A.  Introduction

1  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).

2  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.

4  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.

5  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.

6  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

7  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).

8  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’.

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.

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.