Jump to Content Jump to Main Navigation
EPIL

Self-Defence

Christopher Greenwood

Subject(s):
Self-defence — Armed attack — Collective security — Necessity — Conduct of hostilities — Military necessity

Published under the auspices of the Max Planck Foundation for International Peace and the Rule of Law under the direction of Rüdiger Wolfrum.

A.  Introduction

The right of a State to use force in self-defence is long-established in customary international law. Nevertheless, while it can be traced back at least as far as the correspondence between the United Kingdom and United States governments regarding the Caroline incident in 1837 (Caroline, The), its importance in the 19th century was limited by the fact that international law then recognized a general right of resort to war, so that self-defence was significant (at least in legal, as opposed to political, terms) only with regard to lesser instances of the use of force. It was not until there evolved in international law a general prohibition on recourse to force through the Covenant of the League of Nations (1919), the General Treaty for Renunciation of War as an Instrument of National Policy (Kellogg-Briand Pact [1928]), and the United Nations Charter (1945), that self-defence assumed its modern significance in international law (Use of Force, Prohibition of). Today, as Sir Humphrey Waldock explained, ‘there are few more important questions in international law than the proper limits of the right of self-defence’ (at 461).

The importance of self-defence in contemporary international law derives from its position as the principal exception to the general prohibition of the use of force enshrined in Art. 2 (4) UN Charter. That provision requires all Members of the United Nations to ‘refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any State, or in any other manner inconsistent with the purposes of the United Nations’. Chapter VII UN Charter then goes on to authorize the UN Security Council to take enforcement action, including military measures, should it determine that there is a threat to the peace (Peace, Threat to), breach of the peace (Peace, Breach of), or act of aggression. These provisions were not, however, intended to remove the right of States to use force in self-defence, at least until the UN Security Council had employed its collective security powers. That the right of self-defence was not impaired was taken for granted in the early drafts of the UN Charter prepared at the Dumbarton Oaks Conference (1944), which made no express mention of self-defence. A similar approach had been adopted when the Kellogg-Briand Pact was drawn up in 1928 and the international military tribunals at Nuremberg and Tokyo proceeded, in hearing charges of crimes against the peace, on the basis that a resort to force which fell within the customary law right of self-defence did not violate the provisions of the Pact.

This approach was not, however, considered satisfactory by a number of States, principally those from South America, which were party to collective self-defence agreements and which wanted express recognition of the fact that their right to take action under those agreements was compatible with the UN Charter (Collective Security). The result was Art. 51 UN Charter, which provided that:

[n]othing 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 this right of self-defence shall be immediately reported to the Security Council and shall not in any way affect the authority and responsibility of the Security Council under the present Charter to take at any time such action as it deems necessary in order to maintain or restore international peace and security.

Both the language and the drafting history of Art. 51 make clear that it does not create the right of self-defence but confirms that, within the limits set forth in Art. 51, the customary law right of self-defence is preserved. Moreover, certain features of the customary law right—particularly the requirement that action taken in self-defence must not exceed what is necessary and proportionate (see paras 25–29 below)—continue to apply, notwithstanding that they are not referred to in the text of Art. 51 UN Charter.

Self-defence is, in the language of the International Law Commission (ILC), a circumstance which precludes the wrongfulness of an act which would otherwise be illegal. As Art. 21 Draft Articles on Responsibility of States for Internationally Wrongful Acts puts it, ‘the wrongfulness of an act of a State is precluded if the act constitutes a lawful measure of self-defence taken in conformity with the Charter of the United Nations’ (State Responsibility). The most important consequence is that the resort to force by a State does not violate the prohibition stated in Art. 2 (4) UN Charter if that State is entitled to take action in self-defence and if the action which it takes remains within the limits of the right of self-defence. A State which is entitled to take action in self-defence but which exceeds what is necessary or proportionate will violate the prohibition on the use of force. Moreover, the fact that action is taken in self-defence does not preclude it from constituting a breach of certain rules of international law. The rules of the ius in bello, particularly those of international humanitarian law, apply equally to all parties to an armed conflict, irrespective of whether one or more of them is acting in self-defence. Accordingly, even if a State resorts to force for legitimate reasons of self-defence, the use of a weapon prohibited by international humanitarian law or the deliberate targeting of persons or objects protected by that law will remain unlawful.

Art. 51 UN Charter expressly preserves the right of both individual and collective self-defence. While the text of Art. 51 draws no distinction between them, subsequent practice and case-law have identified certain differences and they will be dealt with separately in this entry. The principal features of individual self-defence are considered in paras 7–34 below, while collective self-defence is considered in paras 35–40 below. The question of whether self-defence may be employed on an anticipatory or pre-emptive basis is discussed in paras 41–51 below.

Although there have been several instances of force being used on the basis of authorization granted by the Security Council pursuant to its collective security powers since the Iraq-Kuwait War (1990–91), and a number of instances of States asserting a right to use force by way of humanitarian intervention, self-defence has remained the legal justification for resort to force most frequently invoked by States.

B.  The Elements of Individual Self-Defence

Although Arts 2 (4) and 51 UN Charter are couched in terms of the rights and obligations of members of the UN, it is now generally accepted that, in almost all respects, they reflect rules of customary international law which are equally applicable to all States, irrespective of whether they are Members of the UN (Military and Paramilitary Activities in and against Nicaragua Case [Nicaragua v United States of America] ‘Nicaragua Case’ paras 187–201). An exception is the procedural provision in Art. 51 UN Charter, which requires that States report to the Security Council measures taken in self-defence. This provision was considered by the International Court of Justice (ICJ) as an innovation (see para. 31 below) which had not yet brought about a change in the customary international law on self-defence (Nicaragua Case paras 200 and 235).

It is generally considered that, for a resort to force to constitute a lawful exercise of the right of self-defence, it must meet the following conditions:

  1. i)  it must be a response to an armed attack;

  2. ii)  the use of force, and the degree of force used, must be necessary and proportionate; and

  3. iii)  it must be reported to the Security Council and must cease when the Security Council has taken ‘measures necessary to maintain international peace and security’.

Each of these requirements will be considered in turn.

1.  Armed Attack

Whatever may once have been the position in customary international law, it is now well established that the right to use force in self-defence exists only in response to an armed attack (though whether that armed attack must have commenced before resort to force in self-defence is undertaken is considered in paras 41–51 below). Self-defence does not give a State the right to resort to force in response to non-military actions and threats, such as economic coercion, no matter how damaging they may be to that State’s rights and interests; any response to such actions must also be of a non-military character.

10  The term ‘armed attack’ is not, however, defined in the UN Charter. Nor does it appear in the principal provisions with which Art. 51 is associated: Art. 2 (4) refers to the ‘threat or use of force’, while Art. 39 speaks of threats to the peace, breaches of the peace, and acts of aggression. The precise relationship of ‘armed attack’ with these other concepts is, therefore, far from easy to determine. Moreover, the French text of Art. 51 uses the term ‘agression armée’. Art. 39, in which the word ‘aggression’ appears in both English and French texts, treats aggression as more serious than a ‘breach of the peace’ and UN General Assembly Resolution 3314 (XXIX) ‘Definition of Aggression’ of 14 December 1974, proceeds on the same assumption. It follows that, if the term ‘agression armée’ in the French text of Art. 51 were to be treated as equivalent to ‘aggression’ in Art. 39, then the French text would point to a very restrictive concept of self-defence which would exclude self-defence in response even to large-scale uses of force. That does not appear, however, to have been the approach taken in international practice. The Iraqi invasion of Kuwait in 1990, for example, was characterized by the Security Council as a breach of the peace, rather than an act of aggression, in UN Security Council Resolution 660 of 2 August 1990 (SCOR 45th Year 19; a proposal to characterize it as an act of aggression having been rejected in consultations on the draft resolution). Nevertheless, the preamble to UN Security Council Resolution 661 of 6 August 1990 (SCOR 45th Year 19) expressly reaffirmed the right of self-defence of Kuwait in the face of that breach. It seems, therefore, that the term ‘agression armée’ is not in practice equated with ‘aggression’ in Art. 39 and should not be regarded as more restrictive than the English ‘armed attack’.

(a)  The Nature and Scale of an Armed Attack

11  An armed attack need not, however, necessarily take the form of action by regular armed forces. In its judgment in the Nicaragua Case, the ICJ held that:

an armed attack must be understood as including not merely action by regular armed forces across an international border, but also ‘the sending by or on behalf of a State of armed bands, groups, irregulars or mercenaries, which carry out acts of armed force against another State of such gravity as to amount to’ (inter alia) an actual armed attack conducted by regular forces, ‘or its substantial involvement therein’ (at para. 195; see also paras 16–18 below).

Similarly, the UN Security Council treated the terrorist attacks on targets in the United States on 11 September 2001, which took the form of crashing hijacked civil airliners into the World Trade Center in New York and the Pentagon, as an armed attack. (see UNSC Res 1368 [2001] [12 September 2001] SCOR [1 January 2001–31 July 2002] 290 and UNSC Res 1373 [2001] [28 September 2001] SCOR [1 January 2001–31 July 2002] 291)

12  In its judgment in the Nicaragua Case, the ICJ also stated that the use of force would not amount to an armed attack for the purposes of the right of self-defence unless it was of a particular scale and effect, and it contrasted an ‘armed attack’ with a ‘mere frontier incident’ (at para. 195). This approach opens up a gap between the use of force, which is prohibited by Art. 2 (4) UN Charter, and the armed attack which gives rise to a right to use force in self-defence. According to the Court, if a use of force did not rise above the level of a ‘mere frontier incident’, then, even though it constituted a violation of Art. 2 (4) UN Charter, the victim of that violation was not entitled to respond by way of action in self-defence, although it was entitled to take countermeasures, with the Court avoiding any clear stance on whether such countermeasures might involve the use of force. This conclusion has proved controversial with both States and commentators. While it is true that there are differences between the language of Arts 2 (4) and 51 UN Charter, there is no evident textual reason for holding that the concept of armed attack excludes some attacks which, though armed, fall below an unspecified threshold of intensity of violence. Nor was such a distinction apparent in State practice prior to the decision of the Court. There is an understandable concern not to open the door to excessive military action in response to minor incidents, but the requirement that action taken in self-defence must be necessary and proportionate already precludes such an excessive response from being brought within the ambit of self-defence. Moreover, the judgment in the Nicaragua Case gives little guidance regarding where the line is to be drawn between those uses of force which are sufficiently serious to constitute armed attacks and those which are not.

13  That dividing line is particularly important when considering whether terrorist attacks constitute an armed attack for the purposes of the law of self-defence. Some such attacks (eg those which took place on 11 September 2001) may be of such severity that they plainly cross the threshold established in the Nicaragua Case. In other cases, however, no one individual incident may be of sufficient gravity to cross that threshold, but if one looks at the totality of such incidents, then the picture changes. In such circumstances, it has often been suggested that the assessment of whether or not there is an armed attack has to be based on the series of terrorist incidents taken as a whole, rather than by examining each one in isolation.

14  In recent years a particular focus of debate has been the possibility of so-called ‘cyber-attacks’, in which computer technology is used to bring down a State’s computer systems, causing extensive economic and social harm (Cyber Warfare). It has sometimes been suggested that such action should be regarded as a modern form of armed attack which would give rise to a right, on the part of the victim State, to take military action in response. Such suggestions need to be treated with considerable caution. The planting of a virus or the use of other computer techniques to undermine, for example, the computer systems regulating a State’s financial system or immigration controls is difficult to see as an armed attack. Although the consequences of such conduct may be very serious, it seems closer to the concept of economic coercion. On the other hand, if such action were used to produce results similar to those which could otherwise be achieved only by the use of armed force, for example, causing aircraft to crash or dams to open and flood areas of a State’s territory, then the argument that such action should be treated as a form of armed attack is more plausible.

(b)  The Source of an Armed Attack

15  Controversy has also surrounded the question of whether an armed attack must be attributable to a State in order to give rise to a right on the part of the victim State to take action in self-defence. This question has assumed increased importance as a number of States have resorted to the use of force in response to terrorist attacks (Terrorism).

16  There is no doubt that a terrorist attack against a State which is carried out by the organs of another State, or for which that other State is in some other way responsible, is capable of amounting to an armed attack. The ICJ has stated, in its judgments in the Nicaragua Case and in the Armed Activities on the Territory of the Congo (Democratic Republic of the Congo v Uganda), as well as in Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory (Advisory Opinion), that such acts of terrorism amount to armed attacks provided that they are of sufficient intensity to cross the threshold identified in the Nicaragua Case and discussed in paras 11–14 above. On the other hand, the Court at times appears to suggest that acts of terrorism for which no State is responsible are not capable of constituting armed attacks, even if the level of violence is such as to cross the Nicaragua Case threshold (see, in particular, Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory [Advisory Opinion] para. 139; Armed Activities on the Territory of the Congo para. 146).

17  Nevertheless, the legal basis for such a limitation of the right of self-defence is unclear. The correspondence between the United Kingdom and United States Governments in the Caroline incident, which is generally treated as the earliest formulation of the right of self-defence in customary international law, runs counter to the existence of any such limitation. That correspondence proceeded on the basis that the attacks from United States territory against the British authorities in Canada, which were the basis for the claim to self-defence, were not the responsibility of the United States. Yet that fact was not seen by either government as excluding the possibility that the right of self-defence might have been triggered. Nor does any such limitation appear in the text of Art. 51 UN Charter, which refers to the need for there to be an armed attack against a State but makes no comment on the source of such an attack. It is true that the practice of States in relation to self-defence focussed, for most of the 20th century, on self-defence against armed attacks which emanated from States. However, that would not have sufficed to cut down the right of self-defence identified in the Caroline incident by limiting it to such a case. For such a limitation to have been grafted onto the right of self-defence, it would have been necessary for that practice to have reflected widespread opinio iuris that international law precluded resort to self-defence in response to other armed attacks. The response of the Security Council to the terrorist attacks against the United States on 11 September 2001 is also instructive. UN Security Council Resolutions 1368 of 12 September 2001 and 1373 of 28 September 2001 condemned the attacks and expressly reaffirmed the right of self-defence but made no mention of whether or not the attacks were in some way the responsibility of a State. In taking military action against Al Qaeda in Afghanistan shortly after the September 2001 attacks, the United States and those States allied with it claimed to be acting in self-defence. Their claim was not contested by the overwhelming majority of other States even though attributing the attacks by Al Qaeda to Afghanistan on 11 September was by no means established at the time.

18  Whether an armed attack, for the purposes of the right of self-defence, must be the responsibility of a State should therefore be regarded as unsettled. On the one hand, the increasing capacity of groups acting outside the responsibility of a State to engage in acts of extreme violence suggests that any such limitation would be an unreasonable restriction on the right of the victim to defend itself (see also Non-State Actors). Contemporary State practice supports the notion that no such broad restriction exists. On the other hand, there is an understandable concern that a State which has been the victim of an attack by a group unconnected with any other State should not inevitably be free to take action against that group in the territory of other States. It is possible, however, that this latter concern can be met by a proper application of the principle that action in self-defence must be limited to what is necessary and proportionate (see paras 25–29 below). If the State, in whose territory a group which has perpetrated a terrorist attack against another State is located, is prepared to take effective action against that group, then military action in that territory by the victim of the terrorist attack cannot be regarded as necessary. Only if the former State has shown itself to be unwilling (or, perhaps, unable) to act effectively against the group it can be said that military action in its territory in the exercise of the right of self-defence meets the criterion of necessity.

(c)  The Target of an Armed Attack

19  Art. 51 UN Charter preserves the right of self-defence ‘if an armed attack occurs against a Member of the United Nations’. For the reasons already given, the right of self-defence is applicable to any State, irrespective of whether it is a member of the UN. What is required, therefore, for a State to be entitled to use force in individual self-defence is that there is an armed attack against that State. That in turn raises the question of what is to be considered as constituting ‘the State’ for these purposes.

20  There is no dispute that the use of force (provided that it meets the criteria in the preceding sections) against the territory of a State is an armed attack against that State. That is so not only in the case of an attack upon the metropolitan territory of a State but also an attack upon islands and overseas territories which may be a considerable distance from the metropolitan territory. For example, most States treated the use of force by Argentina against the Falkland Islands/Islas Malvinas and Falkland Islands Dependencies (overseas territories of the United Kingdom) in 1982 as an armed attack against the United Kingdom.

21  It is also generally agreed that the use of force against the organs of a State outside the territory of that State constitutes an armed attack upon the State. For example, an attack upon units of the armed forces of a State lawfully stationed or operating in the territory of another State can constitute an armed attack upon the State of those forces (the sending State) as well as the State on whose territory the attack occurs (the host State; Military Forces Abroad). Similarly, an attack upon the naval forces or other State ships of a State lawfully operating on the high seas can constitute an armed attack upon that State.

22  A more difficult problem arises with attacks upon privately owned merchant ships flying the flag of a State or upon nationals of that State outside its territory (see also Flag of Ships). Neither are organs of the State and it has therefore been argued that the use of force against them is not an armed attack against the State and cannot, therefore, give rise to a right on the part of the State to take action in self-defence.

23  So far as merchant vessels are concerned, however, the practice of many maritime States has been to use force to protect merchant vessels flying their flag from attacks by the forces of other States and to justify such action by reference to the State’s right of self-defence. That right was frequently invoked in, eg, the Iran-Iraq War (1980–88). In the Oil Platforms Case (Iran v United States of America), the US contended that an attack on a US registered oil tanker, the Sea Isle City, entitled the US to exercise the right of self-defence by taking measures against Iran, whom it blamed for the attack. While the Court rejected that argument, it did so because it concluded that the United States had failed to prove that the attack was attributable to Iran. The judgment appears to accept that an attack by Iran on the Sea Isle City was capable of amounting to an armed attack upon the United States (at para. 64).

24  There is greater controversy surrounding the question whether attacks upon a State’s nationals abroad entitle the State to take action in self-defence. There is an understandable reluctance to see the kind of gunboat diplomacy practised in the 19th century revived under the auspices of self-defence. Nevertheless, the possession of a ‘population’ is one of the requirements of statehood and a case can certainly be made that an attack of sufficient violence upon a substantial number of a State’s nationals, especially where those nationals are selected as victims on account of their nationality and, in particular, where they are attacked in order to harm, or put pressure upon, their State of nationality, is a more serious assault upon the State than some forms of attack upon its territory. Thus the rescue of nationals abroad may well fall within the ambit of the right of self-defence, where the territorial State itself is unable or unwilling to act.

2.  Necessity and Proportionality

25  The fact that a State is entitled to take action in self-defence does not mean that it is entitled to employ unlimited force. It has long been an established principle of customary international law that force used in self-defence must not exceed what is necessary and proportionate. As the ICJ explained in the Nicaragua Case, there is a ‘specific rule whereby self-defence would warrant only measures which are proportional to the armed attack and necessary to respond to it’ (at para. 176). While this important limitation on the right of self-defence is not mentioned in Art. 51 UN Charter, the ICJ has recognized that it applies to measures of self-defence taken under that provision (Legality of the Threat or Use of Nuclear Weapons [Advisory Opinion] [1996] ICJ Rep 226 para. 41; Armed Activities on the Territory of the Congo para. 147).

26  The concepts of necessity and proportionality have often been equated in the literature on self-defence. While closely related, they are, however, separate requirements. The use of force, and the degree of force used, must be necessary to respond to the armed attack and must be proportionate to the threat posed by that attack. In its judgment in the Oil Platforms Case, the ICJ considered the two requirements separately.

27  Both concepts are forward-looking in the sense that compliance with them has to be assessed by reference to the goal which the State acting in self-defence is entitled to seek to achieve. Thus, a State which is the victim of an armed attack is entitled, inter alia, to halt and repel that attack and to recover territory occupied during the attack. It will satisfy the requirement of necessity only if it can demonstrate that it could not have achieved these goals without resort to force and that the degree of force employed did not exceed what was reasonably required for that purpose. It is important, however, to look at the totality of what the State acting in self-defence is entitled to achieve. The fact that the attacking State offers to cease its attack does not render the use of force unnecessary if, for example, the attacking State would thereby be left in occupation of part of the victim State’s territory.

28  In contrast to the concept of proportionality in the law of countermeasures or reprisals, proportionality in self-defence is also a forward-looking requirement. Whether action purportedly taken in self-defence meets the requirement of proportionality is to be assessed not by reference to the degree of force which was employed in the initial armed attack, but rather the threat posed by the armed attack. It is not simply a matter of comparing the number of forces or the types of weapons employed or even the scale of casualties and damage occasioned. For example, an armed attack which benefits from the element of surprise may succeed in gaining control of an area of territory or achieving another of the attacking State’s goals with the employment of comparatively little force. If the attacking State then rapidly reinforces its new positions, then the victim State will be unable to reverse the effects of the attack without employing a far greater degree of force. To prohibit it from doing so would be to reward the initial unlawful attack and place the victim State at a disadvantage which is wholly unreasonable. There is no indication that international law requires such an unjust conclusion. Whether the victim State’s use of force in self-defence meets the criterion of proportionality depends not upon its relation to the force initially used, but upon whether it is required in order to reverse the effects of the armed attack.

29  As the ICJ recognized in the Nuclear Weapons Advisory Opinions, the requirement that the force used in self-defence must be necessary and proportionate may have an impact upon which weapons and methods of warfare the State asserting the right of self-defence is entitled to employ (Warfare, Methods, and Means). It may also affect the geographical scope of operations and other aspects of the conduct of hostilities.

3.  Self-Defence and the Security Council

30  Although Art. 51 UN was designed to confirm that the right of self-defence, as it existed in customary international law at the time of the adoption of the UN Charter, was preserved in the era of the UN Charter, there are two aspects in which Art. 51 UN Charter set out to alter that right, at least for States who are members of the United Nations.

31  First, it introduced a requirement that measures taken in self-defence had to be reported to the Security Council. Although failure to comply with this requirement will probably not suffice to make a use of force which meets all the other requirements of self-defence a breach of Art. 2 (4) UN Charter, the fact that a State has not reported measures which it subsequently claims were taken in self-defence is likely to make that claim less plausible. Failure to comply with the reporting requirement is also, of course, a violation of the UN Charter in its own right.

32  Secondly, Art. 51 UN Charter introduced a temporal limitation on the right to take action in self-defence by providing that the right of self-defence endured only ‘until the Security Council has taken measures necessary to maintain international peace and security’. This requirement has assumed far greater significance since the Iraqi invasion of Kuwait heralded a marked increase in the use by the Security Council of its powers under Chapter VII UN Charter. That increase makes it necessary to consider what steps on the part of the Security Council will curtail the right of a victim State to take action in self-defence.

33  It is generally accepted that the Security Council has the power to impose, by means of a binding decision adopted under Chapter VII UN Charter, a ceasefire. If it does so, then all the parties to a conflict, whether or not acting in self-defence, will be obliged to comply. Nevertheless, it is plain that not every action taken by the Security Council during an armed conflict will have this effect. The reference to ‘measures’ suggests that the mere passage of a resolution, even one which contains a decision binding under the UN Charter, is not sufficient. The fact that UN Security Council Resolution 660 of 2 August 1990 condemned the Iraqi invasion of Kuwait and issued a binding demand for Iraq to withdraw from all Kuwaiti territory was not treated as removing from Kuwait (and its allies) the right to take action in self-defence when Iraq did not comply with that demand. Moreover, Art. 51 UN Charter refers to measures which are ‘necessary to maintain international peace and security’. Only when the Security Council has taken such action as is necessary will the right of self-defence lapse.

34  On a number of occasions, the Security Council has itself made clear that the action which it is taking is not intended to limit the right of self-defence (UNSC Res 661 [6 August 1990], which imposed economic and political sanctions upon Iraq for failing to withdraw from Kuwait contained such a provision). In such a case, there can be no doubt that the right of the victim State to take action in self-defence remains, notwithstanding the fact that the Security Council has taken measures. However, the absence of such a provision in a Security Council resolution should not be taken as meaning that the victim State necessarily loses its right to take action in self-defence. Careful analysis of the resolution in light of all the circumstances, including the nature of any continuing armed attack, the scale of the measures taken, and their likely efficacy seems called for.

C.  Collective Self-Defence

35  The right of collective self-defence attracted relatively little attention until the judgment of the ICJ in the Nicaragua Case. In that judgment, the Court identified three requirements which had to be satisfied if the use of force by a State which was not itself the victim of an armed attack was to be justified as collective self-defence.

36  First, the right of collective self-defence comes into operation only if there is at least one State which is entitled to take action by way of individual self-defence. Only if that State is the victim of an armed attack and has an entitlement to use force to defend itself are other States entitled, by way of collective self-defence, to resort to force to assist it. The Court thus rejected suggestions by some commentators that a group of States might be entitled to take action by way of collective self-defence even if no one State’s right of individual self-defence was engaged.

37  Second, the Court held that the victim State has to declare itself to be the victim of an armed attack before others become entitled to use force to assist it.

38  Third, the Court held that the right of a State which is not itself the victim of an armed attack to resort to force by way of collective self-defence is dependent upon the victim State having requested its assistance. In the aftermath of Iraq’s invasion of Kuwait in 1990, the Government of Kuwait issued a number of such requests and these were relied upon by other States intending to exercise a right of collective self-defence (until the adoption of UN Security Council Resolution 678 [1990] [29 November 1990] [SCOR 45th Year 27] provided a different legal basis for military operations derived from the authorization of the Security Council rather than self-defence).

39  Although these three requirements attracted some criticism at the time, the approach of the Court has generally been followed since the judgment was given in 1986. It is notable, however, that the Court did not require that there be any pre-existing alliance between the States engaged in collective self-defence. Nor did it adopt the suggestion advanced by a number of commentators (and supported by the Nicaragua Case [Dissenting Opinion of Judge Sir Robert Jennings]) that action in collective self-defence was permissible only if the armed attack upon the victim State also posed a threat to the security of the States which sought to rely upon collective self-defence.

40  It is important to distinguish between collective self-defence and various forms of collective security, not least because some organizations engage in both activities. For example, the North Atlantic Treaty Organization (NATO), though established as a means for the exercise of the right of collective self-defence, has assumed responsibilities of a collective security character since the end of the Cold War (1947–91), being used by the Security Council for the purpose of implementing certain collective security decisions. Moreover, an increasing number of modern military operations involve elements of both collective self-defence and collective security (operations in Afghanistan since 2001 are an example). The right of States (whether or not part of a standing alliance such as NATO) to use force by way of collective self-defence is derived from customary international law and is dependent upon the existence of a right to individual self-defence by a victim State which then requests their assistance. The legality of the use of force in a case of collective security is dependent not upon a request from a victim State but upon the authorization of the Security Council under Chapter VII or Chapter VIII UN Charter. Where a regional or other international organization employs force, its own constituent instrument may impose additional limitations upon its actions but it cannot empower the organization or its members to use force in circumstances where general international law does not permit such action or where no Security Council authorization exists.

D.  Anticipatory Self-Defence

41  Perhaps the most controversial question in relation to the right of self-defence is whether that right may be exercised only once an armed attack has begun or whether a State which is threatened with an armed attack may take action to forestall it. This question has given rise to an extensive literature about anticipatory self-defence, pre-emptive action, and other forms of preventative or interceptive measures. However, none of these terms has any precise definition and different commentators attribute different meanings to each of them.

42  Moreover, the uncertainty regarding the concept of ‘armed attack’ extends to marked differences of view as to when an armed attack may be deemed to have begun. It is generally accepted that an armed attack may occur before the victim sustains any casualties or suffers any damage—the contrary view confuses the attack with the effects of the attack; as a matter of logic, the former must precede the latter. An armed attack by air begins before the first bombs land on their targets. Yet there are wide differences of view as to how much earlier the attack can be said to begin. At one extreme it has been suggested that the attack by Japan on the United States in December 1941 began not when the first Japanese aircraft reached Pearl Harbour but when the Japanese fleet sailed from Japan with orders to launch the attack some three weeks earlier. The result is that some commentators who reject the notion of anticipatory or pre-emptive self-defence adopt an approach to the question of when an armed attack begins which means that their application of the law of self-defence to the facts of many cases is very close to that of the proponents of some variations of the anticipatory self-defence theory.

43  Neither State practice, nor jurisprudence, provides clear answers to most of these questions. In the Nicaragua Case, for example, the ICJ made clear that the question of anticipatory self-defence was not before it and that it would not enter into the debate as to whether international law recognized such a concept. Similarly, resolutions adopted by the UN General Assembly tend to avoid the question. The Friendly Relations Declaration (1970) is silent on the subject, while UN General Assembly Resolution 3314 (XXIX) ‘Definition of Aggression’ treats the fact that a State has been the first to resort to force as no more than prima facie evidence of aggression (at Art. 2), which seems to clearly leave open the possibility of a lawful first use of force. UN General Assembly Resolution 60/1 ‘2005 World Summit Outcome’ of 12 September 2005 (GAOR 60th Session Supp 49 vol 1, 3), which was adopted following the publication of the report of the High-Level Panel established by the UN Secretary-General, does not address the question, even though it was much discussed in the preceding debate.

44  Those who maintain that the right of self-defence applies only once an armed attack has commenced generally base their argument on the text of Art. 51 UN Charter, which confirms the continuing existence of the right of self-defence ‘if an armed attack occurs’. This choice of words, it is argued, excludes the possibility of self-defence against an attack which has yet to materialize.

45  There is, however, a strong case that international law still recognizes a right of anticipatory self-defence in circumstances in which an armed attack is imminent. The Caroline correspondence accepted that such a right was part of the customary international law of self-defence and its approach on this subject was reflected in the judgments of the Nuremberg and Tokyo international military tribunals. Since it was originally assumed that the right of self-defence would not be changed by the provisions of Art. 2 (4) UN Charter and Art. 51 UN Charter was introduced only at a later stage in the negotiations to confirm the continuing existence of the right of self-defence, it is plainly arguable that an important change in the ambit of the right of self-defence cannot be inferred from the use by Art. 51 UN Charter of the phrase ‘if an armed attack occurs’.

46  There are also numerous instances of anticipatory self-defence being advanced by States as the legal basis for military action. Debate in those cases tended to concern the question whether or not an armed attack had indeed been imminent when military action was taken to forestall it, whereas this question would be irrelevant if international law did not permit resort to force even in the face of an imminent armed attack.

47  In 2002 the US Government, in its National Security Strategy (a revised version of which appeared in 2005), appeared to assert that the concept of an imminent armed attack was too restrictive. Even in cases where the threat of attack was more remote, it was suggested, there was a right to take pre-emptive action. This theory has attracted considerable criticism. The United Kingdom, for example, which has long supported the existence of a right of anticipatory self-defence in the face of imminent armed attacks, rejected the notion of pre-emptive action unless an armed attack was imminent.

48  The High-level Panel on Threats, Challenges, and Change established by the UN Secretary-General took a similar approach in its report ‘A More Secure World: Our Shared Responsibility’, in which it stated that:

a threatened State, according to long established international law, can take military action as long as the threatened attack is imminent, no other means would deflect it and the action is proportionate (at para. 188).

49  In his report ‘In Larger Freedom: Towards Development, Security and Human Rights for All’, which addressed the recommendations of the Panel, the UN Secretary-General also assumed that the right of self-defence extended to cases in which an armed attack was imminent:

Imminent threats are fully covered by Article 51, which safeguards the inherent right of sovereign States to defend themselves against armed attack. Lawyers have long recognized that this covers an imminent attack as well as one that has already happened. Where threats are not imminent but latent, the Charter gives full authority to the Security Council to use military force, including preventively, to preserve international peace and security (at paras 124–25).

50  Nevertheless, the concept of anticipatory self-defence even in the face of imminent armed attack was criticized by many participants in the debate on ‘In Larger Freedom’, and the resolution 2005 World Summit Outcome adopted by the General Assembly in the wake of that debate does not address the subject.

51  If international law admits a right of self-defence in the face of an imminent armed attack, a further question is when is an attack to be considered imminent. The letter from the United States Secretary of State to the British Government in the Caroline dispute spoke of a threat which was ‘instant, overwhelming, leaving no choice of means and no moment for deliberation’. That statement was made, however, in the context of a scale of threat and means of delivery which were radically different from those of the 21st century. A recent Chatham House study has suggested that the concept of imminence can no longer be viewed only in temporal terms but must take account of the wider circumstances of the threat, including such factors as the gravity of the harm which would be inflicted, the capability of the party threatening the attack and the nature of the attack which is threatened. According to this study, ‘the criterion of imminence requires that it is believed that any further delay in countering the intended attack will result in the inability of the defending State effectively to defend itself against the attack’ and comments that ‘in this sense, necessity will determine imminence: it must be necessary to act before it is too late’ (Wilmshurst 968).

E.  Assessment

52  Although the subject of self-defence has given rise to controversy on a number of issues, particularly, precisely what constitutes an armed attack, when an armed attack begins, and whether a State may take action in the face of an imminent armed attack, the principle of self-defence itself is universally accepted and suggestions of a need for revision of Article 51 of the Charter have met with little support. It remains the principal exception to the general prohibition on the use of force in international relations.

Select Documents