with US intelligence support, Turkey carried out a major ground operation against the PKK in northern Iraq, the first since 2003. Iraq protested about the violation of its sovereignty, and the USA eventually called on Turkey to end its operation.136 Turkey did not report its cross-border operations to the Security Council under Article 51, and it has not offered any detailed legal justification for its use of force. This may be taken as an indication that it was unsure as to the adequacy of its legal case, and that states using force in what they see as a ‘war on terror’ may feel less constrained to offer such legal justification.137 Turkey continued its cross-border air strikes and ground operations against the PKK in Iraq until the PKK proclaimed a ceasefire in 2013. Turkey did so with the cooperation of Iraq and the USA from 2008 until 2012,138 and it did not report to the Security Council under Article 51 to justify these actions.139 From 2008 Turkey also coordinated its military actions with those of Iran against the Kurdistan Free Life (p. 150) Party (PJAK), a Kurdish group said to be linked to the PKK, that operated from northern Iraq against the government of Iran.140 Like Turkey, Iran did not report these operations to the Security Council under Article 51.
Iranian Oil Platforms case
A different set of questions with regard to ‘armed attack’ came up in the Iranian Oil Platforms case.141 Could an attack on a single US-flagged or US-owned merchant vessel amount to an armed attack under Article 51? Could an attack on a single naval vessel? What degree of gravity was necessary for an attack to constitute an armed attack; could a series of minor attacks cumulatively amount to an armed attack? Was it necessary that the attacking state have an intent to attack the particular victim state? This case arose out of the 1980–8 conflict between Iran and Iraq. After Iraq invaded Iran, the conflict spread beyond the land to the waters of the Gulf and affected commerce and navigation in the region. In 1984 Iraq initiated the ‘Tanker War’, attacking ships in the Gulf, particularly tankers carrying Iranian oil; Iran responded against vessels trading with Iraq. In order to ensure the safety of its merchant vessels in the Gulf, Kuwait turned to the USA, the UK, and the Soviet Union to re-flag some of its vessels and thus to ensure their naval protection. The USA agreed to provide all US-flagged ships with a naval escort through the Gulf; these convoys began in July 1987. Other foreign powers took parallel action. But a number of ships suffered attack or struck mines in the Gulf.142
Two specific attacks were central to this case. First, on 16 October 1987 the Kuwaiti tanker Sea Isle City, re-flagged by the USA, was hit by a missile. The USA blamed Iran for this attack, and three days later it attacked Iranian offshore oil installations, claiming to be acting in self-defence on the basis that the oil platforms had been engaged in a variety of actions directed against US vessels and other non-belligerent vessels and aircraft. Second, on 14 April 1988 the US warship Samuel B Roberts struck a mine in international waters in the Gulf; four days later the USA attacked and destroyed further Iranian oil platforms.143
Iran brought a case to the International Court of Justice, saying that the USA had violated Article X of the 1955 Treaty of Amity guaranteeing freedom of commerce and navigation between the two states. The USA counterclaimed that Iran had also broken the treaty ‘in attacking vessels in the Gulf with mines and missiles and otherwise engaging in military actions that were dangerous and detrimental to commerce and navigation between the territories of the two states’. Iran References(p. 151) denied responsibility for these attacks, suggesting that they were committed by Iraq, and claiming that Iran’s attitude was purely defensive. The Court examined the question whether the USA had demonstrated that it had been the victim of an armed attack by Iran such as to justify its using armed force in self-defence; it held that the burden of proof rested on the USA.144 The Court first dealt with the missile attack on the Sea Isle City; it examined in detail the evidence relating to that incident and concluded that the USA had not produced sufficient evidence to support its contentions that Iran was responsible for the missile attack. ‘The burden of proof of the existence of an armed attack by Iran on the United States, in the form of the missile attack on the Sea Isle City, has not been discharged.’145
The USA had not relied solely on the Sea Isle City incident as constituting the ‘armed attack’ to which it claimed to be responding in self-defence. There is considerable doubt as to whether a single attack on a merchant vessel (as opposed to a military vessel) could constitute an armed attack on a state and the Court itself did not directly address this issue.146 The USA asserted that this was the latest in a series of such missile attacks against US-flagged and other non-belligerent vessels in Kuwaiti waters. The alleged pattern of Iranian use of force ‘added to the gravity of the specific attacks, reinforced the necessity of action in self-defence, and helped to shape the appropriate response’. The USA is here clearly trying to address the argument that a single episode could not be serious enough to amount to an armed attack and that its response was disproportionate. The USA set out a series of incidents involving US-flagged or US-owned vessels and aircraft in the period up to the end of the conflict: the mining of the US-flagged Bridgeton, and of the US-owned Texaco Caribbean, the firing on US navy helicopters and the minelaying by an Iranian vessel, the Iran Ajr.147
However, even assuming that all these events were attributable to Iran, the Court held that the USA had not shown that the series of incidents could be categorized as an armed attack. The Court concluded that even taken cumulatively these References(p. 152) incidents did not constitute an armed attack on the USA of the kind which met the Nicaragua test of constituting a ‘most grave’ form of the use of force.148 This left open the question whether an accumulation of events can amount to an armed attack.149 Attacks on US-owned (as opposed to US-flagged) vessels did not amount to an attack on the state.150 There was no evidence that the minelaying alleged to have been carried out by the Iran Ajr, at a time when Iran was at war with Iraq, was aimed specifically at the USA; similarly it had not been established that the mine which struck the Bridgeton had been laid with the specific intention of harming that ship or other US vessels. That is, it had not been established that the incidents were aimed at the USA (as opposed to Iraq). The Court apparently decided that harm by a mine or a missile constitutes an armed attack on a third state during a conflict between two other states only if the attack was specifically aimed at that third state. This is a brief and rather obscure discussion of a difficult issue; the Court does not go into any greater detail as to the element of intent apparently required by the notion of armed attack in this particular context or as to the general significance (if any) of its approach.151
The US State Department Legal Adviser was very critical of the Court’s judgment on this point.152 He claimed that the need to prove a specific intent would undermine international peace and security; a requirement of specific intent would encourage intentionally indiscriminate attacks, since no victim would have the right to defend against them. However, this does not seem convincing, especially given the prohibitions on indiscriminate attacks in international humanitarian law. And it is not clear whether the Court was trying to establish a general requirement for all armed attacks or whether its brief statements on the intent requirement should be limited to the particular and unusual facts of the case where there was US involvement in a conflict between two other states.153
The second US attack on Iranian oil installations took place on 18 April 1988 after the Samuel B Roberts was blown up by a mine in international waters; again the USA claimed to be acting in self-defence.154 The Court noted that the attacks on the oil platforms were not an isolated operation aimed simply at the oil installations, as had been the case with the attacks of 19 October 1987; rather the US attacks of 18 April 1988 formed part of a much more extensive military action, References(p. 153) designated Operation Praying Mantis, and directed against a number of targets, including two Iranian frigates and other Iranian naval vessels and aircraft. The USA had discovered several mines bearing Iranian serial numbers in the vicinity in the days following the attack, and it also adduced other evidence of Iranian minelaying activity. But Iran denied that it had systematic recourse to minelaying in the Persian Gulf, and suggested that the mine that hit the Samuel B Roberts might have been laid by Iraq. The Court held that because mines were being laid by both Iraq and Iran at the relevant time evidence of other minelaying operations by Iran was not conclusive as to responsibility of Iran for the particular mine which had hit the Samuel B Roberts. The evidence produced by the USA was highly suggestive but not conclusive.155 The USA in its communications to the Security Council had linked the attack on the Sea Isle City with a series of offensive acts and provocations by Iranian naval forces against neutral shipping in the Gulf. Before the Court the USA argued (as it had with regard to the Sea Isle City) that the pattern of Iranian use of force ‘added to the gravity of the specific attacks, reinforced the necessity of action in self-defense and helped to shape the appropriate response’. The Court did not exclude the possibility that the mining of a single military vessel might alone be sufficient to bring into play the inherent right of self-defence, but in view of all the circumstances, including the inconclusiveness of the evidence of Iran’s responsibility for the mining of the Samuel B Roberts, the Court was unable to find that the attacks on the oil platforms were justifiably made in response to an armed attack on the US by Iran.156
Gravity of attack
The Oil Platforms judgment raised the question of the gravity of an armed attack. It quoted the passage of the Nicaragua case in which the Court famously and controversially said that ‘It is necessary to distinguish the most grave forms of the use of force (those constituting an armed attack) from other less grave forms.’157 In its examination of the scope of the prohibition of the use of force and the meaning of armed attack it went on to say that mere ‘frontier incidents’ did not constitute armed attacks.158 When the Court came to apply customary international law to the facts of the Nicaragua case, it asked whether the US actions using force against Nicaragua were justified as collective self-defence.159 Did Nicaragua engage in an armed attack against El Salvador, Costa Rica, and Honduras? The Court said that the limited provision of arms from Nicaragua to the opposition in El Salvador did not amount to an armed attack.160 Therefore the concept of frontier incident did (p. 154) not play any clear role in this part of the judgment. In contrast, the Court held that there had been certain trans-border incursions from Nicaragua into Costa Rica and Honduras imputable to the government of Nicaragua.161 Here, it seems that the distinction between frontier incident and armed attack was important to the Court. If these trans-border incursions amounted to armed attacks, then it would be possible that the USA might have a claim to collective self-defence of Costa Rica and El Salvador. However, the Court was rather non-committal (and it did not expressly mention frontier incidents or elaborate on the distinction between armed attack and frontier incident at this point in its judgment). It said only that it had very little information as to the circumstances or possible motivations of the incursions and this rendered it difficult to decide whether they could be treated for legal purposes as amounting either singly or collectively to an armed attack by Nicaragua on either or both of these states.
The Court did not elaborate in any detail on the distinction between frontier incidents and armed attack. The first distinguishing features it mentioned were the ‘scale and effects’ of the attack;162 this formula is comparable to the exclusion of ‘acts or their consequences … not of sufficient gravity’ from the 1974 Definition of Aggression and would seem to cover scale in place and time and also the scale of the impact of the attack.163 It is clear from the context of the Court’s pronouncement that the difference envisaged is one of degree rather than of kind; that is, the concept of frontier incident was not limited to acts of non-state organs.
The second set of distinguishing features mentioned by the Court are more obscure; they are the ‘circumstances and motivations’ of the attack.164 This phrase is very general; the implication seems to be that the Court would include within ‘frontier incident’ episodes where there was no intent to carry out an armed attack, including accidental incursions and incidents where officials disobeyed orders. The question of motivation is a controversial one. Can a state’s motive be anything more than an inference from the action in question? Is the intent of individual soldiers to be attributed to the state? Factors of motive and intent were much discussed during the drafting of the Definition of Aggression; there was fundamental disagreement as to whether an act could constitute an act of aggression simply because the use of force was intentional or whether there should be some further intention on the part of the state to commit aggression (animus (p. 155) aggressionis).165 The Court in Nicaragua left these questions of intent and motive with regard to frontier incidents unresolved.166
Criticism of the distinction between armed attack and frontier incident
At first sight it might seem that the distinction drawn in a sketchy way in the Nicaragua case between armed attacks and lesser incursions such as frontier incidents was illogical and unnecessary. Given that all self-defence, whether individual or collective, must be necessary and proportionate, a minor frontier incursion would justify only a very limited response. Thus there would seem to be no need to distinguish between armed attacks allowing self-defence and mere frontier incidents. The necessity and proportionality requirements would provide adequate safeguard against excessive use of force.167
Many harsh criticisms were made of the Court for its narrow view of armed attack, and its consequent limitation of the US right to act in collective self-defence. As part of this, some writers condemned the Court’s distinction between frontier incidents and armed attacks.168 Reisman accuses the Court of developing a theory that is tolerant of different forms of protracted and low-intensity conflict; he argues that this will lead to an increase in violence in international politics.169 This argument echoes that of Fitzmaurice, writing in 1933, who argued that it was important not to treat frontier incidents as a justified resort to force; this would encourage frontier incidents and place innocent states in a difficult position.170 The writers critical of the reasoning in the Nicaragua case follow the same line as earlier commentators on the UN Charter who said that any attack, even small border incidents, allowed self-defence.171 Thus Brownlie had References(p. 156) expressed doubts about the concept of frontier incidents. He writes of the concept as ‘vague’, and says that from the point of view of assessing responsibility the distinction between frontier incident and armed attack is only relevant insofar as the minor nature of the attack is prima facie evidence of absence of intention to attack, of honest mistake, or simply the limited objectives of an attack. For him, the question as to whether the particular use of force is permissible self-defence is merely one of proportionality.172
Arguments for the distinction between armed attack and frontier incident
But it is clear, despite the criticisms of the Nicaragua case, that there were nevertheless serious reasons for the Court’s distinction between armed attacks and other less grave forms of the use of force, including frontier incidents. Its concern was with collective self-defence; it wanted to limit third state involvement. Its insistence on a high threshold for armed attack would serve to limit third-party involvement. If there was no armed attack, there could be no collective self-defence. The use of necessity and proportionality alone would not exclude third party involvement, merely limit the scope of their permissible response.173 Judge Jennings, in his Dissenting Opinion in the Nicaragua case, expressed some limited sympathy with the Court’s approach to collective self-defence. He said that, ‘It is of course a fact that collective self-defence is a concept that lends itself to abuse. One must therefore sympathize with the anxiety of the court to define it in terms of some strictness. There is a question, however, whether the court has perhaps gone too far in this direction.’174
Moreover, in the Oil Platforms case the Court reaffirmed the gravity requirement in the context of the US involvement in the Iran/Iraq conflict.175 That is, the Court applied the concept to limit the right of individual self-defence, and it accepted a general distinction between armed attacks and less grave forms of the use of force. The US State Department Legal Adviser was again critical of the Court’s judgment on this point.176 He argued that the gravity requirement should be limited to its original context in the Nicaragua case, that of the sending by a state of irregular armed bands; he rejected the proposition that the use of deadly force by a state’s regular armed forces, such as the attacks by Iran in this case, does not qualify as an armed attack unless it reaches a certain level of gravity. Like the earlier critics of the Nicaragua case on this point, he claimed that the (p. 157) requirement that an attack reach a certain level of gravity before triggering a right of self-defence would make the use of force more rather than less likely, because it would encourage states to engage in a series of small-scale military attacks, in the hope that they could do so without being subject to defensive responses. This implausible empirical claim has not been borne out by practice since the Nicaragua case. The counter-argument could be made that the presumption should always be against the use of force and in favour of peaceful settlement. If there is no gravity requirement for an armed attack and self-defence, then an inter-state conflict could arise out of minor cross-border incidents or other minor uses of force. However, the US State Department Legal Adviser expressly accepted the distinction between armed attack and frontier incident.177
The distinction between armed attack and other less grave use of force played a crucial role in the Eritrea/Ethiopia Claims Commission Award on Ethiopia’s Ius ad Bellum Claims 1–8.178 This decision was made in the context of a traditional inter-state conflict. The crucial issue before the tribunal was whether Eritrea had started the 1998–2000 conflict and should be held responsible for all the harm to Ethiopia caused by that conflict. The parties disagreed as to the starting point of the conflict. Ethiopia said it began with armed attacks by Eritrea on 12 May 1998; Eritrea said that those actions were taken in self-defence against Ethiopian forces in illegal occupation of its territory; it also claimed that the 1998 conflict had originated earlier on 6 May. The Commission said that the parties had given very different accounts as to the location of the incidents on 6 and 7 May and of the numbers and types of forces involved. However, it had no need to resolve these differences because it was clear that these incidents were ‘geographically limited clashes’ between small Eritrean and Ethiopian patrols along a remote, unmarked, and disputed border. The Commission was satisfied that these relatively minor incidents were not of a magnitude to constitute an armed attack by either state within the meaning of Article 51. This is a clear affirmation of the need for a use of force to reach a certain level of gravity before it constitutes an armed attack.
Necessity and proportionality
As part of the basic core of self-defence all states agree that self-defence must be necessary and proportionate.179 The requirements of necessity and proportionality are often traced back to the 1837 Caroline incident, involving a pre-emptive attack by the British forces in Canada on a ship manned by Canadian rebels, References(p. 158) planning an attack from the USA.180 This episode has attained an authority far beyond what can be legally justified. States and writers still refer to it, generally to support wide claims to self-defence, but also to support the necessity and proportionality limitation.181 They invoke the famous formula that there must be a ‘necessity of self-defence, instant, overwhelming, leaving no choice of means and no moment of deliberation’. Others challenge the authority of this episode for the modern doctrine of self-defence, seeing it rather as an episode of self-help pre-dating the modern law on self-defence. They challenge the claim that Article 51 preserves a wider customary law right of self-defence and deny that the Caroline incident was an example of pre-Charter self-defence.182 But, irrespective of the very dubious status of the Caroline incident as a precedent, necessity and proportionality have played a crucial role in state justification of the use of force in self-defence and in international response.
The Nicaragua case,183 the Advisory Opinion on the Legality of the Threat or Use of Nuclear Weapons,184 the Oil Platforms case,185 and Armed Activities on References(p. 159) the Territory of the Congo (DRC v Uganda)186 reaffirmed that necessity and proportionality are limits on all self-defence, individual and collective. These requirements are not express in the UN Charter, but are part of customary international law. There was until recently relatively little general academic discussion of these essential characteristics of self-defence, as opposed to discussion in application to particular incidents.187 The question whether self-defence lived up to these requirements was often treated as almost exclusively one of fact. However, there are issues of principle involved: most basically, the different views of the scope of the right of self-defence will affect the scope of proportionality. Thus those who accept a right of anticipatory self-defence or self-defence against an accumulation of events will assert a much wider concept of proportionality.188 Some states and commentators have argued that the ‘war on terror’ justifies a very wide interpretation of necessity and proportionality going beyond a response to an actual armed attack. The conflicts in Afghanistan after the 9/11 attacks, in Lebanon (2006), and Somalia (2006) have given rise to important issues of principle.189
In theory it is possible to draw a distinction between necessity and proportionality, and the ICJ typically applies the two requirements separately. Necessity is commonly interpreted as the requirement that no alternative response to an armed attack be possible.190 Proportionality relates to the size, duration, and target of the response, but clearly these factors are also relevant to necessity.191 It is not clear how far the two concepts can operate separately. If a use of force is not necessary, it cannot be proportionate and, if it is not proportionate, it is difficult to see how it can be necessary. Commentators agree on a few, basic, uncontroversial principles on self-defence in response to an armed attack: necessity and proportionality mean that self-defence must not be retaliatory or punitive; the aim should be to halt and repel an attack. This does not mean that the defending state is restricted to the same weapons or the same numbers of armed References(p. 160) forces as the attacking state; nor is it necessarily limited to action on its own territory. Reprisals are generally agreed to be unlawful. The General Assembly made this clear in the Declaration on Friendly Relations and the Resolution on the Inadmissibility of Intervention. The Security Council also passed Resolution 188 in 1964, in response to a British attack on Yemen, but declaring in absolute terms that it condemned reprisals as incompatible with the purposes and principles of the UN. The distinction between reprisals and self-defence is sometimes problematic in practice.192
In the Nicaragua case the Court treated these limitations of necessity and proportionality as marginal considerations. That is, the use of force by the USA was first held not to qualify as lawful self-defence on other grounds, then its illegality was confirmed because the actions were not necessary or proportionate. The Court applied the requirements in turn. Even if the supply of arms from Nicaragua to opposition forces in El Salvador had amounted to an armed attack, the measures taken by the USA against Nicaragua were not necessary because they were taken months after the major offensive of the opposition against the government of El Salvador had been completely repulsed. Nor were the US activities relating to the mining of the Nicaraguan ports and attacking oil installations proportionate to the aid received by the Salvadoran opposition from Nicaragua.193 Thus the questions of necessity and proportionality are dependent on the facts of the particular case.
The inquiry into necessity and proportionality in the Nicaragua case was not necessary for the Court’s judgment on the merits; the US use of force had already been found to be illegal on other grounds. These criteria of necessity and proportionality were said by the Court to be an additional ground of wrongfulness. The same approach was taken in the Iranian Oil Platforms case; having held that the USA had failed to establish that Iran was responsible for an armed attack, the Court went on to add that the US response was not necessary and proportionate. One aspect to be taken into account in assessing this (as the Court had implicitly done in the Nicaragua case) was the nature of the target. In its communications to the Security Council the USA had indicated the grounds on which it regarded the Iranian platforms as legitimate targets for armed action in self-defence. The USA maintained that the platforms were used for collecting intelligence concerning passing vessels and that they acted as a military communication link coordinating Iranian naval forces and served as staging bases to launch helicopters and small boat attacks on neutral shipping. Iran acknowledged the presence of military personnel and equipment on some of the platforms, but said that their purpose was exclusively defensive and justified by previous Iraqi attacks on the References(p. 161) oil platforms. The Court said that it was not sufficiently convinced that the evidence supported the contentions of the USA and, even if those contentions were accepted, the Court was unable to find that the attacks on the platforms could have been justified as acts of self-defence. In the case of the attack on the Sea Isle City and the mining of the Samuel B Roberts the Court was not satisfied that the attacks on the platforms were necessary to respond to these incidents. There was no evidence that the USA had complained to Iran of the military activities of the platforms, which suggested that the targeting of the platforms was not seen as a necessary act.194
The Court gave separate, though brief, consideration to proportionality. It held that the attack of 19 October 1987 might have been proportionate if it had been found to be necessary, but the attacks of 18 April 1988 were conceived and executed as part of a more extensive operation. The Court could not close its eyes to the scale of the whole operation.195 As a response to a mining by an unidentified agency of a single US warship which was severely damaged but not sunk and without loss of life, the operation was not a proportionate use of force in self-defence.196
Although the Court found in the operative part of its judgment that there had been no armed attack on the USA and that it had not acted in self-defence when it attacked the oil platforms, the Court did not spell out consequences as some judges would have preferred.197 It limited itself to concluding that the actions carried out by the USA against Iranian oil installations on 19 October 1987 and 18 April 1988 could not be justified as ‘measures necessary to protect the essential security interests’ of the USA under Article XX of the Treaty of Amity, since those actions constituted recourse to armed force not qualifying under international law as acts of self-defence. As the Court went on to determine that these actions did not in fact interfere with commerce between Iran and the USA, it therefore did not specify consequences.
References(p. 162) Some judges went further than the Court in holding that these US actions were actually unlawful reprisals as they were not necessary and proportionate.198 The parties had produced radically opposed arguments on the scope of self-defence and its relation to reprisals. The crucial issue was the characterization of the clashes between Iran and the USA: were they a series of individual incidents or should they be seen as an ongoing conflict? Iran argued that as the USA was not a party in the Iran/Iraq conflict it was necessary to determine whether each single incident of the use of force could be justified as self-defence.199 It also took a narrow view of self-defence, arguing that it was only available against an armed attack and could not be anticipatory or preventive. Iran set out a fourfold test to determine whether a particular use of force was lawful self-defence or unlawful reprisals;200 the use of force should be timely, not disproportionate, not premeditated and should be directed against the correct target. Therefore it argued that the US actions were unlawful reprisals because their attacks on the oil platforms took place after the missile and mine had damaged the two ships, involved disproportionate force, were premeditated, and because the missile and mine attacks had not originated from the oil platforms.
The USA’s official position was that it was neutral in the Iran/Iraq conflict and not itself involved in an ongoing armed conflict with Iran. It accordingly tried to argue that the overall pattern of Iran’s continuing deadly and illegal use of force must be taken into account in applying the elements of the law of self-defence.201 It said that the right to self-defence was not limited to repelling an attack while it was in progress. The USA argued that Iran’s conception of self-defence was too limited and would embolden aggressor states. A state could also use force to remove continuing threats to future security.202 But the US position—even if acceptable as between two parties involved in an ongoing conflict—is difficult to maintain where the state claiming such a right is not itself a party. The Court did not have to pronounce on this issue as it found that there was no armed attack by Iran; nevertheless, it may be argued that it is implicit in the Court’s finding that the US use of force was not necessary that its actions were in fact reprisals.
(p. 163) Most recently the Court in DRC v Uganda followed the same approach. After its rejection of Uganda’s claim to be acting in self-defence, it said that there was no need for it to inquire into necessity and proportionality. It said only that ‘The Court cannot fail to observe that the taking of airports and towns many hundreds of kilometres from Uganda’s border would not seem proportionate to the series of transborder attacks it claimed had given rise to the right of self-defence, nor to be necessary to that end.’203
The ICJ’s approach in these cases—that of treating necessity and proportionality as marginal considerations, to be considered after the legality or otherwise of the use of force had already been established on other grounds—may seem a logical approach, but in state practice these factors of necessity and proportionality are often the only factors relied on in deciding the legality of particular actions. They constitute a minimum test by which to determine that a use of force does not constitute self-defence. In Security Council debates states have thus been able to avoid going into doctrinal disputes as to whether self-defence is wide or narrow; they can simply say that the use of force was not necessary or proportionate and therefore illegal. Thus, some states were willing to condemn Israel for disproportionate use of force in Lebanon in 2006, even though they did not go into the controversial issue as to whether the actions of Hezbollah could constitute an armed attack.204 Similarly, some states condemned Israel for its large-scale use of force against Gaza (2008–9, 2014) on the ground that it was not proportionate without going into the controversial question as to whether Israel can invoke self-defence against Gaza if it is still occupied territory.205 Moreover, those states which maintain a controversially wide view of self-defence allowing protection of nationals or anticipatory self-defence are able to make an argument rejecting wide claims to self-defence by other states without undermining their doctrinal position. For example, condemnation of Israel and South Africa for pre-emptive action was possible for states supporting the legality of anticipatory self-defence on the basis that the use of force was not necessary or proportionate on the particular facts.206 In the 2008 Georgian conflict, the USA and the UK were able to condemn Russia for the use of force to protect its nationals in Georgia on the ground that Russia went beyond what was necessary and proportionate, even References(p. 164) though they supported the legality of protection of nationals abroad in principle.207 Necessity and proportionality are also crucial in the rejection by states of the legality of prolonged occupation of territory in the name of self-defence. Thus Israel’s presence in South Lebanon from 1978 to 2000 and South Africa’s occupation of a buffer zone in Angola from 1981–8 were both claimed to be justified as self-defence and both repeatedly and universally condemned as not necessary or proportionate self-defence.208 And similarly the use of force in self-defence has not been accepted as a valid root of title to territory.209
Accumulation of events
The criteria of necessity and proportionality also help states to distinguish unlawful reprisals from lawful self-defence. In cases of repeated cross-border incursions commentators have spoken of the ‘accumulation of events’ or ‘pin prick’ theory of armed attack in order to justify an otherwise disproportionate response.210 That is, they claim that states may use force not in response to each incursion in isolation but to the whole series of incursions as collectively amounting to an armed attack. Such arguments were made by the USA with regard to Vietnam and by Israel, South Africa, and Portugal. Some have claimed that the Security Council has rejected this doctrine of accumulation of events and have criticized it for this.211 In fact the Security Council has not gone so far. It has certainly condemned disproportionate responses by Israel, Portugal, and South Africa, but as usual the condemnation did not address the doctrinal issue of the scope of self-defence; it could be interpreted as based strictly on the special facts of these cases.212 The ICJ in the Nicaragua case seemed to leave open the possibility of an accumulation of events amounting to an armed attack when it said of the trans-border incursions into Honduras and Costa Rica that ‘Very little information is available to the Court as to the circumstances of these incursions or their possible motivations, which renders it difficult to decide whether they may be treated for legal purposes as amounting, singly or collectively, to an armed attack by Nicaragua on either or both States.’213
References(p. 165) Similarly in Cameroon/Nigeria, Iranian Oil Platforms, and DRC v Uganda, the Court apparently contemplated the possibility of an ‘accumulation of events’ model of armed attack, but did not discuss this controversial question. In the first case Cameroon argued that Nigeria was responsible for several frontier incidents; it did not seek a ruling on each individual violation but on the incidents collectively. Nigeria replied that the Court must consider the incidents one by one. It was clear that Cameroon feared that each individual incident would be regarded by the Court as a mere frontier incident not amounting to an armed attack for which Nigeria was responsible. The Court avoided any ruling on the principle; it held that Cameroon had not sufficiently proved the facts or the imputability of the alleged incursions to Nigeria.214 In the Iranian Oil Platforms case the USA also invoked a series of incidents rather than just the attacks on the Sea Isle City and the Samuel B Roberts in order to justify its use of force against the oil platforms as self-defence. Iran specifically challenged the accumulation of events model of armed attack in its Pleadings,215 but the Court did not have to address the issue as it found that the USA had not shown that the events amounted to an armed attack on the USA or that they were imputable to Iran.216 The question also came up in passing in DRC v Uganda, but was not investigated by the Court.217 The related concept of a ‘continuing armed attack’ or an ‘ongoing armed attack’ by terrorists has recently been invoked by some states in the context of counter-terrorism operations in Syria to justify a wide right of self-defence.218
Protection of nationals
The use of force to rescue nationals in a foreign state without the consent of that state is uncommon and has been practised by only a few states since the Second World War. Nevertheless, it attracted a vast amount of academic debate during the Cold War. The interventions in Suez (1956), Lebanon (1958), Congo (1960), Dominican Republic (1965), in the Mayaguez incident (1975), Entebbe (1976), Iran (1980), Grenada (1983), and Panama (1989) have all been exhaustively discussed.219 Between 1989 and 2008 there were no significant instances References(p. 166) of use of force to rescue nationals.220 But in 2008 the Russian intervention in Georgia reignited interest in this issue.221 In these episodes all the states using force invoked self-defence as at least a partial justification for their action. For the most part they expressly referred to Article 51 as covering their operation.222 That is, these states and those who expressly supported them interpret the Charter as allowing the forcible protection not only of a state’s territory but also of its nationals abroad. The UK view was typical; it says that ‘the better view’ is that the justification comes from Article 51 as a form of self-defence: ‘An alternative, less satisfactory view is to seek to derive from customary international law a right of intervention to protect nationals.’223
The international response to these interventions showed a clear division among states, with only a few states accepting a legal right to protect nationals abroad. The legal arguments of Belgium, the USA, Israel, and the UK in favour of such a wide right to self-defence attracted few adherents. The USSR opposed the existence of this right until its dissolution led Russia to abandon this position and to proclaim a right to use force to protect Russian nationals in the former republics.224 The Security Council has generally not taken a collective view or has been prevented by the veto from condemnation.225 Its debates show the radical divisions between states on the doctrinal issue of the permissibility of the use of force to protect nationals. The General Assembly condemned the US interventions in Grenada and Panama, but these condemnations were not unequivocal. First, in References(p. 167) these particular episodes the US action went far beyond the protection of nationals and the USA offered other justifications for its intervention. In Grenada the US forces argued that US nationals were in danger after a socialist coup, but there was considerable controversy as to the reality of this danger. Moreover, the US forces did not simply rescue the nationals; they remained and oversaw the installation of a new government. To justify its intervention the USA used not only protection of nationals but also relied on an invitation by the Governor-General of Grenada and the claim that its action was regional peacekeeping under Chapter VIII of the Charter.226 In justification of its use of force in Panama the USA put more stress on protection of nationals than it had with regard to Grenada, but its actions clearly went far beyond this. The US forces again installed a new government. Moreover, just as in Grenada, there was controversy as to the existence of actual danger to US nationals. Its other main legal argument was that it was acting to defend the integrity of the Panama Canal Treaties. Therefore the grounds for condemnation of the US use of force were not necessarily based on the rejection of a wide doctrine of self-defence that covered protection of nationals.227 Second, the condemnations by the General Assembly in these two cases, especially in the case of Panama, were less than overwhelming.228
Some writers who seek to justify the use of force in protection of nationals seized on this failure to condemn by the Security Council and the failure to take any action against the state using force.229 They discounted the General Assembly votes and the rejection by a majority of states of such a doctrine. There is a clear division among writers on this question; some see intervention as furthering the purposes of the UN and attempt to derive from state practice conditions under which the right may be exercised.230 Essentially these are all variations on the early version offered by the UK over its intervention in Suez in 1956. It said that the relevant conditions were: (a) whether there is an imminent threat of injury to nationals; (b) whether there is a failure or inability on the part of the territorial sovereign to protect the nationals in question; and (c) whether the measures of References(p. 168) protection are strictly confined to the object of protecting them against injury.231 On the other side are those writers who regard intervention to protect nationals as of doubtful value in furthering the purposes of the UN as it may be a pretext for intervention and cause more harm than it prevents.232
Russia’s use of force in Georgia in 2008 revived the debate about the protection of nationals.233 The conflict began in South Ossetia, an Autonomous Region of Georgia, and later spread to Abkhazia, an Autonomous Republic of Georgia. There had been unrest in both areas since before the dissolution of the USSR and Georgia’s admission to the UN in 1992. Separatists resisted the incorporation of South Ossetia and Abkhazia into Georgia and there were violent clashes with Georgian government forces. Russian forces were sent into both areas with the consent of Georgia as part of a Commonwealth of Independent States (CIS) peacekeeping operation.234 The situation in South Ossetia and Abkhazia deteriorated after President Saakashvili came to power at the end of 2003; he sought to reassert full Georgian control over the two regions. Relations between Georgia and Russia also worsened when President Saakashvili sought closer relations with the west, including NATO membership.
Serious conflict broke out in 2008. At the time Georgia and Russia presented competing accounts of the series of events, but the Independent International Fact-Finding Mission on the Conflict in Georgia set up by the EU subsequently found that it was Georgia that started the conflict with its air and ground attack on Tskhinvali, the capital of South Ossetia on 7 August.235 Russia responded by the use of force, driving Georgian forces out of South Ossetia. The conflict spread to Abkhazia and then to the rest of Georgia over the next five days. Georgia and Russia accused each other of aggression. Russia’s legal justification for its use of force was that it was acting in self-defence of its citizens and of its peacekeeping forces under Article 51.236
In the series of Security Council debates on the conflict it was not possible for the USA and the UK to challenge the legality of protection of nationals abroad in principle, given their own support for this right to use force.237 They did, References(p. 169) however, claim that Russia was acting unlawfully because its actions were not necessary and proportionate, in that they extended beyond South Ossetia and Abkhazia into large parts of Georgia and involved the destruction of much of the modern military equipment supplied to Georgia by the USA.238 They challenged Russia’s motives for intervening, claiming that its aims went beyond self-defence and that it was seeking the secession of the two areas from Georgia.239 The UK also challenged Russia’s right to invoke self-defence in protection of its nationals in South Ossetia and Abkhazia, saying that it had issued passports opportunistically just before its intervention, and that the forcible protection of such recent nationals could not be lawful.240 The Independent International Fact-Finding Mission on the Conflict in Georgia accepted this argument;241 overall it took the view that Russia’s use of force to protect nationals abroad was not justified as self-defence in this case.242
Irrespective of the doctrinal divide, most of the above interventions clearly could not be justified as protection of nationals because the action was not necessary or proportionate and was really a pretext for intervention. Only the rescue operation of the Mayaguez, and those in Iran and Entebbe were limited actions.243 In Suez, the Dominican Republic, Grenada, Panama, and Georgia the interventions were prolonged and the states using force added further justifications. The underlying question as to whether an attack on nationals abroad can constitute an attack on a state for the purposes of Article 51 remains important in the context of the ‘war on terror’.244
The same states—the USA, Israel, and the UK—that claim a right under Article 51 to protect their nationals abroad also claim or defend the right to use force, even before their territory or units of their armed forces abroad are actually attacked, if an armed attack is imminent.245 However, the majority of states rejected anticipatory self-defence before the 9/11 attacks.246 The divisions between states as to the scope of the right of self-defence meant that no detailed provisions on self-defence could be included in General Assembly resolutions such as the Declaration on Friendly Relations, the Definition of Aggression, and the Declaration on the Non-Use of Force. It is interesting that those states which argued that self-defence is permissible only against an armed attack made this argument expressly, whereas those states who took a wider view of self-defence adopted a low profile and simply resisted the inclusion of any detailed provisions.247 Also in the International Law Commission (ILC) work on state responsibility when self-defence was considered as a circumstance precluding wrongfulness, the states taking a wide view of self-defence argued that the ILC should not try to define the scope of self-defence; they did not actually send in comments in favour of anticipatory self-defence.248 These differences persisted after 9/11, as was apparent in the debates leading up to the 2005 World Summit.249 The Non-Aligned Movement continues to argue that Article 51 ‘is restrictive and should not be re-written or re-interpreted’.250
Moreover, the actual invocation of the right to anticipatory self-defence in practice was rare. States clearly preferred to rely on self-defence in response to an armed attack if they possibly could. They preferred to take a wide view of armed attack rather than openly claim anticipatory self-defence. It was only where no conceivable case could be made that there had been an armed attack that they resorted to anticipatory self-defence. This reluctance expressly to invoke anticipatory self-defence was in itself a clear indication of the doubtful status of this justification for the use of force. States take care to try to secure the widest possible support; they do not invoke a doctrine that they know will be unacceptable to the (p. 171) vast majority of states. Certain writers, however, ignore this choice by states and argue that if states in fact act in anticipation of an armed attack this should count as anticipatory self-defence in state practice. This is another example of certain writers going beyond what states themselves say in justification of their action in order to try to argue for a wide right of self-defence.251
Thus in 1967 Israel launched what was apparently a pre-emptive strike against Egypt, Jordan, and Syria, but it did not seek to rely on anticipatory self-defence. It argued that the actions of the Arab states in fact amounted to prior armed attack. For example, in the Security Council debates Israel claimed that the blocking by Egypt of the Straits of Tiran to passage by Israeli vessels amounted to an act of war; it was an armed attack justifying self-defence under Article 51. Some states rejected this claim and ruled out the legality of anticipatory use of force; some said that it was not productive to apportion blame; even those supporting the Israeli action did not expressly give their backing to its claim that it had been the victim of a prior attack.252 But, whatever position is taken on the facts of the outbreak of the Six Day War,253 the point of importance here is that Israel did not rely on anticipatory self-defence to justify its actions.
Again in the 1962 Cuban missile crisis, when Cuba was proposing to import nuclear missiles from the USSR, the USA did not rely on anticipatory self-defence to justify its forcible interception of the missiles on the high seas; rather, it relied on regional peacekeeping under Chapter VIII of the UN Charter.254 And in the Iran/Iraq war Iraq first began its justification for its invasion of Iran in 1980 by relying on preventive self-defence, but quickly shifted its position and claimed to be acting in response to a prior armed attack by Iran; this remained its position and this was the view that it put in response to the Secretary-General’s report on the responsibility for the conflict.255
The USA, in its attempt to justify the shooting down by the USS Vincennes of the civilian Iran Airbus Flight 655 in July 1988 during the Iran/Iraq war, made References(p. 172) elaborate argument that its action had been part of an ongoing battle and that it was engaged in a response to an armed attack by Iran.256 It said that its forces had exercised self-defence under international law by responding to an attack by Iran: Iranian aircraft had fired on a helicopter from the USS Vincennes, then Iranian patrol boats had closed in. In the course of exercising its right to self-defence the USS Vincennes fired at what it believed to be a hostile Iranian military aircraft after sending repeated warnings. It is very striking that the USA did not expressly rely on anticipatory self-defence, even though its rules of engagement had been altered to allow its forces to take action against enemy ships and aircraft displaying ‘hostile intent’.257 The USA maintained the same position in its arguments to the ICJ in the Aerial Incident case arising out of the shooting down of the Iran Airbus.258 Iran argued that Article 51 does not allow pre-emptive self-defence and that the US action amounted to aggression.259 Many other states also took this approach. The UK offered support for the US action in general terms, but did not expressly support anticipatory self-defence.260
The USA and the UK (and France until 1996), in patrolling the ‘no-fly’ zones over Iraq, claimed the right to use force in self-defence. After the establishment of the northern zone in 1991 and the southern zone in 1993 there were many clashes between US and UK aircraft, and Iraqi aircraft and ground defences. Iraq denied the legality of the ‘no-fly’ zones; in this it was supported by Russia and China and other states who said that the USA and the UK acted unilaterally and without Security Council authorization in establishing the zones. In 1999 the USA and the UK significantly extended the rules of engagement for their aircraft; they were now to take pre-emptive action against Iraq’s air defences. Not only the direct source of an attack, such as a missile site, but any threat to aircraft, such as a command centre, could be targeted.261 However, the USA and the UK continued to insist that their actions were purely defensive. Their legal position remained that the military actions were taken to ensure the safety of aircraft patrolling the zone in support of Security Council Resolution 688; once a no-fly zone was authorized in accordance with international law it was entirely appropriate to act in self-defence to ensure the safety of those who were imposing the no-fly zone.262 In Security Council debates on Iraq discussion focused on the References(p. 173) legality of the no-fly zones; Russia, China, and Iraq condemned the US and UK actions on the fundamental basis that there was no legal basis for the no-fly zones and therefore no justification for the presence of US and UK aircraft and no right for them to act in self-defence.263
Very occasionally states have expressly used anticipatory self-defence. As was mentioned above, Israel and South Africa both claimed the right to take ‘pre-emptive action’ against incursions from neighbouring states. These claims were expressly rejected by some states on the ground that anticipatory self-defence was unlawful. Other states used other grounds for condemnation. Therefore authoritative pronouncements on the issue of principle—the legality of anticipatory self-defence—were avoided in these cases by the Security Council and the General Assembly.
Israel, in its 1981 attack on the Iraqi nuclear reactor, claimed anticipatory self-defence. It said that it had acted to remove a nuclear threat to its existence; the Iraqi reactor under construction was designed to produce nuclear bombs whose target would have been Israel. Under no circumstances would Israel allow an enemy to develop weapons of mass destruction against it. In the Security Council debate it relied on a series of writers to support its position that anticipatory defence was lawful. But significantly Israel was not able to rely on any clear state practice to support its position.264 In none of these cases did the Security Council make any pronouncement on doctrine; the debates again revealed the divisions between states on the law in this area. The USA, in allowing the condemnation of Israel by the Security Council, said that its judgment that Israel’s actions violated the Charter was based solely on the conviction that Israel had failed to exhaust peaceful means for the resolution of the dispute. Others said that the action was not justified on the particular facts, given that the International Atomic Energy Agency (IAEA) said that there was no evidence that Iraq was planning to use the reactor for the development of nuclear weapons. Other states rejected anticipatory self-defence in principle.265
This reluctance to rely on anticipatory self-defence even by the USA and Israel was not conclusive that they did not believe that it was legal, as it is natural for states References(p. 174) to choose the strongest grounds to justify their claims, but it was strong evidence of the controversial status of this justification for the use of force, as was the deliberate avoidance of the issue of the legality of anticipatory self-defence by the ICJ in the Nicaragua case and more recently in Armed Activities on the Territory of the Congo (DRC v Uganda). In DRC v Uganda the Court recalled that Uganda had insisted that Operation Safe Haven was not a use of force against an anticipated attack. The Court therefore expressed no view on that issue.266 However, some of its subsequent reasoning seems to indicate a narrow view of self-defence which would exclude pre-emptive action. Thus, the Court observed that the official Ugandan document on the conflict made no reference whatever to armed attacks which had already occurred against Uganda. Rather the position of the Ugandan High Command was that it was necessary to ‘secure Uganda’s legitimate security interests’. The specified security interests were essentially preventative—to ensure that the political vacuum in the DRC did not adversely affect Uganda, to prevent attacks from genocidal elements, to be in a position to safeguard Uganda from irresponsible threats of invasion and to deny the Sudan the opportunity to use the territory of the DRC to destabilize Uganda.267 The Court said that Article 51 may justify a use of force in self-defence only within the strict confines there laid down. It does not allow the use of force by a state to protect perceived security interests beyond these parameters. Other means are available to a concerned state, including, in particular, recourse to the Security Council.268
The clear trend in state practice before the terrorist attacks of 9/11 was for states to try to bring their use of force within Article 51 and to claim the existence of an armed attack rather than to argue expressly for a wider right under customary international law. This practice, the cautious approach of the ICJ and the clear and long-standing divisions between states on this issue make it all the more surprising that the High-level Panel set up by the UN Secretary-General269 proclaimed in its December 2004 Report that ‘Long-established customary international law makes it clear that States can take military action as long as the threatened attack is imminent, no other means would deflect it, and the action is proportionate.’270 As we have seen, this is actually an extremely controversial assertion. Nevertheless, the Secretary-General in his own subsequent report of March 2005, In Larger Freedom, asserted that imminent threats are fully covered by Article 51: ‘lawyers have long accepted that this covers an imminent attack as well as one that has already happened’.271 The Secretary-General’s statement References(p. 175) is marginally less controversial in that he refers to ‘lawyers’ rather than ‘customary international law’ as accepting the doctrine of anticipatory self-defence. He also tries to argue that self-defence against an imminent attack comes within the words of Article 51, rather than accepting the High-level Panel’s more controversial position that a wider customary international law right survives. These statements did not attract wide support from states: the Non-Aligned Movement (now with 120 member states) rejected this position,272 and has continued to argue for a restrictive right. The 2005 World Summit for which the two reports had been prepared not surprisingly avoided the issue in its Outcome Document.273 It is clear that states remain fundamentally divided on this question. Many writers assert that the dominant view is that anticipatory self-defence is lawful, but those who study state practice are more cautious.274
The picture that emerges is one of polarization. Before 9/11 few states claimed wide rights of self-defence to protect nationals, anticipate attack, or to respond to terrorist and other past attacks.275 It seems that the lesson they learned from the judgment in the Nicaragua case was that form is more important than substance. As long as they pay lip-service to the need to act in self-defence, and as long as they report to the Security Council invoking the magical reference to Article 51, somehow their action acquires a veneer of legality and their argument will be treated seriously by commentators. A few of these commentators seem prepared to treat any US action as a precedent creating new legal justification for the use of force.276 Thus they use the US actions as shifting the Charter paradigm and extending the right of self-defence. The lack of effective action against the USA as a sanction confirms them in this view. But the vast majority of other states remained firmly attached to a narrow conception of self-defence. This long-standing disagreement between states on interpretation of the UN Charter seemed beyond resolution, and states accordingly sought to avoid doctrinal dispute by appealing where at all possible to doctrines, such as necessity and proportionality, on which there was universal agreement.
The rise of ISIS in Iraq and Syria, its rapid advances and seizure of territory, and its cross-border attacks from Syria into Iraq, led the government of Iraq to seek military assistance from the USA and other states. Their response has brought the right of collective self-defence back into prominence, and has raised fundamental questions about the scope of the right. Before these developments, there had been comparatively little practice on the use of force in collective self-defence in the UN era; states generally avoided direct and open military participation by their armed forces in conflicts between other states. The relatively large number of treaties on collective self-defence was not matched by extensive state practice.277 Commentators listed the following instances where states actually invoked collective self-defence: USA and Lebanon (1958), UK and Jordan (1958), UK and South Arabian Federation (1964), USA and Vietnam (1961–75), USSR and Hungary (1956),278 Czechoslovakia (1968), Afghanistan (1979), France and Chad (1983–4, 1986), USA and others, and Kuwait (1990). The right was invoked by the government of the DRC in support of its request to Angola, Namibia and Zimbabwe for military assistance against Rwanda and Uganda in 1998.279 It was also invoked to justify Operation Enduring Freedom in Afghanistan.280 But this cannot be taken as a definitive list; different commentators produce different lists.281 Controversially, some add the UN-authorized References(p. 177) actions in Korea (1950–3) and Iraq (1990) as further examples of collective self-defence.282
The above list includes some episodes where collective self-defence was invoked and foreign troops were introduced into the ‘victim’ state requesting assistance, but force was not used in actual conflict, or was not used beyond the national border of the victim state. States have invoked collective self-defence as a justification for inviting in foreign troops before any armed attack has occurred, in case collective self-defence is needed in the future; that is, as a deterrent or as a precaution. The sending of troops and the provision of other aid has been much more common than the use of those troops in actual fighting against an attacking state. The US use of force against North Vietnam, Cambodia, and Laos in the name of the collective self-defence of South Vietnam and the US use of force against Nicaragua (in the name of collective self-defence of El Salvador, Costa Rica, and Honduras) are exceptional in that the USA used force outside the ‘victim’ state. Operation Enduring Freedom in Afghanistan in response to the 9/11 terrorist attacks on the USA is another, more recent example, and in 2014 the USA (and others) again invoked collective self-defence of Iraq to justify the use of force against ISIS in Syria.283
That is, although in theory there is a distinction between collective self-defence and assistance in reply to an invitation by a government to respond to external intervention against that government, in practice the line may not be a clear one.284 The states sending in their troops make choices as to the justification they offer. They may invoke collective self-defence before it is actually necessary and conversely they do not always expressly invoke collective self-defence even when a case could be made for it on the basis that there has been or might be an armed attack.
References(p. 178) Thus, for example, Ethiopia, the USSR, and Cuba all tended to play down the presence of Soviet and Cuban troops in Ethiopia (1977–8) even though there had been an armed attack by Somalia into the Ogaden region of Ethiopia after the overthrow of Emperor Haile Selassie, and the armed response by Ethiopia was limited to driving out the invading forces.285 Also with regard to the collective self-defence of Angola against attacks by South Africa, Cuba at first simply stressed that it had been invited in by the Movimento Popular de Libertação de Angola (MPLA), which subsequently formed the government on the coming to independence of Angola in 1975. Angola referred to Article 51 in relation to the presence of the Cuban troops only from 1983, after the issue of ‘linkage’ had become more prominent. On the basis of ‘linkage’, South Africa argued that its withdrawal from Namibia was linked to that of Cuba from Angola, thus implying an equivalence between the two situations. Angola replied that the Cuban presence had been requested by the legitimate government of Angola for the clear and express objective of repulsing the open and flagrant invasion by South Africa. The first invasion (in 1975) was repulsed by the Angolan people with the assistance of Cuban troops, but South African aggression had continued. There was a continued need for the assistance of Cuban forces in full conformity with Article 51, as every state has the right to individual or collective self-defence.286 These episodes pre-date the judgment in the Nicaragua case.287
After the judgment the USA occasionally invoked collective self-defence against Nicaragua in Central America again.288 And subsequently claims to collective self-defence to justify the use of force in defence of Kuwait,289 Tajikistan,290 and the DRC291 revealed the complexity of collective self-defence. The legality of the third-state use of force was controversial in almost all these cases, both those References(p. 179) before and those after the Nicaragua case, but the disagreements between states on the legality of these uses of force have generally centred on the facts rather than the law. In almost all these cases the controversy concerned the question whether there had been an armed attack, and also whether there had been a genuine request for help by the victim state. In contrast, Operation Enduring Freedom in Afghanistan raised fundamental legal questions about the concept of ‘armed attack’.292 The question of the existence of an armed attack provoked debate more recently in the context of the invocation of collective self-defence in conflicts in Iraq and Syria.293
On the whole, however, the states directly involved and those responding to their use of force through international organizations did not disagree as to the content of the applicable law. This may seem surprising, given that the theory of collective self-defence has been controversial since the debate over its express inclusion in the UN Charter. Collective self-defence was included at the instance of the Latin American states to make clear the compatibility of the existing American system and the new UN system. After prolonged debate, collective self-defence was included in Chapter VII on the powers of the Security Council rather than in Chapter VIII on regional arrangements.294 There has subsequently been controversy as to whether collective self-defence was a new concept when it was included in the Charter in 1945.295 Some of the judges in the Nicaragua case took the view that it was an innovation. Thus, for example, Judge Oda said that the term ‘collective self-defence’ was unknown before 1945 and therefore expressed doubt as to whether it was an inherent right.296 Judge Jennings agreed that it was a novel concept.297 Whether or not collective self-defence was a totally new concept, the post-1945 practice has been crucial in its crystallization.
Early debates on Article 51 of the UN Charter focused on whether collective self-defence was an autonomous right allowing any third state to use force in defence of the victim of an armed attack, or whether it was a collection of rights to individual self-defence only to be exercised if the third state was itself a victim or if References(p. 180) the interests of the third state were somehow engaged.298 What were the conditions for its exercise: did it require a pre-existing treaty arrangement for collective action? Some argued that there was a need of prior agreement for collective self-defence, otherwise the use of force would be contrary to the spirit of Article 51;299 other writers insisted on a common interest rather than a pre-existing treaty.300
The Nicaragua Case
The ICJ decision in the Nicaragua case on the legality of the US use of force and intervention in Nicaragua renewed the passion of the debate on the scope of collective self-defence. The judgment has been much attacked and much misinterpreted. It plays a crucial role in this area. The Court’s decision, its first extended discussion of the law on the use of force, was based on customary international law because of the US reservation to its Optional Clause acceptance. The Court found that the US multilateral treaty reservation prevented it from applying the UN Charter and other multilateral treaties, such as the (Organization of American States) OAS Charter and the Rio Treaty, which in fact bound the parties. However, the reservation did not stop the Court from deciding the case on the basis of customary international law, which continued to exist alongside treaty law.301 Moreover, the Court could properly adjudicate because the provisions of multilateral treaties did not diverge from customary international law to such an extent that a judgment of the Court on custom would be a wholly pointless exercise. The Court went on to say that, although it had no jurisdiction to determine whether the conduct of the USA constituted a breach of the Charter of the UN and that of the OAS, it could and must take them into account in ascertaining the content of customary international law.302
References(p. 181) The Court’s exposition of the law on collective self-defence, the justification used by the USA to support its use of force and intervention in and against Nicaragua, was relatively brief. The parties, in view of the circumstances in which the dispute had arisen, had relied only on the right of self-defence in the case of an armed attack which had already occurred; the lawfulness of a response to the imminent threat of armed attack was not raised.303 Also the parties agreed that any exercise of self-defence must be necessary and proportionate. The Court accordingly went on to define the other specific conditions which had to be met for the exercise of collective self-defence.304
First, the Court considered what constituted an armed attack: the sending of armed bands rather than regular army could constitute an armed attack, provided that the scale and effects of the operation were such as to be classified as an armed attack and not a mere frontier incident.305 Second, ‘it is also clear that it is the State which is the victim of the armed attack which must form and declare the view that it has been so attacked. There is no rule in customary international law permitting another state to exercise the right of collective self-defence on the basis of its own assessment of the situation. Where collective self-defence is invoked, it is to be expected that the State for whose benefit this right is used will have declared itself to be the victim of an armed attack’.306 Third, the Court held that ‘there is no rule permitting the exercise of collective self-defence in the absence of a request by the State which regards itself as the victim of an armed attack’.307 The Court also held that the requirement in Article 51 of the UN Charter that the state claiming to use the right of individual or collective self-defence must report to the Security Council was not a customary law requirement, although ‘the absence of a report may be one of the factors indicating whether the state in question was itself convinced that it was acting in self-defence’.308
References(p. 182) The Court was criticized for its treatment of collective self-defence in the separate and dissenting opinions on contrasting grounds. Judge Ruda said that the Court should not have gone into the topic at all, given that it had held that there was no armed attack.309 Judge Oda said that if it was going to consider collective self-defence, it had been far too brief on this controversial topic.310
The judgment on the merits in Nicaragua attracted strong criticism, especially from US writers.311 They were unhappy at the brevity of the Court’s reasoning on collective self-defence and at its approach to customary international law. Or it may be more accurate to say that, because some of the writers were unhappy with the substantive conclusions of the Court that the USA had illegally used force and intervened in Nicaragua, they therefore attacked its legal reasoning. How far were the Court’s conclusions on collective self-defence justified on the basis of customary international law and compatible with treaty law? To what extent were they based on sound policy considerations?
State practice on collective self-defence
A central issue in all the episodes where collective self-defence was expressly invoked by states was whether there had been an armed attack such as to justify the third-state assistance to the victim state.312 During the US intervention in Lebanon in the name of collective self-defence in 1958 there was initially some uncertainty on this issue.313 The USA and Lebanon did not at first mention armed attack, although they both reported to the Security Council that the US intervention was in response to a request by Lebanon under Article 51 of the UN Charter. But subsequently Lebanon expressly argued that there was an armed attack by the United Arab Republic (Egypt and Syria) and that there was no difference between a regular army and irregular forces for the purposes of Article 51.314 China also took this position and no state challenged it.315 The reason why the claim to collective self-defence was controversial in Security Council and General Assembly debates was that states were sceptical as to whether there (p. 183) had in fact been any armed attack, whether by regular or irregular troops; they claimed that the USA was simply trying to protect an unpopular leader from internal unrest at a time of growing Arab nationalism and republicanism.
Again, in the case of Vietnam, the USA argued that the infiltration from North Vietnam amounted to an armed attack justifying collective self-defence of South Vietnam. It famously asserted that from 1959 until 1964 the North infiltrated over 40,000 men into the South. It said that in these circumstances armed attack was not as easily fixed by date and hour as in the case of traditional warfare, but the infiltration of thousands of men clearly constituted an armed attack under any reasonable definition.316 States did not deny that the actions of irregular troops could be attributed to a state, but they doubted whether in fact there was an invasion of one state by armed bands from another rather than an uprising throughout Vietnam.317 Similarly, with regard to its intervention in Afghanistan in 1979 the USSR claimed collective self-defence against an at first unspecified ‘foreign intervention’; it was not controversial that the actions of armed bands could constitute an armed attack, but there was doubt as to the existence of such an attack.318
Later, in Tajikistan there was controversy as to the existence of armed attacks from Afghanistan against Tajikistan.319 After Tajikistan attained independence in 1991 civil war broke out and opposition forces operated against the government from Afghanistan. Russia argued that it was justified in using force in collective self-defence of Tajikistan against these incursions. In 1993–5 Russia and Tajikistan repeatedly accused Afghanistan of involvement in the attacks; Afghanistan denied these claims. This continued even after the conclusion of a border agreement between Afghanistan and Tajikistan, an Agreement on a Temporary Cease-fire and the Cessation of Other Hostile Acts on the Tajikistan/Afghanistan Border and within the Country between the warring parties in Tajikistan, and the creation of a UN observer force (UNMOT) to monitor the border/ceasefire. The UN Secretary-General made various reports on the situation, but did not come to any public conclusions as to the occurrence of armed attacks and the right of Tajikistan and Russia to act in collective self-defence against Afghanistan.
In state practice the supply of arms, money, and logistic support have not generally been treated as armed attacks in the context of collective self-defence. (p. 184) Occasionally there have been hints of such a position. For example, in Lebanon in 1958 the USA and Lebanon at first simply said that the infiltration of armed men, arms, and supplies from Syria in the UAR threatened the independence of Lebanon and that this gave the right of collective self-defence. They did not expressly mention armed attack at this stage; they invoked Article 51 as a precaution, saying that their forces were not there to engage in hostilities.320 In the Security Council debates some states (the UK, France, Canada, and China) supported the right of Lebanon to request and the USA to send troops, but on the facts this cannot be interpreted as amounting to an endorsement of the actual use of force in the exercise of collective self-defence. It can be seen as simply an endorsement of the right of the USA to send in its troops to help the government of Lebanon. Certain states such as the UAR, the USSR, and Sweden, denied the existence of an armed attack and said that because of this the USA had no right to use force against the UAR.321 Lebanon, in later defending its position against the criticisms that had been made by other states in the Security Council, expressly justified its invitation to the USA on the basis that there had been an armed attack. It affirmed a conception of armed attack that did not include the mere supply of arms, but only armed attack by regular army and irregular troops.322
The US justification of its intervention in Vietnam also at first seemed to be based on a wide view of armed attack. In its 1966 Department of State Memorandum on The Legality of US Participation in the Defense of Vietnam it said in the first paragraph that it was assisting South Vietnam to defend itself against armed attack from the North and that ‘this armed attack took the form of externally supported subversion, the clandestine supply of arms and the infiltration of armed personnel’.323 But the USA subsequently focused on the movement of troops across the border between North and South Vietnam; in its reports to the Security Council the USA, in claiming to act in collective self-defence, generally spoke only of the use of force by regular and irregular troops.324 The lesson that emerges from this practice is that the supply of arms alone does not constitute an armed attack.325
References(p. 185) The requirement of a request
As mentioned above, the Court’s findings in the Nicaragua case on the nature of the incursions from Nicaragua into Costa Rica and Honduras were somewhat inconclusive as regards the question whether these were frontier incidents not amounting to an armed attack. After the Court had looked at these incursions and at the supply of arms to the opposition in El Salvador it continued, ‘There are however other considerations which justify the court in finding that neither these incursions nor the alleged supply of arms to the opposition in El Salvador may be relied on as justifying the exercise of the right of collective self-defence.’326 The Court went on to apply the conditions that it had identified as limiting the right to collective self-defence in its earlier discussion of customary international law. It found that there had been no timely declaration by El Salvador that it was the victim of an attack and no declaration at all by Honduras and Costa Rica. Also none of the three had made any request for help to the USA before its forcible intervention. These factors together all showed that the USA was not acting in self-defence of the three states.327
The only authority the Court mentioned for its requirement of a request by the victim state was the Rio Treaty, Article 3(2), which says that measures of collective self-defence are decided ‘on the request of the state or states directly attacked’.328 As regards its requirement of a declaration by the victim state that it had been the victim of an armed attack, the Court offered no authority. That is, it offered almost no justification for the conditions it apparently imposed on collective self-defence. One obvious inference is that the Court was influenced by the fact that the parties were actually bound by a treaty commitment that the victim state request assistance.329 Moreover, the approach adopted by the Court seems correct in principle, as any other approach would allow the third state to pronounce on the existence of an armed attack and to decide that it was going to use force even against the wishes of the victim state.330
Judges Jennings and Schwebel, however, attacked the Court’s reasoning and conclusion on these points. Their arguments seem to be based on policy considerations. Both were critical of the Court’s ‘formalistic’ model of collective self-defence. Schwebel’s concern was with covert action; he asked, ‘Where is it written that a victim state may not informally and quietly seek foreign assistance?’331 References(p. 186) Jennings showed some sympathy with the Court’s desire to limit the scope of collective self-defence:
Obviously the notion of collective self-defence is open to abuse and it is necessary to ensure that it is not employable as a mere cover for aggression disguised as protection, and the Court is therefore right to define it somewhat strictly. Even so, it may be doubted whether it is helpful to suggest that the attacked state must in some more or less formal way have ‘declared’ itself the victim of an attack and then have as an additional ‘requirement’ made a formal request to a particular third state for assistance.
Jennings’ argument is apparently based on policy. He goes on, ‘It may readily be agreed that the victim state must both be in real need of assistance and must want it and that the fulfilment of both these conditions must be shown. But to ask that these requirements take the form of some sort of formal declaration and request might sometimes be unrealistic.’332
Some writers have been similarly critical of the Court’s reasoning on the grounds of its formalism.333 Simma’s commentary on the UN Charter, states categorically, but without any attempt at justification beyond references to secondary sources, that the Court was wrong to require an express request by the victim state.334 Macdonald agrees that the requirement of a request for help was a ‘wholly new and unconsidered limitation on the right to collective self-defence’.335 These pronouncements by writers appear misguided as a matter of principle and in the light of the mass of state practice.
Moreover, in the Oil Platforms case the Court reasserted its position. It said ‘Despite having referred to attacks on vessels and aircraft of other nationalities, the United States has not claimed to have been exercising collective self-defence on behalf of neutral states engaged in shipping in the Persian Gulf; this would have required the existence of a request made to the United States by the State which regards itself as the victim of an armed attack.’336 That is, although it was not actually called on to pronounce on collective self-defence, the Court treated the Court’s position in the Nicaragua case as authoritative, and again made it clear that the existence of a request by the victim state was a necessary element of self-defence.
References(p. 187) State practice also generally supports the Court’s position that normally a request and a declaration would be made. Some collective self-defence treaties, like the Rio Treaty, expressly require a request by the victim state.337 In contrast Article 5 of the NATO Treaty does not include such a requirement.338 And recently Japan has taken the position that collective self-defence does not necessarily require a request by the victim state as long as there has been an armed attack. But in every case where a third state has invoked collective self-defence it has based its claim on the request of the victim state, even where there was no express treaty provision requiring this.339 Also the state claiming to be the victim has generally asserted that it has been the victim of an armed attack. But in almost all the cases of collective self-defence listed above there has been controversy over the existence or the genuineness of the request. With regard to the interventions in Lebanon and Jordan in 1958, the USSR argued that the USA put pressure on Lebanon to issue an invitation,340 and that the UK intervention was planned before the request from Jordan and that the request was not free. The true motive of the USA and the UK was to repress the rise of Arab nationalism.341 With regard to Vietnam, those who challenged the legality of the US intervention said that South Vietnam was not a separate state and had no right to seek outside assistance.342 In the Soviet invasions of Czechoslovakia and Afghanistan the invitations came from governments installed by the invading state.343 In Chad there References(p. 188) was an ongoing civil war and the legitimacy of the government and its right to request outside help was not always clear. But Libya relied on a request in 1980 and said that it had left when so requested.344 France also said that a request was necessary and it responded to a request.345 In 2001 the UK and Australia, acting with the USA in Operation Enduring Freedom, did not expressly refer to a request by the USA in their letters to the Security Council, but it was clear that there had been such a request. More recently, almost all the states (apart from Germany and Turkey) coming to the assistance of Iraq against ISIS in Syria expressly stated that they did so at the request of Iraq, or referred to the request by Iraq to the USA to lead international efforts.346
Third state interest?
After his criticism of the Court’s requirement of a declaration and a request by the victim state, Judge Jennings went on to say that the reasoning was also objectionable in that the Court was giving the impression that the third state need not itself have an interest for it to exercise collective self-defence.347 A few others follow the Jennings approach.348 Some have even argued that the right to collective self-defence is essentially the right of the party giving aid to the victim, and that the ICJ itself should not be taken to have rejected this position.349 But this insistence on a third-party interest all seems rather far-fetched in the light of state practice since 1945. States themselves have not used this argument; criticisms by states of the legality of actions taken in the name of collective self-defence have not mentioned the absence of a third-state interest or of a treaty commitment as a ground of illegality. In many of the episodes the intervening state did in fact have a pre-existing treaty relationship with the ‘victim’ state,350 but in the other cases References(p. 189) where there was no such treaty this was not mentioned as a ground of illegality even by those otherwise critical of the use of force.351
The duty to report to the Security Council under Article 51
Also the failure of the USA to report on its use of force to the Security Council under Article 51 was taken by the Court as an indication that the USA was not exercising the right of collective self-defence.352 Judge Schwebel criticized this, but the Court’s position is an accurate reflection of earlier practice on collective self-defence. The USA itself, with regard to its intervention in Vietnam, pointed out that it was not bound under Article 51 to report because neither North nor South Vietnam were members of the UN. Nevertheless, it said it would report because Article 51 was an appropriate guide.353 The other states claiming to use collective self-defence in other episodes also reported under Article 51. Indeed, there was a contrast here with individual self-defence, with regard to which states’ reporting was much more erratic and which improved only after the Nicaragua case. In collective self-defence all the states expressly invoking collective self-defence duly reported.354 And after the decision on the merits of the Nicaragua case the Central American states and the USA referred their subsequent claims to collective self-defence to the Security Council.355 Moreover, several collective self-defence treaties specifically require parties to report to the Security Council.356 It may not be mandatory to report in the sense that failure to report will not in itself References(p. 190) mean that the action cannot be self-defence, but failure will be evidence that the action was not in fact self-defence. As with the UK’s controversial intervention to protect the South Arabian Federation (before it was a member of the UN), the USSR and the USA both said that the UK should have gone to the Security Council earlier if its action had been justified as self-defence.357 The UK itself repeated this argument against the USSR and its failure to turn earlier to the Security Council over Afghanistan.358
Iraq and the Complexities of Collective Self-Defence
Several states relied on collective self-defence to justify their use of force in the large-scale conflicts in Iraq and Syria since 2014. These claims were diverse and their interrelationship was not always clear. The dramatic rise of ISIS in Iraq in 2014 led the government of Iraq to seek foreign assistance.359 The USA began air strikes in August 2014 in response to this request.360 Initially US action was limited to operations inside Iraq and so the legality of the air strikes was not contested. The USA gradually increased the scope of its military action in Iraq after a change of government; the new al-Abadi government pledged itself to abandon the sectarian policies of the al-Maliki government (2006–14) whose exclusion of the Sunnis from positions of power had contributed to the rise of ISIS and to its success in seizing control of Sunni-occupied territory in Iraq. At first the USA claimed to be using force only for the protection of its own troops, but it went on to humanitarian efforts to save minority groups of Yazidis and Christians under attack by ISIS, and to more far-reaching action to roll back the ISIS threat. The USA then announced on 10 September 2014 that it would take military action in Syria as well as Iraq, in order to ‘degrade and destroy’ ISIS which had taken advantage of the civil war in Syria to spread beyond Iraq: it seized Syrian oil fields and also gained control of border areas that allowed it to operate across the border into Iraq.361 The USA would lead a broad coalition of forty (later sixty) states. It would not send in ground troops.362
On 20 September 2014 Iraq reaffirmed its earlier request and made a more far-reaching request for military assistance.363 It did not explicitly invoke collective (p. 191) self-defence against ISIS in Syria, but the substance of the letter was clear. It welcomed military assistance in coordination with Iraqi armed forces and in accordance with international law, respecting Iraq’s sovereignty. It did not refer to Article 51; it did not use the language of armed attack or even of imminent attack. It requested the USA to lead international efforts to strike ISIS sites ‘with our express consent’. The aim was to end the constant threat to Iraq, protect Iraq’s citizens, and ultimately arm Iraqi forces to regain control of Iraq’s borders. The letter to the Security Council made no express mention of use of force in Syria; it simply said that ISIS had established ‘a safe haven outside Iraq’s borders’ that was a direct threat to Iraq’s security. ISIS had secured for itself the ability to carry out terrorist operations across Iraq’s borders. It had made those borders impossible to defend.
The controversial question of the scope of self-defence against non-state actors, the legality of action against ISIS in Syria, will be discussed in Chapter 5. Clearly a wide right of individual self-defence can bring with it a wide right of collective self-defence. Here the focus will be on the issues particular to collective self-defence. There is one fundamental question: who decides on the scope of the action to be taken? In theory the answer is simple: this should be Iraq, the state inviting assistance. It seems to follow from the framework set out in the Nicaragua case that it is for the victim state to ask for help, and therefore to establish the limits of the action to be taken by the third state, and not for the intervening state to determine what is required. It is interesting to note Iraq’s repeated stress on its sovereignty in its letter to the Security Council. It said: ‘Although Iraq is in great need of the assistance of its friends in combatting this evil terrorism, it nonetheless attaches great importance to preserving its sovereignty and its ability to take decisions independently, both of which must be honoured in all circumstances.’ Iraq ‘requested the USA to lead international efforts to strike ISIL sites and military strongholds with our express consent’.
However, in practice, implementation of this principle may not be straightforward when a powerful state is coming to the assistance of a weaker state; the exercise of collective self-defence in Syria proved complex because of the very different interests of the states involved. Some of those states did not rely only on the right of collective self-defence of Iraq. They also claimed a separate right of individual self-defence against ISIS and other groups in Syria.364 There is a considerable diversity among their letters to the Security Council. The first states to respond to Iraq’s request by carrying out air strikes in Syria were the USA and four Gulf States.365 The Gulf States did not write to the Security Council to report that References(p. 192) they were acting in self-defence. The USA in its letter said that there was a serious threat of continuing attacks from ISIS coming out of safe havens in Syria.366 ISIS and other terrorist groups in Syria were a threat not only to Iraq but also to many other countries, including the USA and its partners in the region and beyond. The USA referred to both collective and individual self-defence ‘as reflected in Article 51’ in general terms, saying that states must be able to defend themselves. The USA also reported that it had initiated military actions in Syria against Al Qaida elements ‘to address terrorist threats that they pose to the US and our partners and allies’. This right of individual self-defence—based on a very wide doctrine of self-defence—gave the USA the freedom to choose targets other than those necessary for the collective self-defence of Iraq.367
The extension of military action beyond Iraq into Syria was controversial, and other states that supported military action in Iraq did not immediately join the USA in air strikes on Syrian territory.368 Syria’s position was set out in many letters to the Security Council: one state was inviting others to use force on the territory of another; states were intervening against terrorist organizations in Syria at the request of a government other than that of Syria.369 It said that ‘Any attempt to invoke Article 51 of the Charter to justify military action on Syrian territory without coordination with the Syrian government manipulates, distorts, and misinterprets the provisions of that Article.’370
A year later in September 2015 Canada, Australia, and France also began military action in Syria. Canada’s letter to the Security Council reported that it was acting in collective self-defence of Iraq, but also asserted that ISIS was a threat to Canada and invoked individual self-defence.371 In contrast, Australia invoked only collective self-defence ‘as part of international efforts led by the USA’.372 France was not specific. It said briefly that ‘In accordance with Article 51 … France has taken actions involving the participation of military aircraft in response to attacks carried out by ISIL from the territory of the Syrian Arab Republic.’373 After the Paris (p. 193) nightclub attacks—attributed to ISIS—of November 2015, France expressly said that earlier it had been acting in collective self-defence of Iraq, but now it would act in individual self-defence. It now itself requested assistance from other EU member states in collective self-defence.374 Although the attacks in Paris had been carried out by French and Belgian nationals, France asserted that ‘the acts of war’ were decided upon, planned, and prepared in Syria.375
Like the USA after 9/11, France did not act through NATO under the provision for collective self-defence in Article 5 of the NATO Treaty.376 Instead it invoked the mutual assistance clause, Article 42(7), of the Treaty of European Union which provides: ‘If a Member State is the victim of an armed aggression on its territory, the other Member States shall have towards it an obligation of aid and assistance in accordance with Article 51 of the UN Charter.’ This was the first time that this provision was invoked. It provides for bilateral cooperation rather than any involvement of the institutions of the EU. More states now joined in the air strikes on Syria. Again their letters to the Security Council are not all entirely clear as to the basis of the military action. The UK’s letter mentions both individual and collective self-defence.377 But it does not make clear whether this is collective self-defence of France or of Iraq or both.
This distinction is important because it affects the type of military action to be taken. The aim of action in collective self-defence of Iraq would be to drive ISIS out of Iraq, stop ongoing attacks from Syria, and to re-establish government control. It would be for the government of Iraq to decide what is required, and when such operations should end. As ISIS loses territorial control and the ability to carry out cross-border attacks into Iraq from Syria, the question arises as to how long the collective self-defence of Iraq can continue. If the action is taken in collective self-defence of France, the aim would be to respond to attacks on France. The choice of targets, and the identity of the decision-maker, could differ according to the legal basis for the action. Germany’s letter was also less than clear.378 It said that ISIS has carried out armed attacks against Iraq, France, and ‘other states’. Germany was exercising the right of collective self-defence and would support the military measures of those states that have been subject to attacks by ISIS. In contrast, Denmark reported that it was acting in collective self-defence as part of the international efforts led by the USA.379 The Netherlands reported that References(p. 194) it was acting in collective self-defence of Iraq.380 Belgium subsequently reported that its actions were in collective self-defence of Iraq, but it also said that it would support the military measures of those states that had been subjected to attacks by ISIS.381 The scope of these claims to collective self-defence was not clear: were they intended to extend to military assistance to Turkey?
Turkey’s involvement in Syria further complicated the situation with regard to collective self-defence. Its policy interests in the region differed from, or even conflicted with, those of other states acting in collective self-defence of Iraq. Thus Turkey was hostile to the Kurds in Syria whom it perceived as linked to the PKK terrorist organization in Turkey. In contrast, the USA regarded the Syrian Kurds as effective allies in the struggle against ISIS.382 Also Turkey had intervened in the conflict in Syria since 2011; it was for several years a major conduit into Syria for weapons and opposition fighters, including those with links to ISIS and to Al Qaida, to fight against the government of President Assad. There were border clashes between the two territories. In June 2012 Syria shot down a Turkish plane over its territory.383 NATO met at Turkey’s request, and issued a condemnation of Syria’s action. However, it did so under Article 4 of the NATO treaty (which provides only for consultation in the case of threats), rather than Article 5 on collective self-defence.384 It was the July 2015 ISIS terrorist attack on Suruç (in Turkish territory) that led Turkey to join openly in military actions in Syria, and to end its 2013 ceasefire with the PKK. It now invoked individual and collective self-defence as the basis for its military actions against ISIS in Syria, ‘including in coordination with individual members of the Global Coalition’.385
As regards Turkey’s claim to individual self-defence, Syria rejected this, saying that Article 51 entitled Syria to defend its people and land against terrorists who were coming over its borders with the support of the Turkish regime. However, Turkey had no right to invoke Article 51, because it was itself the party exporting the terrorism to which it referred.386 As regards Turkey’s general claim to collective self-defence, some serious questions arose. Unlike the other states writing to the Security Council under Article 51, Turkey did not refer to any request by Iraq. Syria asked whether Turkey genuinely intended to fight ISIS and other terrorist organizations associated with Al Qaida or was it merely making that claim References(p. 195) in order to strike the Kurds in Syria and Iraq.387 This suspicion about Turkey’s motives was borne out by subsequent events. Initially Turkey undertook only very limited military operations against ISIS in Syria; at the same time as it began these operations it once again sent air forces (and later ground forces) into Iraq in major operations against the PKK for the first time since the 2013 ceasefire.388 Turkey claimed that the PKK and the Syrian Kurds (PYD) were one and the same, even though the Syrian Kurds were fighting against ISIS in Syria.389 Iraq repeatedly protested about these Turkish actions against the PKK on its territory as an assault on its sovereignty. It invoked Article 51 against Turkey and called on the Security Council to order Turkey to withdraw its forces.390 The Arab League and the USA called on Turkey to withdraw, but it did not do so.391 In February 2016 Turkey also attacked Syrian Kurds across the Syrian border.392 It seems that Turkey had invoked collective self-defence (apparently of Iraq) mainly as a pretext to justify its own actions against the Kurds. In August 2016 it began Operation Euphrates Shield in Syria, ostensibly against ISIS, but with the apparent aim of establishing a buffer zone in Syria, to prevent the Kurds from controlling this area.393
A further—and dangerous—complication with regard to collective self-defence arose out of Turkey’s membership of NATO and its bad relations with Russia from 2011 to mid-2016. As was discussed in Chapter 3, Turkey was supporting the armed opposition seeking to overthrow President Assad, whereas Russia provided support to Assad. In September 2015 Russia began air strikes in Syria to halt opposition advances.394 Turkey complained about Russian overflights and in November 2015 it shot down a Russian fighter jet which it said was passing briefly over Turkish territory.395 This was the first direct clash between a NATO (p. 196) member state and Russia since Soviet participation in the 1950–3 Korean War. Turkey called a meeting of NATO and there was concern that it might draw NATO into a conflict in pursuit of its own interests, but NATO was divided. It did not expressly invoke Article 5 on collective self-defence and merely called for diplomacy and de-escalation.396
Writers on collective self-defence are clearly split into two camps, and their reactions to the Nicaragua case reflect these different viewpoints. First, some view collective self-defence as a valuable means to help protect weak victim states from oppression.397 They therefore attack the ICJ’s limitations on collective self-defence. For them the Court’s view of armed attack is too narrow, and the requirements of a declaration and request by the victim state are unduly formalistic and restrictive. They argue that the Court’s approach will encourage aggression of a low-key kind.398 This sort of enthusiasm for collective self-defence was also apparent in earlier writers on the Charter; they were clearly writing under the influence of the Second World War and saw the provision for collective self-defence in Article 51 as a useful means to protect small states. For example, McDougal and Feliciano said that defence must be collective if it is not to be an exercise in individual suicide.399
The opposing camp comprises those writers who have taken a much more suspicious approach to collective self-defence. They see it rather as a threat to world peace. Thus they argue that there is a need for a high threshold of armed attack and distinction between armed attack and lesser use of force in order to reduce the involvement of major powers. Otherwise there would be a risk of the internationalization of civil conflicts and the expansion of inter-state conflicts.400 They References(p. 197) also said that there is a danger that Article 51 on collective self-defence would help remote, undemocratic states.
In this regard it is interesting that the Court itself in Nicaragua, although concerned to limit the right of collective self-defence, expressly ruled out any consideration of the motives of states engaged in collective self-defence. Thus it declined to undertake an examination of any additional motive beyond the protection of El Salvador, Costa Rica, and Honduras that the USA might have in using force against Nicaragua.401 The USA asserted that it had responded to requests for assistance from El Salvador, Honduras, and Costa Rica in their self-defence against aggression by Nicaragua. Nicaragua claimed that the references made by the USA to the justification of self-defence were merely pretexts for its activities. The true motive was to impose its will on Nicaragua and force it to comply with US demands. However, the Court said that if the USA could establish that Nicaragua had supported the opposition in El Salvador and that this support amounted to an armed attack and the other appropriate conditions for collective self-defence were met, then it could legally invoke collective self-defence. The possibility of an additional motive, even one perhaps more decisive for the USA, could not deprive the USA of its right to resort to collective self-defence. The only significance of the alleged additional motive was that special caution was called for in considering the allegations of the USA concerning conduct by Nicaragua which might provide a sufficient basis for self-defence. This provides a marked contrast to the policy-oriented approach of McDougal and Feliciano. In their discussion of collective self-defence they say, ‘A first step in the determination of reasonableness (that is lawfulness) is thus an inquiry into the substantiality of the collective “self” alleged for security and defence, and into whether a purported grouping for common protection is in reality a facade for other unlawfully expansive purposes.’402
Does state practice reflect the ideal picture of collective self-defence as a protection for small states rather than as a pretext for furthering Cold War or neo-colonial interests? For the most part, the more cynical view appears to be the more accurate as regards the actual use of force. All the state practice on collective self-defence since the Second World War has been controversial. The USSR subsequently disavowed its invasion of Hungary and Czechoslovakia and acknowledged that the Brezhnev doctrine of limited sovereignty was not compatible with international law.403 Some of the other episodes may be seen as showing a fundamental clash of perceptions; the situation could be seen either as one of a civil (p. 198) war with outside interference to further the political aims of the third state or as collective self-defence against an outside attack. The US intervention in Vietnam is just the most dramatic instance of this. The episodes where the USA or the UK intervened in Arab states may also be seen in this way: were they propping up unpopular rulers against regional pressure for change or were they saving the victims of outside aggression? And the same question arose of the French intervention in Chad. The use of collective self-defence in Operation Enduring Freedom in Afghanistan and in Syria raised other questions.404
On the positive side, it is possible to argue that during the Cold War the simple existence of collective self-defence treaties—not only NATO and the now defunct Warsaw Pact, but also the treaties between the USA, the USSR and former colonial powers and smaller states—may have acted as a deterrent to attack and thus protected small states.405 But this conclusion is necessarily speculative. The same state practice would be equally open to an alternative construction that these treaties in fact served to legitimate intervention by states parties.
These debates remain topical today. Japan’s reinterpretation of its Constitution to allow it to act in collective self-defence of its allies provoked significant domestic and regional concern.406 The USA’s relationships with the coastal states of the South China Sea—at a time of great instability caused by China’s sweeping maritime claims and its militarization of the artificial islands it has constructed—have also given rise to discussion about possible US military action in the region in the name of collective self-defence. Western attacks on Syrian territory on the basis of collective self-defence of Iraq have been evaluated in very different ways. Some saw collective self-defence as a pretext for anti-Assad intervention in Syria; others supported a wide right of self-defence against ISIS even without Syria’s consent. Turkey’s assertion of a right to act in collective self-defence in that conflict was criticised by many as a pretext for attacks on the PKK rather than genuine collective self-defence of Iraq against ISIS. Turkey’s involvement in the Syrian conflict also brought the possibility that it would involve NATO to further its References(p. 199) own agenda against Russia. The military interventions of Russia in non-NATO member states Georgia (2008) and Ukraine (2014) have led to renewed discussion of the role of NATO in Europe, and in particular as to whether it should refocus on its traditional function of collective self-defence against the perceived danger from Russia.
1 Cot, Pellet and Forteau, La Charte des Nations Unies (3rd edn 2005), 1329; Simma et al. (eds), The Charter of the United Nations: a Commentary (3rd edn 2012), 1399; Alexandrov, Self-Defense Against the Use of Force in International Law (1996); Bowett, Self-Defence in International Law (1958); Brownlie, International Law and the Use of Force by States (1963); Zourek, ‘La notion de légitime défense en droit international’, 56 Annuaire de l’Institut de Droit International (1975) 1.
2 See Chapter 5.
3 See Treves, ‘La Déclaration des Nations Unies sur le renforcement de l’efficacité du principe du non recours à la force’, 1987 AFDI 379; Gray, ‘The Principle of Non-use of Force’ in Lowe and Warbrick (eds), The United Nations and the Principles of International Law (1994), 33 at 38.
4 On the inclusion of such attacks on land, sea, or air forces as armed attacks giving rise to a right to self-defence, see Simma, note 1 above at 1411. See also the Definition of Aggression, GA Res 3314 (1974) Article 3(d); NATO Treaty Article 6(1); and the Report of the EU Independent International Fact-Finding Mission on the Conflict in Georgia (2009), Vol II, Chapter 6, 264–7. In the Oil Platforms case (Iran v USA) the Court said 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’, ICJ Reports (2003) 161 at para 72.
6 ICJ Reports (2002) 303, paras 308–24. This was predominantly a boundary dispute between Cameroon and Nigeria, but Cameroon also claimed that Nigeria had illegally used force against it, including a full-scale invasion of the Lake Chad area in 1987 and a series of attacks into the Bakassi peninsula. Cameroon argued that in thus invading and occupying its territory Nigeria had violated Article 2(4) of the UN Charter and the principle of non-intervention. Nigeria replied that it was in peaceful possession of the disputed territory and that any use of force had been in self-defence. Although the parties had produced very extensive arguments on the use of force, the Court dealt with this part of the case only briefly. It simply noted that its decision on the location of the boundary meant that Nigerian forces and administration were in place in areas which the Court had determined were Cameroonian territory. Therefore Nigeria was under an obligation expeditiously and without condition to withdraw its administration and forces from those areas. The Court refused to order Nigeria to make guarantees of non-repetition as it ‘cannot envisage a situation’ where either party would fail to respect the territorial sovereignty of the other (paras 318–19). Thus the Court effectively avoided a decision on the use of force in self-defence in this case. For further discussion, see Gray, ‘The Use and Abuse of the International Court of Justice: Cases concerning the Use of Force after Nicaragua’, 14 EJIL (2003) 467. The ICJ also avoided a decision on whether Nicaragua had violated Article 2(4) in Certain Activities carried out by Nicaragua in the Border Area (Costa Rica v Nicaragua), ICJ Reports (2015), paras 96–7.
7 ICJ Reports (2003) 161. In this case the parties again produced very extensive argument on the use of force during the 1980–8 Iran/Iraq war. But the Court held that the USA had failed to establish that Iran was responsible for armed attacks on US-flagged vessels and aircraft entitling the USA the right to use force in self-defence: see 150 below.
8 ICJ Reports (2005) 168. In this case, as regards a large part of the DRC claims against Uganda for unlawful use of force, the Court was able to avoid the problematic task of establishing the facts because they were relatively uncontested between the parties; the crucial issue was how the facts should be characterized (paras 55, 72). Uganda admitted that its troops were present in the DRC and that they had carried out certain operations, but it justified these actions on the basis of consent by the DRC and self-defence. The Court rejected these justifications, and so it followed that Uganda had acted unlawfully. However, in some instances Uganda denied the presence of its troops and the Court had to set out its approach to the facts (paras 55–71) and to try to establish what had happened. See Teitelbaum, ‘Recent Fact-finding Developments at the ICJ’, 6 The Law and Practice of International Courts and Tribunals (2007) 129. In contrast the Eritrea/Ethiopia Claims Commission in Ethiopia’s Ius ad Bellum Claims 1–8, 45 ILM (2006) 430, did not include any detailed discussion of the crucial issue of the establishment of the facts: see Gray, ‘The Eritrea/Ethiopia Claims Commission oversteps its boundaries: a Partial Award’, 17 EJIL (2006) 699.
9 Request for Provisional Measures, ICJ Reports (2008) 353. The EU Independent International Fact-finding Mission, note 4 above, found that it was Georgia that had started the conflict.
10 UN Press Release SG/SM/7410, 22 May 2000; SC Res 1171 (1998), 1226 (1999), 1297 (2000). The Eritrea/Ethiopia Claims Commission in Ethiopia’s Ius ad Bellum Claims 1–8, 45 ILM (2006) 430, nevertheless showed no hesitation in asserting jurisdiction to decide this controversial question, and in finding that it was Eritrea which was responsible for a violation of Article 2(4).
11 SC Res 1862 (2009). The SC Fact-finding Mission said that there was a link between these events and the Eritrea/Ethiopia peace process. A breakthrough in that process would help secure Eritrean cooperation on demilitarizing its border with Djibouti (2008 UNYB 315, 2009 UNYB 297). Ethiopia remains in occupation of Eritrean territory in 2017, and there were further clashes between the two states in 2012 and 2016.
12 1991 UNYB 165; UN doc S/23273, 9 December 1991. After the overthrow of Saddam Hussein the new government of Iraq announced that it had been the aggressor in the 1980–8 conflict, Keesing’s (2005) 46656.
13 For example, Greece 1947–8 UNYB 63, 337; 1948–9 UNYB 238; Indonesia/Netherlands 1947–8 UNYB 369; 1948–9 UNYB 212; Korea 1950 UNYB 245, 251-1; Laos 1959 UNYB 62; Cambodia/Thailand 1959 UNYB 80. On the early practice of the UN in establishing responsibility, see Higgins, The Development of International Law through the Political Organs of the United Nations (1963), 166.
15 For example, Bowett, Self-Defence in International Law (1958); Schwebel, ‘Aggression, Intervention and Self-Defense in Modern International Law’ 136 Hague Recueil des Cours (1972-II) 463; McDougal and Feliciano, Law and Minimum World Public Order (1961).
16 See Tladi, ‘An assessment of Bethlehem’s principles on the use of force against non-state actors’,107 AJIL (2013) 570; see also Kammerhofer, ‘The resilience of the restrictive rules on self-defence’, in Weller (ed), Oxford Handbook of the Use of Force in International Law (2015), 627.
18 After 9/11 some states and writers apparently assumed that the right of self-defence in Article 51 is susceptible to expansion in the context of the ‘war on terror’ and thus not frozen as it had been in 1945: see Chapter 5.
21 On Grenada, UN doc S/PV 2677 (1987); on Tripoli, 1986 UNYB 247. In contrast, there was little discussion in the Security Council of the US action against sites in Afghanistan and Sudan in response to the terrorist attacks on its embassies in Kenya and Ethiopia in August 1998: ‘Contemporary Practice of the US’, 93 AJIL (1999) 161. There was also relatively little discussion of the legality of the 2006 conflict in the Lebanon and of the 2006 Ethiopian intervention in Somalia: see 218 below.
22 Nicaragua case, Jurisdiction and Admissibility, ICJ Reports (1984) 551, paras 92–3; the Court rejected this argument, saying that the USA was attempting to transfer municipal law concepts of separation of powers to the international plane, whereas these concepts are not applicable to the relations among international institutions for the settlement of disputes. Also the fact that a matter is before the Security Council should not prevent it being dealt with by the Court.
23 Combacau, ‘The exception of self-defence in UN practice’, in Cassese (ed), The Current Legal Regulation of the Use of Force (1986), Chapter 13; Delivanis, La légitime défense en droit international public moderne (1971).
26 On the conflict in the DRC, see Chapter 3.
27 See Chapter 5.
36 Ethiopia’s Ius ad Bellum Claims 1–8, 45 ILM (2006) 430 at para 11. The EU Independent International Fact-Finding Mission on the Conflict in Georgia, note 4 above at 246, 269, noted that both Georgia and Russia had reported to the Security Council.
37 See 218 below. Ethiopia did not report its 2011 intervention in Somalia against AlShabaab. The Report of the Secretary-General, UN doc S/2011/759 at para 8, said that it was acting with the government. Nor did it report its March 2012 incursion into Eritrea.
38 There was, however, some genuine concern about the issue; the UN Secretary-General in The Report of the Special Committee on Enhancing the Effectiveness of the Principle of the Non-Use of Force in International Relations (1986), UN doc A/41/41, called for consideration of the possibility that the Security Council might inquire into episodes when the states involved had not reported.
39 This is apparent from a search of communications to the Security Council. A misleading impression was given by writers who still relied on an earlier, pre-Nicaragua account: Simma (ed), The Charter of the United Nations: A Commentary (1st edn 1994), 677, note 148, (2nd edn 2002), 804, note 152; Schacter, ‘Self-Defense and the Rule of Law’, 83 AJIL (1989) 259, Greig, ‘Self-Defence and the Security Council’, 40 ICLQ (1991) 366, and Ronzitti, ‘The expanding law of self-defence’, 11 JCSL (2006) 343 at 356 all rely on Combacau’s earlier, pre-Nicaragua account based on The Repertoire of Practice of the Security Council up to 1974. Combacau, ‘The Exception of Self-defense in UN Practice’, in Cassese (ed), The Current Legal Regulation of the Use of Force (1986), Chapter 13, also took a formalistic approach in that he distinguished between a special report and ordinary communications to the Security Council or statements in debates. Thus there is a danger that a myth of non-reporting will be perpetuated. The most recent edition of Simma’s Commentary, note 1 above at 1424–5, abandoned this position.
40 The existence of an ongoing armed conflict may be controversial, as in the USA’s claim of an ongoing conflict with Al Qaida inside and outside Afghanistan, and later of an ongoing conflict with ISIS. The USA justifies its controversial targeted killings programme on the basis of an ongoing right of self-defence, initiated in response to the 9/11 attacks, the USA does not report individual drone attacks or air strikes under Article 51: see Chapter 5.
41 Repertoire of the Practice of the Security Council 1985–88, Part XI, note 141.
42 Repertoire of the Practice of the Security Council 1981–84, Part XI, note 73.
45 Iran took both these cases to the ICJ. In the first, the Case concerning Oil Platforms, the Court held that the USA had not acted in self-defence (ICJ Reports (2003) 161 at paras 38–78). The second case, Aerial Incident of 3 July 1988, was withdrawn in 1996 after a settlement between the parties and the payment by the USA of ex gratia compensation.
49 Higgins, The Development of International Law through the Political Organs of the United Nations (1963) at 198, 206; Waldock, ‘The Regulation of the Use of Force by Individual States in International Law’, 81 Hague Recueil des Cours (1952-II) 496. On drafting history, see Halberstam, ‘The Right to Self-Defense once the Security Council takes action’, 17 Michigan JIL (1995–6) 229; see also Chayes, Reisman, and Schacter in Damrosch and Fisher (eds), Law and Force in the New International Order (1991), 1, 26, 65; Franck and Patel, ‘UN Police Action in Lieu of War’, 85 AJIL (1991) 63; Rostow, ‘Until What? Enforcement Action or Collective Self-Defense?’, 85 AJIL (1991) 506.
51 See De Guttry and Ronzitti (eds), The Iran–Iraq War (1980–1988) and the Law of Naval Warfare (1993) at 219, 226; Gray, ‘The British Position with regard to the Gulf Conflict’, 37 ICLQ (1988) 420 at 427, 40 ICLQ (1991) 464 at 466.
52 SC Res 661 (1990). Greenwood, ‘New World Order or Old?’, 55 MLR (1992) 153; Warbrick, ‘The Invasion of Kuwait by Iraq’, 40 ICLQ (1991) 482; UN Blues Book Series Vol IX, The UN and the Iraq/Kuwait Conflict 1990–1996 (1996) at 16.
53 Repertoire of Practice of the Security Council 1975–1980, 311. This question also came up over Sierra Leone when those who wanted to defend the supply of arms to the legitimate government claimed that the arms embargo applied only to those who had seized power in a coup. However, this was not expressed in SC Res 1132 (1997). See UK Parliamentary Report of the Sierra Leone Arms Investigation (1998).
54 Gray, ‘Bosnia and Herzegovina: Civil War or Inter-State Conflict? Characterization and Consequences’, 67 BYIL (1996) 155; Report of the Secretary-General pursuant to GA Resolution 53/35 (1998), ‘Srebrenica report’, UN doc A/54/549 (1999).
55 The ICJ, in Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Provisional Measures) ICJ Reports 1993, 3, 325, was also faced with a claim by Bosnia for the lifting of the arms embargo, but it decided that this was not within its jurisdiction; see note by Gray, 43 ICLQ (1994) 704; ‘Srebrenica report’ note 54 above, paras 99–102.
56 Rwanda successfully campaigned for the lifting of the arms embargo imposed on it because of the internal conflict on the grounds that this made it vulnerable to outside interference: 1995 UNYB 347.
62 See note 4 above.
63 On the special question of nuclear weapons, see Advisory Opinion on the Legality of the Threat or Use of Force, ICJ Reports (1995), 226 at paras 35–6, 38–48; Boisson de Chazournes and Sands (eds), International Law, the International Court of Justice and Nuclear Weapons (1999). For an early discussion of naval mines and modern missiles, see O’Connell, The Influence of Law on Sea Power (1975), 70.
64 As regards radar-guided missiles, some states argue that an armed attack begins when the radar guiding the missile is locked on ready to fire. The rules of engagement of their armed forces reflect this approach. For example, in 1998 US aircraft in the no-fly zone over Iraq fired at a missile battery when its radar had locked on to planes patrolling the zone. There was controversy over whether the radar had actually locked on (and over the right of the planes to fly over Iraq), but the idea that an armed attack started when the radar locked on was apparently accepted by Iraq and other states: Keesing’s (1998) 42368. This contrasts with the hostile reaction that the USA and the UK met later when they further extended their rules of engagement to allow a wider range of targets (see 362 below).
65 With regard to naval mines, the difficulty of fitting these into the traditional conception of self-defence became apparent during the Iran/Iraq war. The USA provided convoys for US-flagged vessels through the Gulf, and some of its vessels were harmed by mines. The USA held Iran responsible and used force in response, claiming self-defence. Also the Iran Ajr was detected laying mines; the USA boarded and seized the Iranian vessel. Because the USA claimed not to be a party to the conflict, it had to justify its actions in protection of US-flagged vessels incident by incident rather than invoke self-defence once to cover its entire operation. It had to explain its actions against Iranian minelayers in terms of Article 51 rather than just the laws of war. Accordingly it argued that its actions taken to intercept minelaying vessels were in self-defence. See Gray, ‘The British Position in Regard to the Gulf Conflict’, 37 ICLQ (1988) 420 at 427; Thorpe, ‘Mine Warfare at Sea’, 18 Ocean Development and International Law (1987) 255; Nordquist and Wachenfeld, ‘Legal Aspects of Reflagging Kuwaiti Tankers and the Laying of Mines in the Persian Gulf’, 31 German YIL (1988) 138.
Also, in response to mine damage to the USS Samuel B. Roberts in 1988, the USA attacked and destroyed Iranian oil platforms which it said had been used as a base for Iranian military operations. The USA reported its actions to the Security Council under Article 51 as self-defence, but the justification that it offered made the action appear more like a reprisal: Gray, ‘The British Position in Regard to the Gulf Conflict, Part II’, 40 ICLQ (1991) 464; 1987 UNYB 235, UN doc S/19149, 22 September 1987; see also De Guttry and Ronzitti (eds), The Iran–Iraq War (1980–1988) and the Law of Naval Warfare (1993) at 195–7, 222–3. The legality of the US actions was considered by the ICJ in the Case Concerning Oil Platforms, ICJ Reports (2003) 161, paras 72, 38–78. The Court said 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” ’, but on the facts the US use of force was held not to be lawful self-defence.
66 NATO has decided that cyber defence is part of its core task of collective self-defence, and that it has recognized cyberspace as a domain of operations in which NATO must be defending itself. A decision as to when a cyber attack would lead to the invocation of collective self-defence would be taken on a case-by-case basis. NATO Cyber defence <www.nato.int/cps/en/natohq/topics_78170.htm>.
67 See Simma note 1 above at 1419; Fleck, ‘Searching for international rules applicable to cyber-warfare’, 18 JCSL (2013) 331); Roscini, Cyber Operations and the Use of Force in International Law (2014).
69 See Gray, ‘The ICJ and the Use of Force’, in Tams and Sloan (eds), The Development of International Law by the ICJ (2013), 237. For a different view, see Green, The ICJ and Self-Defence in International Law (2009).
74 For example, Franck, ‘Some Observations on the ICJ’s Procedural and Substantive Innovations’, 81 AJIL (1987) 116; Norton Moore, ‘The Nicaragua case and the Deterioration of World Order’, 81 AJIL (1987) 151; Macdonald, ‘The Nicaragua case: New Answers to Old Questions’, 1986 Canadian YIL 127; Higgins, Problems and Process (1994), 251.
75 For early support for this view, see Brownlie, ‘International Law and the Activities of Armed Bands’, 7 ICLQ (1958) 712; Gill, ‘The Law of Armed Attack in the Context of the Nicaragua case’, 1 Hague YIL (1988) 30. See also Judge Schwebel, Dissenting Opinion, Nicaragua case, paras 157–8.
76 Judge Ago, in his Separate Opinion, 181, para 7, expressed reservations about the legal significance of General Assembly resolutions. In contrast, Judge Schwebel was prepared to accept the Definition of Aggression as reflecting customary international law (Dissenting Opinion, para 168). See Gray, ‘The Principle of Non-Use of Force’, in Lowe and Warbrick (eds), The United Nations and the Principles of International Law (1994), 33.
77 See Chapter 5.
80 Nevertheless Judge Schwebel in his otherwise sweeping rejection of the majority judgment recognized that there was ‘room for the Court’s construction of the legal meaning of armed attack’, Dissenting Opinion, para 15.
81 Dissenting Opinion, paras 543–4; Jennings’ argument echoes that of Reisman, ‘Coercion and self-determination: construing Charter Article 2(4)’, and is open to the rebuttal by Schacter, ‘The legality of pro-democratic invasion’, 78 AJIL (1984) 642, 646.
83 Franck, ‘Some observations on the ICJ’s Procedural and Substantive Innovations’, 81 AJIL (1987) 116 at 120; Norton Moore, ‘The Nicaragua case and the deterioration of World Order’, 81 AJIL (1987) 151 at 154.
84 This is the view taken in Simma note 1 above at 1415.
86 Judge ad hoc Kateka in his Dissenting Opinion, paras 13–34, criticized this narrow approach and said that the Court should not have followed the decision of the Court in the Nicaragua case, but should have adopted that of Judge Jennings.
87 Brownlie, ‘International Law and the Activities of Armed Bands’, 7 ICLQ (1958) 731; Cot, Pellet, and Forteau (eds), La Charte des Nations Unies (3rd edn 2005), 1343; Rifaat, International Aggression: A Study of the Legal Concept (1979), Chapter 15; Lamberti-Zanardi, ‘Indirect Military Aggression’, in Cassese (ed), Current Legal Regulation of the Use of Force (1986), 111.
88 This question of state involvement in irregular actions arises in several different contexts: the definition of intervention, aggression, armed attack, and use of force. The drafting of the GA resolutions on Friendly Relations, Definition of Aggression, and Non-Use of Force reflected differences between states on these issues. There are also questions about the degree of state involvement necessary for the acts of armed bands to give rise to state responsibility. Since 9/11 much has been written on the relationship between issues of state responsibility and the definition of armed attack: see Becker, Terrorism and the State (2006).
92 The Court was able to do so because Uganda had not clearly argued for such a wide view of armed attack. In its written pleadings Uganda did not justify its use of force against the DRC on the basis that self-defence extended to action against non-state actors in a third state in the absence of substantial involvement of that state in the sending of the armed bands. Rather it argued for a wide, four-fold concept of armed attack by a state, to include: (1) sending by a state; (2) provision of logistical support in the form of arms, training, or financial assistance by a state; (3) the operation of armed groups forming part of the command structure of the state; and (4) a conspiracy between the state and the armed bands (Counter-Memorial, para 359). Uganda went further in its oral pleadings, arguing that ‘toleration of armed bands by the territorial state generates responsibility and therefore constitute armed attacks for the purpose of Article 51. Failure to control the activities of armed bands creates a susceptibility of action in self-defence by neighbouring states’: Ugandan Oral Pleadings, CR 2005/7, para 80, quoted by Judge Kooijmans in his Separate Opinion para 21. This conflation of state responsibility for unlawful intervention by irregular forces and armed attack has proved attractive to some commentators, but it confuses primary rules (the definition of armed attack) and secondary rules (the attribution of state responsibility for the breach of those primary rules). The Ugandan oral argument was also implicitly rejected by the Court in this case in its finding that failure to control does not amount to toleration of armed bands in violation of the duty of vigilance (DRC v Uganda, ICJ Reports (2005) 168, paras 277, 300–1; see Chapter 3).
94 Judge Simma, Separate Opinion, paras 4–15; Judge Kooijmans, Separate Opinion, paras 16–31. Judge Kooijmans went on to suggest that such a reaction by the attacked state might be called an act under the state of necessity or ‘extra-territorial law enforcement’, but he gave no detailed discussion of the significance of these terms. Dinstein takes this view in War, Aggression and Self-Defence (5th edn 2011) 268.
95 See Chapter 5 .
97 See, for example, Murphy, ‘Self-defense and the Wall Opinion’, 99 AJIL (2005) 62; Wedgwood, ‘The ICJ Advisory Opinion on the Israeli security fence and the limits of self-defense’, 99 AJIL (2005) 52; Tams, ‘Light treatment of a complex problem: the law of self-defence in the Wall case’, 16 EJIL (2005) 963.
99 As Tams seems to have assumed in ‘Light treatment of a complex problem: the law of self-defence in the Wall case’, 16 EJIL (2005) 963. He is more cautious in ‘Note Analytique: Swimming with the tide or seeking to stem it’, 18 Revue quebecoise de droit int (2005) 275.
100 Judge Buergenthal, Separate Opinion, para 6. Judge Higgins, Separate Opinion, paras 33–5. Judge Higgins made the important point that the invocation of Article 51 was not appropriate in this situation as the construction of the wall was not a forcible action. Both judges were unhappy with the Court’s view that there could not be self-defence where the attacks emanate from occupied territory.
102 On self-defence against non-state actors, see Becker, Terrorism and the State (2006); Jinks, ‘State responsibility for the acts of private armed groups’, 4 Chicago JIL (2003) 83; Kammerhofer, ‘The Armed Activities case and non-state actors in self-defence law’, 20 Leiden JIL (2007) 89; Murphy, ‘Terrorism and the concept of armed attack in Article 51 of the UN Charter’, 43 Harvard JIL (2002) 41; Ruys and Verhoeven, ‘Attacks by private actors and the right of self-defence’, 10 JCSL (2005) 289; Travalio and Altenburg, ‘Terrorism, State responsibility and the use of military force’, 4 Chicago JIL (2003) 97; Tams, ‘Note Analytique: Swimming with the tide or seeking to stem it’, 18 Revue quebecoise de droit int (2005) 275.
103 For example, 1966 UNYB 117, 122; 1967 UNYB 123, 131; 1968 UNYB 159; 1969 UNYB 135, 137, 140; 1970 UNYB 187, 191, 192; 1971 UNYB 113, 116, 119, 121; 1972 UNYB 136; 1973 UNYB 109. See Alexandrov, Self-Defense against the Use of Force in International Law (1996), 179.
104 For example, South Africa argued host state support by Angola: 1981 UNYB 217, Mozambique, 1981 UNYB 228, Botswana, 1985 UNYB 189, Zambia, 1980 UNYB 263. See Alexandrov, Self-Defense against the Use of Force in International Law (1996), 180.
105 For example, 1969 UNYB 200, 1970 UNYB 227, 1978 UNYB 295, 1982 UNYB 428 at 431. See Alexandrov, Self-Defense against the Use of Force in International Law (1996), 174. For more recent cross-border action by Israel, see Chapter 5. In recent years Israel has responded forcibly to attacks by Hezbollah across its border with Lebanon; it has claimed to be acting in self-defence. It has taken care to attribute responsibility for the Hezbollah attacks to Lebanon because of its alleged collusion and support for Hezbollah.
107 UN doc S/PV 1944 (1976); 1985 UNYB 180 at 184. Hot pursuit was rejected by India and Nigeria (UN doc S/PV 2606, 1985), by Trinidad (UN doc S/PV 2607, 1985). In UN doc S/PV 2616 (1985) Mozambique said South Africa has abandoned hot pursuit. See Kwakwa, ‘South Africa’s May 1985 Military Incursions into Neighbouring African States’, 12 Yale JIL (1987) 421. The white minority government in Rhodesia also espoused hot pursuit; see Luttig, ‘The legality of the Rhodesian military operations inside Mozambique—the problem of hot pursuit on land’, 3 South African YIL (1977) 136.
113 On pre-emptive action by South Africa against the ANC, see, for example, UN doc S/PV 2598 (1985); against SWAPO, UN doc S/PV 2606 (1985), Botha, ‘Anticipatory Self-Defence and Reprisals Re-examined’, 11 South African YIL (1985–86) 138. On pre-emptive action by Israel, see, for example, 1982 UNYB 428 at 435, O’Brien, ‘Reprisals, Deterrence and Self-Defense in Counterterror Operations’, 30 Virginia JIL (1990) 421; Alexandrov, Self-Defense against the Use of Force in International Law (1996), 174, 180.
116 SC Res 387 (1976), 1976 UNYB 171. Most extreme was the US statement in 1981 when it vetoed a draft resolution condemning the large-scale invasion by South Africa into Angola. The USA said that the draft blamed South Africa alone for the escalation of violence, but the presence of Cuban troops and USSR military advisers in Angola had fuelled the explosive atmosphere: 1981 UNYB 217. The USA made a similar statement in 1987, but was nevertheless prepared to vote for the resolution condemning South Africa’s invasion of Angola: 1987 UNYB 167.
120 1993 UNYB 382. Tajikistan said that mujahedin and sub-units under the ministry of defence of Afghanistan were responsible for a major incursion into Tajikistan on 13 July 1993: Russia UN doc S/26110, 19 July 1993; Tajikistan UN doc S/26092, 16 July 1993; Afghanistan UN doc S/1994/310.
134 Keesing’s (2007) 48265. See also, US DoD News Briefing, 14 December 2007, <www.defenselink.mil.transcripts.aspx?transcriptid=4106>.
137 See Chapter 5.
138 Iraq again protested about the violation of its airspace in July and October 2012, Reuters, ‘Iraq warns Turkey against violating airspace’, 17 July 2012, <www.reuters.com/assets/print?aid=USBRE86GO0220120717>, ‘Iraq tells Turkey to stop pursuing Kurdish rebels over border’, 2 Oct 2011, <www.reuters.com/assets/print?aid=UKBRE89117Q20121002>.
144 Ibid, para 51. Some judges—Judges Buergenthal, Higgins, Kooijmans, Owada, and Parra-Aranguren—argued that the Court had acted outside its jurisdiction in considering this question of self-defence. The Court had decided that there was no violation of the 1955 Treaty of Amity because there was no actual interference with commerce between the territories of the two parties; there was therefore no need for the Court to consider whether the US use of force had been justified under Article XX of the Treaty of Amity as action necessary to protect its essential security interests.
145 ICJ Reports (2003) 161, paras 51–61. Some judges were critical of the Court’s approach to the burden of proof and the standard of evidence: Higgins, paras 30–9; Buergenthal, paras 33–46; Owada, paras 41–52.
146 For the debate (and useful references) on this issue see ICJ Pleadings, Iran Reply para 7.36; US Rejoinder, paras 5.16, 5.19. The EU Independent International Fact-Finding Mission on the conflict in Georgia stated in its Report (hereafter the Georgia Report), Vol 2, p 266, that attacks on Russian private property in Georgia did not amount to armed attacks: <http://www.mpil.de/en/pub/publications/archive/independent_international_fact.cfm>.
148 ICJ Reports (2003) 161, paras 62–4. See Chapter 5.
151 The USA addressed the Iranian arguments on intent which were accepted by the Court in its ICJ Pleadings (US Rejoinder, paras 5.23, 5.26). However, the US argument focused on the laws of naval warfare rather than the concept of armed attack.
153 The Georgia Report said that specific intent was necessary to establish that there had been an armed attack on Russia, given that there was a conflict between two other parties, Georgia and South Ossetia, note 86 above, Vol 2, p 266. See 168 below for further discussion of this conflict.
163 The distinction between mere frontier incidents and other more significant uses of force was discussed during the drafting of the Definition of Aggression. The proposal to include a de minimis clause (to exclude minor incidents) was first made by Finland in 1972: see Ferencz, Defining International Aggression (1975) at 367; Schwebel, ‘Aggression, Intervention and Self-Defense in Modern International Law’ 136 Hague Recueil des Cours (1972-II) 463 at 467–8; Fitzmaurice, ‘The Definition of Aggression’, 1 ICLQ (1952) 137 at 139.
164 Ibid, para 231.
165 Schwebel, ‘Aggression, Intervention and Self-Defense in Modern International Law’, 136 Hague Recueil des Cours (1972-II) 463; Ferencz, ‘Defining Aggression: Where it stands and where it’s going’, 66 AJIL (1972) 491.
167 For example, this is the view of Higgins, Problems and Process (1994), 251. Others support the position that only serious acts count as armed attacks, and that small border incidents did not count: Mullerson, ‘Self-Defense in the Contemporary World’ in Damrosch and Scheffer (eds), Law and Force in the New International Order (1991), 13; Lamberti Zanardi, ‘Indirect Military Aggression’ in Cassese (ed), The Current Legal Regulation of the Use of Force (1986), 111; Rifaat, International Aggression (1979), Chapter 11.
168 Dinstein, note 94 above, 555; Schacter, ‘In Defense of International Rules on the Use of Force’, 53 University of Chicago Law Review (1986) 113; Macdonald, ‘The Nicaragua case: New Answers to Old Questions’, 1986 Canadian YIL 127 at 151; Hargrove, ‘The Nicaragua Judgment and the Future of the Law of Force and Self-defense’, 81 AJIL (1987) 135 at 139.
169 Reisman, ‘Allocating Competences to use Coercion in the post Cold-War World, Practises, Conditions, and Prospects’, in Damrosch and Scheffer (eds), Law and Force in the New International Order (1991), 26 at 39–40.
171 For example, Kunz, ‘Individual and Collective Self-Defense in Article 51 of the Charter of the United Nations’, 41 AJIL (1947) 872 at 878; Badr, ‘The Exculpatory Effect of Self-defense in State Responsibility’, 10 Georgia JICL (1980) 1.
173 For some support for this view, see Farer, ‘Drawing the Right Line’, 81 AJIL (1987) 112; Diaz Barrado, El Consentimento Causa de Exclusion de la Ilicitud del Uso de la Fuerza en Derecho Internacional (1989).
177 Ibid at 302.
179 A few writers have rejected these limits on self-defence as not established in customary international law: Kunz, ‘Individual and Collective Self-defence in Article 51 of the Charter of the UN’, 41 AJIL (1947) 872; Delivanis, La légitime défense en droit international public moderne (1971), Chapter 2. See also, Gardam, ‘Proportionality and Force in International Law’, 87 AJIL (1993) 391 and Gardam, Necessity, Proportionality and the Use of Force by States (2004).
181 For example, the UAE referred to the Caroline incident in UN doc S/PV 2616 (1985); the GDR invoked it over the clashes between the USA and Libya (UN doc S/PV 2677, 1986). When an advisory report to the Dutch government recommended use of the Caroline incident to determine whether an armed attack was imminent (Pre-emptive Action, No.36, AIV/No.15, CAVV, July 2004) the government response was, ‘Whether the Caroline criteria are adequate today for deciding whether a threat is sufficiently imminent to justify exercising the right of self-defence, only experience will tell. While these criteria are a useful tool, one cannot exclude the possibility that they will have to be refined at some point.’ (Government Letter to the House of Representatives, 29800 V, no.56, 29 October 2004). More recently the USA and the UK have invoked the Caroline incident to support their wide claims to self-defence in the context of counter-terrorism: see 241 below.
182 Cot, Pellet, and Forteau (eds), La Charte des Nations Unies (3rd edn 2005), 1330; Brownlie in Butler (ed), The Non-Use of Force in International Law (1989), 17; Kearley, ‘Raising the Caroline’, 17 Wisconsin International Law Journal (1999) 325. Judge Schwebel, in the Nicaragua case, Dissenting Opinion, para 200, argued that the narrow criteria of the Caroline incident concerned anticipatory self-defence only. This issue also came up in the Oil Platforms case, where Iran invoked the Caroline incident as imposing limits on the right of self-defence and the USA argued that it was not relevant to the use of force in a continuing engagement: ICJ Pleadings US Rejoinder, para 5.32, Counter Memorial, para 4.44.
184 ICJ Reports (1996) 226, para 141; the Court went on at para 143 to refuse to decide the issue whether the effects of any use of nuclear weapons would be so serious that it could not constitute a necessary and proportionate measure. It said, ‘Certain states have in their written and oral pleadings suggested that in the case of nuclear weapons, the condition of proportionality must be evaluated in the light of still further factors. They contend that the very nature of nuclear weapons, and the high probability of an escalation of nuclear exchanges, mean that there is an extremely strong risk of devastation. The risk factor is said to negate the possibility of the condition of proportionality being complied with. The Court does not find it necessary to embark upon the quantification of such risks; nor does it need to enquire into the question whether tactical nuclear weapons exist which are sufficiently precise to limit those risks: it suffices for the Court to note that the very nature of all nuclear weapons and the profound risks associated therewith are further considerations to be borne in mind by States believing they can exercise a nuclear response in self-defence in accordance with the requirements of proportionality.’
187 See Gardam, Necessity, Proportionality and the Use of Force by States (2004); Okimoto, The Distinction and Relationship between Ius ad Bellum and Ius in Bello (2011), Chapter 3.
188 See, for example, the wide views of the US State Department Legal Adviser, Taft ‘Self-Defense and the Oil Platforms Decision’, 29 Yale JIL (2004) 295, and of the UK Attorney-General, House of Lords debates, 21 April 2004, Hansard Column 370. The latter said that where use of force is to be used in self-defence in anticipation of an imminent armed attack, ‘First, military action should be used only as a last resort. It must be necessary to use force to deal with the particular threat that is faced. Second, the force used must be proportionate to the threat faced and must be limited to what is necessary to deal with the threat.’
190 There is controversy as to whether self-defence should be immediate, as in the Caroline incident: see Gardam, Necessity, Proportionality and the Use of Force by States (2004), at 149–53; Corten, The Law against War (2010), 485.
191 See the Georgia Report, note 146 above, Vol 2, 269–74.
194 ICJ Reports (2003) 161, paras 73–6. The US State Department Legal Adviser criticized this statement as unduly restrictive and without basis in international law or practice, Taft, ‘Self-Defense and the Oil Platforms Decision’, 29 Yale JIL (2004) 295. But it is not clear that the ICJ was laying down a mandatory requirement for all self-defence; rather the fact that the USA had not complained was part of the evidence that it was not necessary to use force against the oil platforms in response to the attacks on the US ships.
196 Again the US State Department Legal Adviser was critical, Taft, ‘Self-Defense and the Oil Platforms Decision’, 29 Yale JIL (2004) 295 at 305. He said that there is no requirement that a state exercising the right of self-defence must use the same degree or type of force used by the attacking state in its most recent attack. Rather the proportionality of the measures is to be judged according to the threat being addressed. This is a very wide and controversial view of proportionality and one which will be examined in more detail in Chapter 5.
198 Judge Simma (Separate Opinion, para 15) concluded that the USA used these two incidents to teach Iran a broader lesson: ‘nowhere in these materials do we find any trace of the considerations that an international lawyer would regard as necessary in order to justify action taken in self-defence’. Kooijmans (Separate Opinion, paras 52, 55, 62) found it hard to avoid the impression that in reality a punitive intent prevailed; Judge Elaraby (Dissenting Opinion, para 1.2) said that the USA’s aim was punitive and that its actions were reprisals; it would have been advisable for the Court to insert a decisive and straightforward statement that defined the legal character of the US use of armed force.
203 ICJ Reports (2005) 168, para 147. Judge Kooijmans partly dissented on this point in his Separate Opinion, paras 33–4. He said that the seizure of towns and airports in an area contiguous to the border zone was not unnecessary or disproportionate to the purpose of repelling the persistent attacks of the Ugandan rebel movements.
206 See, for example, the US statement with regard to South Africa’s pre-emptive action against Angola in 1985, that there is no inherent right to engage in military activity across one’s border on the basis that it is a pre-emptive strike. The question is whether it is self-defence, a necessary, reasonable, and proportionate response to the danger posed. In this instance the USA said that it was not: UN doc S/PV 2616 (1985).
208 On Israel, see 1978 UNYB 295, 306. The Security Council called for Israel to end its occupation in SC Res 425 (1978). On South Africa, see 1982 UNYB 312; the Security Council called for it to withdraw in SC Res 545 (1983); the General Assembly also called for this in GA Res 36/9 (1981).
210 Feder, ‘Reading the UN Charter connotatively: toward a new definition of armed attack’, 19 New York University JILP (1987) 395; Higgins, The Development of International Law through the Political Organs of the United Nations (1963), 201.
216 Judge Simma strongly rejected the accumulation of events doctrine: ‘Also, there is in international law on the use of force no “qualitative jump” from iterative activities remaining below the threshold of Article 51 of the Charter to the type of “armed attack” envisaged there.’ ICJ Reports (2003) 161, Separate Opinion, para 14.
219 For early practice on protection of nationals, see Bowett, ‘The Use of Force for the Protection of Nationals Abroad’, in Cassese (ed), The Current Legal Regulation of the Use of Force (1986), 39; Schweisfurth, ‘Operations to Rescue Nationals in Third States’, 23 German YIL (1980); Ronzitti, Rescuing Nationals Abroad (1985). On Suez see Marston, ‘Armed Intervention in the 1956 Suez Canal Crisis: the Legal Advice tendered to the British Government’, 37 ICLQ (1988) 773; on Entebbe, see Repertoire of Practice of the Security Council 1975–80, 286; on the attempted rescue of the Iranian hostages, see Stein, ‘Contempt, Crisis and the Court’, 76 AJIL (1982) 499; on Grenada, ‘Contemporary Practice of the US’, 78 AJIL (1984) 200; on Panama, ‘Contemporary Practice of the US’, 84 AJIL (1990) 545.
220 Though this doctrine has been invoked in rather bizarre form in the American Servicemembers Protection Act (2002), para 3008 (popularly called the Hague Invasion Act); this authorized the US President to use all necessary means to bring about the release of American nationals detained by the International Criminal Court at the Hague. The US government’s hostility to the ICC apparently led it to claim a right to use force in protection of nationals. See 96 AJIL (2002) 975. This apparent authorization to use force met strong protests from the Netherlands government.
221 See Gray, ‘The Protection of Nationals abroad: Russia’s Use of Force in Georgia’, in Constantinides and Zaikos (eds), The Diversity of International Law (2009) 133. Russia also hinted at the possible use of this doctrine to justify use of force in Crimea and Ukraine, but as it denied intervention it did not pursue this argument: see Green, ‘The annexation of Crimea’, 1 Journal on the Use of Force (2014) 3.
222 With regard to Iran, the USA reported its action to the Security Council under Article 51, Case concerning United States Diplomatic and Consular Staff in Tehran, ICJ Reports 1979, at 18. The ICJ did not pronounce on the legality of the US action; Judges Morozov and Tarazi, in their Dissenting Opinions, ICJ Reports 1980, 57, 64, said that the US action was not lawful self-defence. With regard to Grenada, the USA invoked Article 51 (UN doc S/16076, 25 October 1983). With regard to Panama, it invoked Article 51 as giving an inherent right of self-defence to protect American lives (UN doc S/21035, 20 December 1989).
224 Georgia Report, note 4 above, Vol 2, 285.
225 On Grenada the draft resolution was defeated by 11–1 (USA) –3 (France, Canada, UK), 1983 UNYB 211. On Panama the draft resolution was defeated by 10–4 (France, UK, USA, Canada) –1, 1989 UNYB 172 at 174.
226 For the USA justification for Grenada, see 1983 UNYB 211, UN doc S/16076, 25 October 1983; UN doc S/PV 2487, 2489, 2491 (1983). Gilmore, The Grenada Intervention (1984); Weiler, ‘Armed Intervention in a Dichotomized World: The Case of Grenada’, in Cassese (ed), Current Legal Regulation of the Use of Force (1986), 241.
227 For the USA justification for Panama, see UN doc S/21035, 20 December 1989; 84 AJIL (1990) 545. Contrasting assessments of the intervention are given by Henkin, ‘The Invasion of Panama under International Law: a Gross Violation’, 29 Columbia JTL (1991) 293 and D’Amato, ‘The Invasion of Panama was a Lawful Response to Tyranny,’ 84 AJIL (1990) 516.
229 Arend and Beck, International Law and the Use of Force: Beyond the UN Charter Paradigm (1993), 107–10; Franck, Recourse to Force (2002), 76–96.
230 Bowett, ‘The Use of Force for the Protection of Nationals Abroad’, in Cassese (ed), The Current Legal Regulation of the Use of Force (1986), 39; Dinstein, note 94 above, 217; Ruys, ‘The Protection of Nationals Doctrine’, 13 JCSL (2008) 233); Grimal and Melling, ‘The Protection of Nationals Abroad’, 16 JCSL (2011) 541.
232 Brownlie, International Law and the Use of Force by States (1963), 432; Akehurst, ‘Humanitarian Intervention’, in Bull (ed), Intervention in World Politics (1984), 95; Ronzitti, Rescuing Nationals Abroad (1985). See discussion by Forteau in Weller (ed), The Oxford Handbook of the Use of Force in International Law (2015), 947.
235 Georgia Report, note 4 above, Vol 1 on the series of events; Vol 2, 238 on the start of the conflict.
238 UN docs S/PV 5953, 5961 (2008). Russia’s reply was to refer to the conflict in Kosovo and to remind the USA that NATO military actions there spread far beyond Kosovo to attack the infrastructure of the rest of Serbia. The Independent International Fact-Finding Mission on the Conflict in Georgia Report, Vol 2, 269–75, found that Russia’s actions were not necessary and proportionate.
241 Georgia Report, note 4 above, Vol 1, para 12; Vol 2, 147.
243 There have also been many instances of states sending in troops to extract nationals and others from dangerous situations where a state was involved in a civil war or domestic unrest. However, issues of legality have not been raised in the UN with regard to these episodes. The state using force has not reported it to the Security Council under Article 51 and the state where the intervention took place did not raise the matter. Nor did other states protest about the use of force. These can therefore be seen as cases of consent or perhaps implied consent by the government to the rescue operation. But many of these cases occurred when there was no effective government: the previous government had been overthrown and the state was in confusion. It seems that third states were willing to acquiesce in the forcible evacuation of nationals; their concern is roused only with regard to those rescue missions where the territorial state objects to the intervention or where the protection of the nationals was just a pretext for an invasion with wider objectives. See discussion by Ruys, and by Grimal and Melling, note 230 above.
244 See Chapter 5.
245 The UK Attorney-General formally reasserted the longstanding UK support for this position in 54 ICLQ (2005) 767; see also his statement to the House of Lords, Hansard House of Lords Debate, 21 April 2004, Cols 369–70. The US position was set out in the 2002 US National Security Strategy at 15.
247 Ferencz, ‘Defining Aggression: Where it stands and where it’s going’, 66 AJIL (1972) 491; Gray, ‘The Principle of Non-Use of Force’, in Lowe and Warbrick (eds), The United Nations and the Principles of International Law (1994), 33.
249 See Ghafur Hamid, ‘The legality of anticipatory self-defence in the 21st century world order’, 54 Netherlands ILR (2007) 441; Gray, ‘A Crisis of Legitimacy for the UN Collective Security System?’, 56 ICLQ (2007) 157.
251 Both Alexandrov, Self-Defense against the Use of Force in International Law (1996) and Arend and Beck, International Law and the Use of Force: Beyond the UN Charter Paradigm (1993), include a very wide range of incidents under the heading collective self-defence; they do not restrict themselves to those episodes where states actually invoked the doctrine. For a detailed examination of state practice, see Corten, note 246 above, and Ruys, note 61 above.
253 It is the subject of much controversy whether the government of Israel genuinely believed that Arab states were about to mount an imminent attack: see, for example, Shlaim, The Iron Wall: Israel and the Arab World (2001) and Oren, Six Days of War (2002) for contrasting accounts.
254 Alexandrov, Self-Defense against the Use of Force in International Law (1996) at 154 uses these as examples of anticipatory self-defence despite the choice of the USA and Israel not to invoke this doctrine. Chayes, ‘Law and the Quarantine of Cuba’, 41 Foreign Affairs (1963) 550.
262 For an early statement of this position, see ‘UK Materials on International Law’, 64 BYIL (1993) 728; more recently, ‘Contemporary Practice of the United States Relating to International Law’, 94 AJIL (2000) 102; ‘UK Materials on International Law’, 73 BYIL (2002) 861 at 867.
264 1981 UNYB 275; Israel explained its action in UN doc S/14510, 8 June 1981; in the Security Council debates Israel referred to writers in support of the doctrine of anticipatory self-defence: UN doc S/PV 2280 (1981) at paras 98, 99, 100; UN doc S/PV 2288 (1981) at para 38. The General Assembly voted to condemn Israel’s action as a premeditated and unprecedented act of aggression in GA Res 36/27 (1981) (109-2-34). Many of those who abstained said that they did so because it was for the Security Council rather than the General Assembly to act. D’Amato defended the legality of the use of force: ‘Israel’s Air Strike upon the Iraqi Nuclear Reactor’, 77 AJIL (1983) 584.
268 Ibid at para 148. For further discussion of the debate on pre-emptive self-defence, see Chapter 5.
272 NAM Comments on the High-level Panel Report, 28 Feb 2005, <www.un.int/malaysia/NAM/NAM/html>, paras 26–8, UN doc A/59/PV.85, 14–15. The NAM repeated its view that Article 51 is a restrictive provision at its 2006 Havana Conference, UN doc S/2006/780, 29 Sept 2006, para 19. See also UN doc S/2014/573.
275 See Chapter 5.
276 Arend and Beck, International Law and the Use of Force (1993); Weisburd, The Use of Force (1997); D’Amato, Prospect and Process (2nd edn 1995). This tendency is again apparent with regard to wide claims to self-defence by the USA after 9/11 and the Syrian conflict, see Chapter 5.
277 The main multilateral treaties are the NATO Treaty (1949) 34 UNTS 243; the (now defunct) Warsaw Pact (1955) 219 UNTS 24; the Rio Treaty (1947) 21 UNTS 77; the Security Treaty between Australia/New Zealand/USA (1951) 131 UNTS 83; South East Asia Collective Defense Treaty (1954) 209 UNTS 20; the Baghdad Pact (1955) 233 UNTS 199; Pact of the Arab League (1945) 70 UNTS 237; Arab League Treaty of Joint Defence (1955) 49 AJIL Supplement (1955) 51; Commonwealth of Independent States Collective Security Treaty (1992); African Union Non-Aggression and Common Defence pact (2005). There are also hundreds of bilateral treaties which provide for collective self-defence. On the more than sixty Soviet bloc treaties made in the 1970s, see Zipfel, Die Freundschafts und Kooperationsverträge der Kommunistischen Staaten (1983). The USA, the UK, and France each have an extensive network of treaties. On French treaties, see The French White Paper on Defence and National Security (2008), 159. For a discussion of the different types of collective self-defence treaty, see Dinstein, War, Aggression and Self-Defence (5th edn 2011), 283.
278 The inclusion of the Soviet intervention in Hungary as an instance of collective self-defence being invoked by a state is very doubtful; apart from a reference to the intervention being in accordance with the Warsaw Treaty, the USSR did not refer to collective self-defence to justify its action (1956 UNYB 67).
281 Cot, Pellet, and Forteau (eds), La Charte des Nations Unies (3rd edn 2005), 1354; Alexandrov, Self-Defence Against the Use of Force in International Law (1996), 216. It is noteworthy that this list is exclusively collective self-defence of territory. The issue of collective self-defence of ships at sea came up in the Iran–Iraq conflict when the USA, on 29 April 1988, decided to extend the protection offered by its naval forces in the Gulf to friendly neutral vessels. It announced that, following a request from the vessel under attack, assistance would be rendered by a US warship or aircraft. It used this power twice in 1988 to protect a Danish and a Panamanian vessel, but apparently did not expressly rely on collective self-defence; it spoke of assistance to vessels in distress, following a request from the vessel under attack: Gray, ‘The British Position with regard to the Gulf Conflict (Iran-Iraq): Part 2’, 40 ICLQ (1991) 465 at 468. See de Guttry and Ronzitti, The Iran–Iraq War (1980–1988) and the Law of Naval Warfare (1993), 196, 304. When this case went to the ICJ the USA did not invoke collective self-defence as justification for its use of force (Case Concerning Oil Platforms, ICJ Reports (2003) 161, para 51). Collective self-defence of ships is now topical again because of the competing territorial claims and military activity in the South and East China Seas.
282 Alexandrov, Self-Defense Against the Use of Force in International Law (1996) at 252; he also includes regional action under this heading of collective self-defence. See also Dinstein, War, Aggression and Self-defence (5th edn 2011), 299; Wood in Weller (ed), The Oxford Handbook of the Use of Force in International Law (2015), 649.
283 See Chapter 5.
284 Mullerson, ‘Intervention by Invitation’, in Damrosch and Scheffer (eds), Law and Force in the New International Order (1991), 127; Diaz Barrado, El Consentimento, Causa de Exclusion de la Ilicitud del Uso de la Fuerza en Derecho Internacional (1989), 78.
286 For example, UN docs S/PV 2481 (1983), 2565 (1984). On the history of Cuban involvement in Angola, see Repertoire of the Practice of the Security Council 1975–1980, 260. The Security Council expressly rejected the doctrine of linkage; see, for example, SC Res 539 (1983).
289 The USA and the UK invoked collective self-defence after the Iraqi invasion of Kuwait in 1990 to justify their naval operations undertaken after the imposition of an economic embargo on Iraq but before the specific authorization of force by the Security Council. Kuwait and other states in the region requested third-state assistance, and Kuwait reported its action under Article 51 to the Security Council (UN Publications, The UN and the Iraq–Kuwait Conflict 1990–1996 at 16.) The USA and the UK imposed a ‘naval interdiction’ to stop ships violating the embargo. In this instance controversy over legality centred on the question whether the USA and the UK were entitled to act after the imposition of an economic embargo without Security Council authority. See Warbrick, ‘The invasion of Kuwait by Iraq’, 40 ICLQ (1991) 482, 964; Greenwood, ‘New World Order or Old?’, 55 MLR (1992) 153 at 161, 164–5.
292 See Chapter 5.
294 Judge Oda and Judge Schwebel go into the history of the drafting of the UN Charter provisions on collective self-defence in their Dissenting Opinions, 1986 ICJ Reports 212 at paras 91–6; 266 at para 194. See also Alexandrov, Self-Defense Against the Use of Force in International Law (1996) at 90; Franck, Recourse to Force (2002) at 48.
295 Brownlie, International Law and the Use of Force by States (1963), 328, 229-30; Higgins, The Development of International Law through the Political Organs of the United Nations (1963), 208; Kelsen, Law of the United Nations (1950) at 793, Delivanis, La légitime défense en droit international public moderne (1971), Part II, Chapter 2; Macdonald, ‘The Nicaragua case: New Answers to Old Questions’, 1986 Canadian YIL 127 at 143.
298 Alexandrov, Self-Defense Against the Use of Force in International Law (1996) at 101; Higgins, The Development of International Law through the Political Organs of the United Nations (1963) at 208; see also Nicaragua case, Jennings, Dissenting Opinion at 544–6.
299 Delivanis, La légitime défense en droit international public moderne (1971), Part II, Chapter 2; Kulski, ‘The Soviet System of Collective Security compared with the Western System’, 1950 (44) AJIL 453.
302 Ibid, para 183. Judge Ago expressed ‘serious reservations with regard to the seeming facility with which the Court—while expressly denying that all the customary rules are identical in content to the rules in the treaties—has nevertheless concluded in respect of certain key matters that there is a virtual identity of content as between customary international law and the law enshrined in certain major multilateral treaties’ (Separate Opinion at 183, para 6). Judge Jennings was similarly sceptical as to whether custom could have developed since the adoption of the UN Charter on the basis of the rules in the Charter (Dissenting Opinion at 531). Other judges in their Separate Opinions argued that the US multilateral treaty reservation should not be given any effect and that the Court could apply the UN Charter and other multilateral treaties (Judge Sette Camara, Separate Opinion 192; Judge Ni, Separate Opinion 201).
303 Nicaragua case, para 194. The Rio Treaty, by which the parties were in fact bound, requires an armed attack in its express provision for collective self-defence in Article 3 (see Nicaragua case, paras 196–7). There is an important distinction between Article 3 which allows collective self-defence in cases of armed attack, and Article 6, which provides for cooperation in response to other types of outside intervention. Many other collective self-defence treaties make the same distinction. Judge Schwebel, in his Dissenting Opinion, blurred this distinction. The NATO Treaty, Article 5 also requires an armed attack (see note 338 below).
In state practice there were no instances of anticipatory collective self-defence being expressly invoked to justify the actual use of force, except perhaps in the Harib fort incident, 1964 UNYB 181; this use of force by the UK was condemned by the Security Council as a reprisal. Judge Schwebel (Dissenting Opinion, paras 172–3) apparently argued that there is a right of anticipatory self-defence, but he did not support this by reference to any state practice on collective self-defence.
311 For criticism of the Court’s doctrine of collective self-defence, see, for example, Franck, ‘Some Observations on the ICJ’s Procedural and Substantive Innovations’, 81 AJIL (1987) 116; D’Amato, ‘Trashing Customary International Law’, 81 AJIL (1987) 101; Hargrove, ‘The Nicaragua Judgment and the Future of the Law of Force and Self-defense’, 81 AJIL (1987) 135; Norton Moore, ‘The Nicaragua case and the Deterioration of World Order’, 81 AJIL (1987) 151; see also Macdonald, ‘The Nicaragua case: New Answers to Old Questions’, 1986 Canadian YIL 127 at 149.
312 Judge Schwebel, Dissenting Opinion, paras 157–8; Brownlie, ‘International Law and the Activities of Armed Bands’, 7 ICLQ (1958) 712; Gill, ‘The Law of Armed Attack in the Context of the Nicaragua case’, 1 Hague YIL (1988) 30.
324 Thus on the extension of the war into Cambodia, see 1972 UNYB 153; Stevenson, ‘US Military Actions in Cambodia: Questions of International Law’, in Falk (ed), The Vietnam War and International Law, Vol 3 (1972), 23 at 31.
325 Some commentators rejected this position: Norton Moore, ‘The Nicaragua case and the Deterioration of World Order’, 81 AJIL (1987) 151 at 154; Norton Moore, ‘The Secret War in Central America and the Future of World Order’, 80 AJIL (1986) 43; Reisman, ‘Allocating Competences to Use Coercion in the Post-Cold War World’, in Damrosch and Scheffer (eds), Law and Force in the New International Order (1991), 26.
330 See contra Macdonald, ‘The Nicaragua case: New Answers to Old Questions’, 1986 Canadian YIL 127 at 143. This issue raised crucial questions with regard to the collective self-defence of Iraq in 2014: see 190 below.
331 Schwebel, Dissenting Opinion, paras 191, 221–7. He argued particularly that these requirements were not appropriate in cases of covert action. But this seems to mistake the nature of the right and the role of the Security Council.
334 Simma (ed), The Charter of the United Nations: A Commentary (2nd edn 2005), 803 at para 38 at para 38. The 3rd edition (2012), 1421 slightly modified this position, but it still asserts that there is no need for a formal and express request. It is interesting to note that Germany in reporting to the Security Council with regard to action in Syria did not mention any request by Iraq, UN doc S/2015/946.
337 The Arab League, Treaty of Joint Defence, Article IV(3), 55 AJIL Supplement 51, and the France/Djibouti Protocol (1982), 1430 UNTS 103 also require a request. Other treaties require ‘consultation’ (UK/Mauritius Agreement on Mutual Defence and Assistance (1968) 648 UNTS 3; UK/Malta Agreement on Mutual Defence (1964) 588 UNTS 55) or ‘agreement’ on the response to an armed attack (USA/Liberia Agreement on Cooperation (1959) 357 UNTS 94).
338 Article 5 provides: ‘The Parties agree that an armed attack against one or more of them in Europe of North America shall be considered an attack against them all and consequently they agree that, if such an armed attack occurs, each of them, in exercise of the right of individual or collective self-defence recognized by Article 51 of the Charter of the UN, will assist the Party or Parties so attacked by taking forthwith, individually and in concert with the other Parties, such action as it deems necessary, including the use of armed force, to restore and maintain the security of the North Atlantic area.’
339 USSR/Hungary (1956), 1956 UNYB 67; USA/Lebanon (1958) UN doc S/PV 827 (1958), 1958 UNYB 36 at 38; UK/Jordan (1958) 1958 UNYB 41; USA/Vietnam (1965) UN doc S/6174 (1965), 60 AJIL (1966) 565; USSR/Czechoslovakia (1968) UN doc S/PV 1441 (1968), 1968 UNYB 298; USSR/Afghanistan (1979), 1980 UNYB 296; Libya/Chad (1980) 1981 UNYB 222; France/Chad (1983, 1986) 1983 UNYB 180, 1986 UNYB 168, 1987 UNYB 176, UN docs S/17837, 18 February 1986, S/18554, 2 November 1987; Angola/Cuba (from 1975) UN docs S/PV 2440, 2481(1983), 1983 UNYB 173; USA/Honduras (1988) UN doc S/PV 2802 (1988), UN doc S/19643, 17 March 1988, 1988 UNYB 170; USA and UK/Kuwait (1990) UN Publications, The UN and the Iraq/Kuwait Conflict 1990–1996 at 16; Russia/Tajikistan (1993) UN doc S/26241, 5 August 1993, 1993 UNYB 514; Angola, Namibia and Zimbabwe/DRC 1998 UNYB 82-6. On Operation Enduring Freedom in Afghanistan, see Chapter 5. On the exceptions to this general practice, see note 346 below.
345 France 1983 UNYB 180, 1984 UNYB 185, 1986 UNYB 168, 1987 UNYB 176, UN docs S/17837, 18 February 1986, S/18554, 2 January 1987, UN doc S/PV 2721 (1986); Repertoire of the Practice of the Security Council (1981-84) 261; see Alibert, ‘L’Affaire du Tchad’, 90 RGDIP (1986) 368.
346 The striking exceptions are Turkey, UN doc S/2015/563, (see 184 above) and Germany, UN doc S/2015/946 (see note 334 above).
348 Dinstein, War, Aggression and Self-Defence (5th edn 2011), 280; Bowett, Self-Defence in International Law (1958) at 216; Macdonald, ‘The Nicaragua case: New answers to old questions’, 1986 Canadian YIL at 151; Delivanis, La légitime défense en droit international public moderne (1971).
349 Macdonald ‘The Nicaragua case: New Answers to Old Questions’, 1986 Canadian YIL 127 argued that ‘if there is an armed attack, what the victim believes to have occurred is otiose because the aid-giving state is also subject to the armed attack’.
350 There were pre-existing treaties between Hungary, Czechoslovakia, Ethiopia, Afghanistan, and the USSR; the UK and the South Arabian Federation; El Salvador, Costa Rica and Honduras and the USA; Chad and France (France invoked a 1976 Cooperation Agreement, but had to stretch its terms; see Alibert, ‘L’Affaire du Tchad’, 90 RGDIP (1986) 343). To justify its collective self-defence of Tajikistan Russia invoked a bilateral Treaty of Friendship (UN doc S/26110, 19 July 1993) and an agreement between five members of the CIS (UN doc S/26892, 18 December 1993, 1993 UNYB 514). In some of these cases the treaty was concluded not long before the use of force, so it seems that even though it is not a legal requirement it may be seen as adding legitimacy. NATO invoked Article 5 of its constituent treaty for the first time and only in response to 9/11: see Chapter 5.
351 Thus the absence of a treaty in the cases of USA/Lebanon (1958), UK/Jordan (1958), Cuba/Angola (from 1975), USA and UK/Kuwait (1990), and Angola, Namibia, and Zimbabwe/DRC (1998) was not singled out as a ground for criticism.
353 As discussed in the previous chapter, the USA did not just report to the Security Council once at the start of the conflict, but made several separate reports of individual actions or series of actions; 1965 UNYB 185, 1966 UNYB 153, 1970 UNYB 215.
354 USA/Lebanon, UN doc S/PV 827 (1958), 1958 UNYB 38; Jordan/UK, UN docs S/4053, 17 July 1958, S/4071, 1958 UNYB 40; USA/Vietnam, 1965 UNYB 185, 1966 UNYB 146, 1970 UNYB 215, 1972 UNYB 153, S/10631; USSR/Czechoslovakia, UN doc S/PV 1441 (1968), 1968 UNYB 298; Libya/Chad, 1981 UNYB 223; France/Chad, UN doc S/PV 2721 (1986), UN docs S/17837, 18 February 1986, S/18554, 2 January 1987, S/19136, 15 September 1987, 1983 UNYB 180, 1984 UNYB 185, 1986 UNYB 168, 1987 UNYB 176; USSR/Afghanistan, 1980 UNYB 296 at 299,300; Cuba/Angola, UN docs S/PV 2440, 2481 (1983), 1983 UNYB 173; USA/Honduras, 1988 UNYB 170; USA and the UK/Kuwait, 1990 UNYB 195; Russia/Tajikistan, UN docs S/26110, 19 July 1993, 26241, 5 August 1993, 1993 UNYB 514; DRC/Angola, Namibia and Zimbabwe, UN doc S/1998/891, 1998 UNYB 85-6; USA/Afghanistan, UN docs S/2001/946, S/2001/947.
362 Keesing’s (2014) 53953; Weekly Address: the World is United in the Fight against ISIL, 20 September 2014, <www.whitehouse.gov/the-press-office/2014/09/20/weekly-address-world-united-fight-against-isil>.
365 The USA and the Gulf States had all provided support for armed opposition forces in Syria aiming at the overthrow of the Assad government. The government was supported by Russia, Iran, and Iraq. See Chapter 3.
367 Russia accused the USA of double standards: the USA was helping the government of Iraq against Islamists at the same time that it was funding armed opposition groups against the government of Syria, UN doc S/PV 7271 (2014).
368 For example, the UK in its letter under Article 51 reported that it was taking measures in support of the collective self-defence of Iraq as part of international efforts led by the USA at the request of the government of Iraq (UN doc S/2014/851). But it did not at first take a direct part in military action in Syria, following an earlier vote in Parliament against such action: Keesing’s (2013) 52848. The UK nevertheless carried out a targeted killing in Syria in August 2015 and relied on a wide doctrine of individual self-defence to justify this. Its letter reporting its targeted killing it also said ‘ISIL is engaged in an ongoing armed attack against Iraq and therefore action against ISIL in Syria is lawful in the collective self-defence of Iraq’ (UN doc S/2015/688).
376 Although the member states of NATO invoked Article 5 on collective self-defence for the first time after the 9/11 attacks, the USA chose not to act through NATO or the UN. NATO did provide assistance to the USA’s military operations and it later led a UN authorized force (ISAF) in Afghanistan from 2003–14.
383 Keesing’s (2012) 52082. At the time Turkey said that ‘A short-term border violation can never be a pretext for an attack’, <www.bbc.co.uk/news/world-middle-east-18584872>. Later Turkey shot down several Syrian aircraft allegedly in its airspace. This attracted little public attention in the West. See ‘Syrian–Turkish border incidents during the Syrian civil war’ in Wikipedia.
393 Shaheem, ‘Turkey send tanks into Syria in operation aimed at Isis and Kurds’, The Guardian, 24 Aug 2016, <www.theguardian.com/world/2016/aug/24/turkey-launches-major-operation-against-isis-in-key-border-town>; Chulov, ‘Turkey and Russia sceptical of Trump’s plan to create safe havens in Syria’, The Guardian, 26 Jan 2017, <www.theguardian.com/world/2017/jan/26/turkey-russia-trump-syria-safe-zones>.
395 Keesing’s (2015) 64401, 54402. Turkey claimed that the plane was over Turkish territory for 17 seconds; Russia rejected this claim: UN docs S/2016/906, S/2015/962, S/2016/116. Turkey later expressed its regret to Russia for shooting down its plane: ‘Turkey moves to mend relations over downed Russian jet’, The Guardian, 28 June 2016. The two states re-established good relations, and at the end of 2016 they acted together to conclude a ceasefire in Syria to end the five-year conflict: see UN docs S/2016/1082, S/2016/1133. The Security Council passed SC Res 2336 (2016) unanimously, welcoming their efforts to end the conflict. Russia and Turkey also carried out joint air operations against ISIS in Syria: ‘Syria conflict: Russia and Turkey in first joint air strikes on IS’, <www.bbc.co.uk/news/world-middle-east-38667895>.
398 Reisman, ‘Allocating Competences to use Coercion in the post Cold-War World, Practises, Conditions, and Prospects’ in Damrosch and Scheffer (eds), Law and Force in the New International Order (1991), 26.
400 Farer, ‘Drawing the Right Line’, 81 AJIL (1987) 112; Higgins, ‘The Attitude of Western States towards Legal Aspects of the Use of Force’, in Cassese (ed), Current Legal Regulation of the Use of Force (1986); also Delivanis, La légitime défense en droit international public moderne (1971), Part II, Chapter 2.
404 See Chapter 5.
405 NATO struggled to establish a new role for itself after the end of the Cold War. After 9/11 it agreed at the Prague Summit (2002) on a New Military Concept on Defence against Terrorism, 42 ILM (2003) 244. In 2006 it adopted a Comprehensive Political Guidance setting out the direction for NATO’s continuing evolution. This said: ‘The character of potential Article 5 challenges is continuing to evolve. Large scale conventional aggression against the Alliance will continue to be highly unlikely; however, as shown by the terrorist attacks on the US in 2001 … future attacks may originate from outside the Euro-Atlantic area and involve unconventional forms of armed assault.’ More recently, following Russia’s intervention in Crimea and Ukraine, NATO has refocused on its traditional role of securing member states—now including many former Soviet bloc states—against threats from Russia. See NATO’s 2016 Warsaw Summit Communiqué, <www.nato.int/cps/en/natohq/official_texts_133169.htm>.