Self-defence

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 recent ICJ cases, Cameroon v Nigeria (2002),⁶ Iranian Oil Platforms (2003),⁷ and Armed Activities on the Territory of the Congo (DRC v Uganda) (2005).⁸ The issue is left unresolved in the vast majority of cases; 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 ceasesfire rather than any attribution of responsibility.9 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

The 1980–88 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 1980.¹¹ This willingness to identify the outbreak of a conflict and to determine responsibility was more common in the early days of the UN.¹² 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.13 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 ACADEMIC DEBATE

As far as writers are concerned, 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 fifty 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.14 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. 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.¹⁵ 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.

Those still supporting the wide right of self-defence today – in contexts outside the war against terror – discount the rejection of their position by the large majority of states in practice since 1945; for these writers the Charter preserves customary law as it allegedly was in 1945. Thus the term ‘inherent right of self-defence’ in Article 51 is not for them a dynamic term capable of shifting in meaning over time; the scope of the right was fixed in customary international law in 1945 and was apparently not susceptible of restriction in the light of subsequent state practice.¹⁶ An alternative approach invokes 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).¹⁷ Again this argument is at variance with the mass of state practice and has to discount the views of the vast majority of states.

In practice, states making their claims to self-defence 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.18 There was a similarly protracted discussion of the US 1983 intervention in Grenada and its 1986 bombing of Tripoli.¹⁹

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 International Court of Justice should not pronounce on claims of self-defence because Article 51 provides a role in such matters only for the Security Council.²⁰

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.21 Thus Combacau takes a rather formalistic approach. He argues 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 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.²² More recently, and exceptionally, Kuwait’s right to self-defence was affirmed by the Security Council after the Iraqi invasion.²³ And Resolution 1234 on the conflict in the DRC affirmed in general terms the right of individual or collective self-defence in accordance with Article 51.²⁴

Other resolutions respond to the use of force by states; in so far 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.25

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 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 have been 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’.²⁶ Judge Schwebel, in his Dissenting Opinion, strongly criticized this as unacceptable in the case of covert self-defence.²⁷ 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.28 And after the USSR intervention in Afghanistan 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.²⁹ 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.³⁰ However, it is clear that the reporting requirement is merely procedural; failure to comply does not of itself destroy a claim to self-defence.³¹

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. 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 51.³² 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.³³

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.34 After Nicaragua it can no longer be maintained that the reporting requirement is rarely observed.³⁵

Indeed, there is now 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, if relying on self-defence, go to the Security Council at the start of that conflict. However, when there is a prolonged conflict the states parties tend 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 interpret 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 may significantly increase the burden on the state claiming self-defence in that it has 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 conflict ³⁶ and in the practice of the UK and Argentina in the Falklands conflict.³⁷ Similarly in the 1998–2000 conflict between Eritrea and Ethiopia both parties repeatedly invoked self-defence.³⁸

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.³⁹ Here we see self-defence being invoked with reference to the protection of US ships and aircraft; this is sometimes referred to as ‘unit self-defence’ as opposed to ‘national self-defence’ of a state’s territory. 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.40

Such repeated reporting may seem to play partly a propaganda role.⁴¹ 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 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.⁴² 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.⁴³

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.44 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, since Argentina, the aggressor, remained in occupation of the islands.⁴⁵ The question came up 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.⁴⁶

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.⁴⁷

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 here, 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.⁴⁸ 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 51.⁴⁹ At the 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 United Nations 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.⁵⁰

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.

This question came up again with regard to Rwanda.51 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.⁵² 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.53 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’.⁵⁴ The claims by Liberia and Ethiopia that the embargoes were unlawful as they denied the states concerned the right to defend themselves were not successful.⁵⁵

THE SCOPE OF SELF-DEFENCE

Armed attack

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. The paradigm case is obviously an invasion by the regular armed forces of one state into the territory of another state. 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.⁵⁶ Thus the concept of armed attack by modern missiles⁵⁷ and naval mines⁵⁸ has given rise to special questions. Questions as to the regulation of cyber-attacks have also arisen in recent years.59 There are also 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 International Court of Justice has considered the concept of armed attack in a series of cases, starting with the Nicaragua case. In that case, Oil Platforms,⁶⁰ Armed Activities on the Territory of the Congo (DRC v Uganda)⁶¹ 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 Territory⁶² 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.⁶³

The concept of armed attack was central to the International Court of Justice’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, 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.64 However, the Court’s description of 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. 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.⁶⁵ 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 9/11.⁶⁶

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.⁶⁷ This 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.68 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 while denying potential victims what in some cases may be their only hope of survival.

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.⁶⁹ 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.’⁷⁰

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.⁷¹ 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. 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 Afghanistan, Yugoslavia, and Rwanda, has never identified such interventions as an armed attack.

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.72 But the support Uganda offered for this argument in its Pleadings was extremely weak and the Court did not change its view.⁷³

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.⁷⁴ 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