- Armed attack — Self-defence — Paramilitary groups — Weapons — Conduct of hostilities
Published under the auspices of the Max Planck Foundation for International Peace and the Rule of Law under the direction of Rüdiger Wolfrum.
1 Art. 51 UN Charter makes an ‘armed attack’ the condition for the exercise of the right of self-defence. In the judgment on the merits of the Military and Paramilitary Activities in and against Nicaragua Case (Nicaragua v United States of America) (‘Nicaragua Case’) the International Court of Justice (‘ICJ’) declared, however, that ‘a definition of “armed attack” which, if found to exist, authorizes the exercise of the “inherent right” of self-defence, is not provided in the United Nations Charter, and is not part of treaty law’ (Nicaragua Case para. 176). The statement was one link in the chain of arguments which the Court used to show that Art. 51 UN Charter was actually referring to pre-existing customary international law. Customary law thus determines the content of the term ‘armed attack’. While the Court stressed that ‘[t]he areas governed by the two sources of law [ie the Charter and customary law] do not overlap exactly, and the rules do not have the same content’ (ibid), it nevertheless confirmed in a later part of the judgment that ‘[t]here appears now to be general agreement on the nature of the acts which can be treated as constituting armed attacks’ (Nicaragua Case para. 195).
2 To illustrate the agreement, the Court cited the description, contained in Art. 3 para. (g) of the definition of aggression annexed to Resolution 3314 (XXIX), which the General Assembly of the UN had adopted in 1974. Given that the French text of Art. 51 UN Charter uses the term ‘agression armée’ for ‘armed attack’, some scholars argue that ‘armed attack’ is a subcategory of ‘aggression’. But this view is not shared by the majority opinion (Alexandrov 105–7, Constantinou 60–2). As, counter to the assumption of the Court, no consensus of opinion as to the acts which constitute armed attacks exists, States involved in the process of self-defence have to determine themselves whether an armed attack has occurred, at least until the Security Council takes measures necessary to restore peace (Alexandrov 98).
3 One must, however, take into account that both judgments in the Nicaragua Case (1984 and 1986) declared the law as the Court found it at the time of the judgments. That law may not be the same today. The cardinal point in the Court’s reasoning was the emphasis on the customary nature of the applicable international law, and custom is a dynamic body subject to modification by a change of opinio iuris confirmed by corresponding State practice. To determine the law as it now stands therefore requires a review of its application during the last decades and its possible development in that process.
4 The words ‘…if an armed attack occurs…’ suggest that they refer to an event which has just taken place or is presently taking place. It would, however, be unrealistic to expect a State which detects preparations of a large-scale attack to remain a sitting duck. But it is equally obvious that a State may feign the imminence of an attack to camouflage its own aggressive aims against another, or it may honestly mistake a military operation by a neighbouring state, eg manoeuvres near its border, as an impending attack where none is intended. On balance the majority opinion accepts nevertheless that a manifestly imminent armed attack which is objectively verifiable, ie an attack in progress, falls within the meaning of Art. 51 UN Charter (Constantinou 112–15; Duffy 156–7). The Japanese attack on Pearl Harbour is usually cited as an example where interceptive self-defence would be permissible (Dinstein  190–1). The lawfulness of this type of anticipatory self-defence was also endorsed by the UN High Level Panel on Threats, Challenges and Change in 2004 (at para. 188 self Defence, Anticipatory). As presently construed, the word ‘occur’ thus includes demonstrable imminence, but does not include the threat of only a possible attack at some point in the future (Bothe 229, 231–2).
B. Armed Attack by a State
5 Although the introductory words of Art. 51 UN Charter ‘…if an armed attack occurs’ do not reflect it, one may assume that the drafters had States and governments in mind as ‘attackers’ because, at that time, they alone had armed forces at their disposal which could launch an armed attack. The ICJ still subscribes to that view, as did or do some scholars (Brownlie 244–5), although it is hard to imagine that an ‘inherent’ right should only protect against a particular type of assailant (Schwebel 482).
6 In the decades following the founding of the United Nations, attacks by irregular forces became more frequent and the ICJ reacted to that development in its 1986 Nicaragua Case judgment. Borrowing from the law of state responsibility and from the definition of aggression, it broadened the term ‘attacker’ by including ‘the sending of or on behalf of a State of armed bands, groups, irregulars or mercenaries’ in the list of acts that could constitute an armed attack in the sense of Art. 51 UN Charter (Nicaragua Case para. 195). According to the Court, such an attack is imputable to the ‘sending’ State if the latter has ‘effective’ control of it (Nicaragua Case paras 109, 115). This view was disputed by the Appeals Chamber of the ICTY in the Tadić Case, which considered ‘overall’ control sufficient (Prosecutor v Tadić at para. 145). However, when the occasion arose in the judgment on the Application of the Convention on the Prevention and Punishment of the Crime of Genocide Case (Bosnia and Herzegovina v Serbia and Montenegro), the ICJ reaffirmed its earlier opinion (at paras 402–7). Yet it does not look as if that has put an end to the controversy (see A Cassese ‘The Nicaragua and Tadić Test Revisited in the Light of the ICJ Judgment on Genocide in Bosnia’ (2007) 18 EJIL 649–68).
7 The judgment in the Nicaragua Case focused on the ‘scale and effects’ (Nicaragua Case at para. 195) of an armed attack to distinguish it from a ‘mere frontier incident’ (ibid). The criteria seem adequate for the subsequent evaluation of an armed attack in judicial proceedings, but they are less so as operational guidance for the attacked State at the onset of the attack, when it may not be easy to recognize what is what. But the thrust of the Court’s thinking is clear: an isolated minor incident which, by the manner in which it takes place, cannot be mistaken for a threat to the safety of the State would not qualify as armed attack under Art. 51 UN Charter. However, with more empirical data available, the Court reflected decades later in the Oil Platforms Case (Iran v United States of America), on whether a series of minor attacks could cumulatively be considered an armed attack in the sense of Art. 51 UN Charter (Oil Platforms Case at para. 64). Although it did not find so in the case, it is significant that it entertained the idea not with standing inconclusive academic opinion (Wandscher 170–2), equivocal State practice (Ruys 174), and the Security Council’s longstanding refusal to accept the ‘accumulation of events’ theory (Oellers-Frahm 510).
8 The judgment in the Nicaragua Case added still another aspect to the evaluation by distinguishing ‘the most grave forms of the use of force (those constituting an armed attack) from other less grave forms’ (at para. 191). Dinstein commenting thereon has justly observed that the Court had failed to indicate what threshold must be reached for the use of force to qualify as an armed attack and concluded: ‘There is certainly no cause to remove small-scale armed attacks from the spectrum of armed attacks’ (Dinstein  195). It is, moreover, unclear whether the distinction is not just another expression of the ‘scale and effects’ criterion mentioned above. A passage in the Oil Platforms Case suggests so, since in that case the Court found that armed attacks of a lesser gravity, even when made by the armed forces of a State, did not justify self-defence (Oil Platforms Case at para. 51). That makes the Court’s idea of the proper response to an attack which takes one of the ‘less grave forms’ of the use of force something of an enigma. The answer in the Nicaragua Case judgment that the use of force, although not permitted as self-defence, may be permitted as countermeasure may appear as artificial distinction but is significant because it excludes the possibility of collective action (Nicaragua Case para. 210). Inexplicably, however, the Court undermined its own argument by adding the suggestion that forceful third-party assistance may be justified ‘in exercise of some [sic] right analogous to the right of self-defence’ (ibid). Implicitly, the Court’s suggestion thus recognized as legal the use of force outside the Charter system (Oellers-Frahm 508).
9 The judgment in the Oil Platforms Case named yet another requirement for an attack to qualify as armed attack in the sense of Art. 51 UN Charter. The Court found that the attack must be undertaken with the ‘specific intention of harming’ (Oil Platforms Case para. 64). By invoking the ‘intention’ of the attacker, the Court has apparently borrowed from domestic criminal law. But the motive of an attacker can only be established in judicial proceedings. Taking the intention into account is thus suitable in a domestic legal system where such proceedings are mandatory. The occasional judicial proceedings of the current international system do not warrant making harmful intentions a condition of unlawfulness in the application of a norm which usually has to be applied by States without the aid of a court. How should the victim discover the motive of the attacker in such circumstances? Unless clear indications point to the contrary (Ruys 167), one would assume that the attack itself was unrebuttable proof of harmful intentions (Constantinou 62).
10 Constantinou (at 64) has comprehensively summarized the elements which satisfy the conditions of ‘scale and effect’. She states: ‘…armed attack implies an act or the beginning of a series of acts of armed force of considerable magnitude and intensity (ie scale) which have as their consequence (ie effects) the infliction of substantial destruction upon important elements of the target State namely, upon its people, economic and security infrastructure, destruction of aspects of its governmental authority, ie its political independence, as well as damage to or deprivation of its physical element namely, its territory’ (ibid at 63–4) and furthermore adds the ‘use of force which is aimed at a State’s main industrial and economic resource and which results in the substantial impairment of its economy…’ (ibid at 64). In sum, it is submitted that regardless of the dispute over degrees in the use of force, or over the quantifiability of victims and damage, or over harmful intentions, an armed attack even when it consists of a single incident, which leads to a considerable loss of life and extensive destruction of property, is of sufficient gravity to be considered an ‘armed attack’ in the sense of Art. 51 UN Charter.
11 The state of knowledge and experience in 1945 suggests that the drafters of Art. 51 UN Charter thought of ‘armed’ attack in conventional military terms, meaning use of the standard weaponry of World War II, with the probable addition of the German V-rockets and the nuclear bombs. That view had not changed by 1956, as the proceedings of the UN Special Committee on the Question of Defining Aggression, which inter alia considered defining ‘armed attack’. It was only in 1961 that Brownlie proposed the additional consideration of bacteriological, biological, and chemical weapons since they ‘are employed for the destruction of life and property’ (Brownlie 255–6; Biological Weapons and Warfare; Chemical Weapons and Warfare).
12 The continuing development in weapons technology makes it increasingly difficult to determine when an armed attack begins or when it can be considered as being manifestly imminent (Gray 108). It is doubtful whether the present concept of demonstrable imminence, and hence that of permissible interception, is adequate for an attack with Weapons of Mass Destruction (‘WMD’), especially when delivered with missiles, because that attack has a distinct quality: it may cripple the State as a whole and annihilate its capacity to defend itself if it is not blocked in time. The reluctance of the ICJ in the Nuclear Weapons Advisory Opinion to pronounce definitely on the legality of the use of nuclear weapons in self-defence shows how ambivalent the current legal opinion is on the use of, or the defence against, WMD (Legality of the Threat or Use of Nuclear Weapons [Advisory Opinion]  ICJ Rep 226 para. 97; Nuclear Weapons Advisory Opinions).
13 Recent events demonstrated the possibility of yet another manner of attack. Computer network attacks (‘CNAs’) are actions taken through the use of computer networks to disrupt, degrade, or destroy information resident in computers or computer networks, or the computers and networks themselves (Harrison Dinniss 4). Viruses, worms, Trojans and similar devices destroy or alter data and programmes, while denial of service attacks (‘DoS’) flood an internet site, a server, or a router with more requests for data than it can process to shut it down. A CNA can either be undertaken in conjunction with a conventional (kinetic) armed attack or as self-contained end in itself. Provided that the autonomous CNA has the ‘scale and effect’ which the ICJ named as condition of an armed attack (see para. 7), the mainstream ‘effect based approach’ (Roscini 129), which draws on the conclusions of the debate after 9/11 concerning the definition of ‘arm’ (see para. 21), considers it an ‘armed attack’ in the sense of Art. 51 UN Charter. The supplementary question, whether a device that causes damage only indirectly (eg by shutting down the computer system controlling the electric grid of an area) can initiate an ‘armed attack’, has already been settled in connection with bacteriological, biological, or chemical weapons (see para. 11). However, since the probability of reliably identifying the source of a CNA with the currently available technology is small, it may be doubtful against whom a right of self-defence exists. Opinions are, moreover, divided over the permissible defence against an autonomous CNA, some advocating a limitation to measures in kind (Roscini 120), while others consider a response with conventional weapons legal (Schmitt 928).
C. Armed Attack by a Non-State Actor
14 Since the attack of ‘9/11’ the question whether a terrorist attack of that magnitude qualifies as ‘armed attack’ in the sense of Art. 51 UN Charter has become the subject of debate. The ICJ maintains the position expressed in the Nicaragua Case judgment, that only acts attributable to a State can constitute an ‘armed attack’. It has reiterated its view in the Israeli Wall Advisory Opinion (Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory) (at para. 139) and in the Armed Activities on the Territory of the Congo Case (Democratic Republic of Congo v Uganda) (at paras 146 and 160; Armed Activities on the Territory of the Congo Cases). But, as Ruys (473) observes, ‘it did so in a rather confusing manner that is bound to exacerbate legal uncertainty’. Hence it is not surprising that some judges challenged the majority view and appended declarations or separate opinions to the Israeli Wall Advisory Opinion (Judges Buergenthal, Higgins, and Kooijmans) and the judgment in the Congo v Uganda case (Judges Kooijmans and Simma). Judge Simma remarked on that occasion: ‘Such a restrictive reading of Article 51 might well have reflected the state, or rather the prevailing interpretation, of the international law on self-defence for a long time. However, in the light of more recent developments not only in State practice but also with regard to accompanying opiniojuris, it ought urgently to be reconsidered, also by the Court’ (Congo v Uganda [Separate Opinion Judge Simma] para. 11).
15 The ICJ’s position disregards resolutions adopted by the Security Council after 9/11. In a letter by the US Permanent Representative to the Security Council the United States had justified its reaction to the attack with the exercise of its right to self-defence (UNSC ‘Letter Dated 7 October 2001 from the Permanent Representative of the United States of America to the United Nations Addressed to the President of the Security Council’ [7 October 2001] UN Doc S/2001/1946). The Security Council expressly referred to that right in the preambles of the relevant resolutions 1368 (2001) and 1373 (2001). It can be argued that the Council thereby accepted the claim of the United States and implicitly recognized the terrorist attack as ‘armed attack’ in the sense of Art. 51 UN Charter (Congo v Uganda [Separate Opinion Judge Simma] para. 11). This view was challenged by academics who maintain that the resolutions ‘do not provide a clear indication whether they intend to refer to a wide concept of armed attack which would comprise also acts which are not attributable to a State’. It was also suggested that the terrorist acts of 9/11 should be regarded rather as conventional crimes than as armed attack. Other scholars retorted that the significant scale and effect of the attack, as well as the subsequent reactions of the international community, make a credible case for an armed attack (Gray 164–5). The admission that ‘in fact, the incidents can properly be characterized as both a criminal act and an armed attack’ (Murphy 49) expresses the ambiguity of the unsettled legal situation.
16 The analysis of the text of Art. 51 UN Charter does not clarify the issue. It is obvious that the words ‘…if an armed attack occurs…’ do not indicate the nature of the attacker. Scholarly examinations have found no evidence in the legislative history of Art. 51 UN Charter (Scholz 101), or in the practice of States subsequent to 9/11 (Bruha and Tams 95–9), that supports the Court’s majority view. Their findings rather tend to support the opposite conclusion: nothing suggests that an armed attack can only be launched by a State or that the right of self-defence would, consequently, only be available against inter-State attacks. The majority view of the Court is rather the consequence of the specific conceptual construction of international law as law between States and some entities created by them, a conception which is still shared by many, but increasingly criticized by others.
17 Treating international terrorists as initiators of an ‘armed attack’ in the sense of Art. 51 UN Charter is, however, complicated by their lack of international personality and of territory in the sense of international law. No generally agreed definition of international terrorism and its actors exist (Duffy 18–46). Terrorist acts which occurred prior to 9/11 were not connected with Art. 51 UN Charter. This is explained by the Court’s construction of Art. 51 UN Charter which requires the imputability of the attack to a foreign State. Acts of ‘indigenous terrorists’, based and operating on the territory against which their acts are directed, even when that territory is only occupied, are not attributable to a foreign source. As the Court observed in the Israeli Wall Advisory Opinion of Palestinian terrorist acts against Israel: ‘…Israel does not claim that the attacks against it are imputable to a foreign State’ (at para. 139). As for attacks by ‘transboundary terrorists’, based in neighbouring States and operating therefrom, they may sometimes show elements of an armed attack, but did not, in the past, amount to the scale required by the ICJ.
18 The situation changed fundamentally with 9/11. Never before had an operation of that magnitude been planned and carried out by a globally acting terrorist organization, which had ‘sleeping’ members in many countries and could call them together for a specific operation that was directed by headquarters of unknown locality, communicating with its members through the internet. The little that is known of internationally active terrorists suggests that they operate as a loose network of cells with a horizontal rather than a vertical structure. They have no territory of their own, but operate from the territory of one or several States. Whether the ‘host’ State thereby violates its duty under international law which the ICJ had characterized in the Corfu Channel Case as ‘every State’s obligation not to allow knowingly its territory to be used for acts contrary to the rights of other States’ (Corfu Channel Case 22), a statement which was recently reiterated in the case Congo v Uganda (at para. 162), is a matter to be determined eventually under the law of State responsibility (Dinstein  206). Sometimes, when the ‘host’ State was unwilling or unable to prevent cross-border attacks, the victim treated them as ‘armed attack’ in the sense of Art. 51 of the Charter warranting measures of self-defence (Bothe 233), but the relevant State practice is inconsistent (Ruys 455, 486–7). It excludes, at any rate, purely preventive measures (Dinstein  208; Self-Defence, Pre-Emptive).
19 9/11, taken as a paradigmatic case of a terrorist ‘armed attack’, unquestionably had the magnitude and intensity, and hence the significant scale, required by the ICJ’s Nicaragua Case judgment for qualifying as armed attack in the sense of Art. 51 UN Charter. But it also had distinctive features of its own. While there is no doubt about the substantive destruction of civilians and civilian objects, the purpose of destruction and the manner in which it was carried out make 9/11 different from an inter-State attack.
20 A State attacks in pursuit of a strategic, political, or economic aim which it wishes to achieve through the attack. A terrorist attack has the ideological aim to create terror through damage and through demonstrating the vulnerability of the object-State, but as a physical act it is an end in itself. If it is not foiled before it takes place, defensive action can only deflect it by chance. Terrorists taking part in the attack may be prepared or even determined to lose their own life. Since the attack runs its course with the destruction, self-defence strictosensu is not possible. If the attacked State attempts to eliminate the persons or organizations behind the attack, it is an act of self-help rather than of self-defence.
21 The criteria for arms which would qualify a terrorist attack as ‘armed’ in the sense of Art. 51 UN Charter are not different from those applying to inter-State attacks. But what about highjacked civil airliners? They were used in the attacks of 9/11 for the destruction of human life and property, and had the same effect as bombs or missiles had they been used, which are conventional weapons. Hence it is submitted that it is neither the designation of a device, nor its normal use, which make it a weapon but the intent with which it is used and its effect. The use of any device, or number of devices, which results in a considerable loss of life and/or extensive destruction of property must therefore be deemed to fulfil the conditions of an ‘armed’ attack, and that includes CNAs.
22 As far as armed attacks attributable to States are concerned, the interpretation of Art. 51 UN Charter has been thoroughly discussed in academic writings and was further clarified in judgments and opinions of the ICJ. It may only need refinement to make the distinction between attacks entitling self-defence, and others causing State responsibility, more transparent (see para. 8). More intensive consideration must, however, be given to cover WMD and computer network attacks as instruments of attack, since the technical progress increases the likelihood of their use in the future (see paras 12, 13).
23 The legal evaluation of attacks by non-State actors is still controversial. Even if a large-scale spectacular assault like 9/11 has not happened again, it may in the future, for which reason achieving a consensus of opinion on the matter is clearly indicated. The most significant obstacle in the endeavour is the interpretation of Art. 51 UN Charter presently preferred by the majority of the judges of the ICJ (see para. 14).
- I Brownlie ‘The Use of Force in Self-Defence’ (1961) 37 BYIL 183–268.
- SM Schwebel ‘Aggression, Intervention and Self-Defence in Modern International Law’ (1972) 136 RdC 411–98.
- SA Alexandrov Self-Defence Against the Use of Force in International Law (Kluwer The Hague 1996).
- MN Schmitt ‘Computer Network Attacks and the Use of Force in International Law: Thoughts on a Normative Framework’ (1999) 37 Columbia JTransnational Law 885–937.
- A Constantinou The Right of Self-Defence under Customary International Law and Article 51 of the UN Charter (Sakkoulas Athènes 2000).
- Y Dinstein ‘Computer Network Attacks and Self-Defence’ in MN Schmitt and BT O’Donnell (eds) Computer Network Attack and International Law (Naval War College Newport 2002) 96–105.
- SD Murphy ‘Terrorism and the Concept of “Armed Attack” in Article 51 of the UN Charter’ (2002) 43 HarvardILJ 41–52.
- M Bothe ‘Terrorism and the Legality of Pre-emptive Force’ (2003) 14 EJIL 227–40.
- C Gray International Law and the Use of Force (OUP Oxford 2nd edn 2004)
- T Bruha and CJ Tams ‘Self-Defence Against Terrorist Attacks: Considerations in the Light of the ICJ’s “Israeli Wall” Opinion’ in K Dicke and others (eds) Weltinnenrecht: Liber amicorum Jost Delbrück (Duncker & Humblot Berlin 2005) 85–100.
- Y Dinstein War, Aggression and Self-Defence (4th edn CUP Cambridge 2005).
- H Duffy The ‘War on Terror’ and the Framework of International Law (CUP Cambridge 2005).
- K Oellers-Frahm ‘The International Court of Justice and Article 51 of the Charter’ in K Dicke and others (eds) Weltinnenrecht: Liber amicorum Jost Delbrück (Duncker & Humblot Berlin 2005) 503–18.
- M Scholz Staatliches Selbstverteidigungsrecht gegen terroristische Gewalt (Duncker & Humblot Berlin 2006).
- C Wandscher Internationaler Terrorismus und Selbstverteidigungsrecht (Duncker & Humblot Berlin 2006).
- M Roscini ‘World Wide Warfare: Jus ad Bellum and the Use of Cyber Force’ (2010) 14 Max Planck YUNL 85–130.
- T Ruys ‘Armed Attack’ and Article 51 of the UN Charter: Evolutions in Customary Law and Practice (CUP Cambridge 2010).
- K Zemanek ‘Responses to a Terrorist Attack—A Clarification of Issues’ (2010) 15 ARIEL 199–226.
- H Harrison Dinniss Cyber Warfare and the Laws of War (CUP Cambridge 2013).
- Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v Serbia and Montenegro) (ICJ 26 February 2007).
- Armed Activities on the Territory of the Congo (Democratic Republic of the Congo v Uganda) (Judgment)  ICJ Rep 168.
- Corfu Channel (United Kingdom and Northern Ireland v Albania) (Merits)  ICJ Reports 4.
- Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory (Advisory Opinion)  ICJ Rep 136.
- Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America) (Jurisdiction and Admissibility)  ICJ Rep 392.
- Military and Paramilitary Activities in and against Nicaragua (Nicaragua v United States of America) (Merits)  ICJ Rep 14.
- Oil Platforms (Islamic Republic of Iran v United States of America)  ICJ Rep 161.
- Prosecutor v Tadić (Appeals Judgment) ICTY-94-1 (15 July 1999).
- UNGA ‘A More Secure World: Our Shared Responsibility, Report of the High-Level Panel on Threats, Challenges and Change’ in UNGA ‘Note by the Secretary-General’ (2 December 2004) UN Doc A/59/565, 8.
- UNGA ‘Report of the Special Committee on the Question of Defining Aggression’ UN GAOR 12th Session Supp 16 (1957) A/3574.
- UNGA Res 3314 (XXIX) ‘Definition of Aggression’ (14 December 1974) GAOR 29th Session Supp 31 vol 1, 142.
- UNSC Res 1368 (2001) (12 September 2001) SCOR [1 January 2001–31 July 2002] 290.
- UNSC Res 1373 (2001) (28 September 2001) SCOR [1 January 2001–31 July 2002] 291.