We know now that the uncertainty surrounding the ‘origin’ of the concept of self- defence in Article 51 is largely irrelevant for its scope. The focus of recent debates on self-defence under the Charter has therefore been on the sixteenth to twentieth words of Article 51, which read: ‘if an armed attack occurs’. This section will discuss whether an ‘armed attack’ needs to be in some sense ‘present’ for a threat or use of force to be justifiable as self-defence, i.e. whether an ‘armed attack’ is a necessary condition for self-defence. Later sections will discuss what is meant by
‘armed attack’ (Sections 2.2.2 and 2.2.3) and ‘presence’ (Section 2.2.4).
The question is narrow and must be distinguished from the legality of anticipa- tory, preventive or pre-emptive self-defence. Some writers base their arguments in favour of these modes of self-defence on the absence of an ‘armed attack condi- tion’, but most of the recent proponents of such kinds of doctrines do acknow- ledge the conditionality. The converse argument, namely that the armed attack conditionality necessitates negating the possibility of anticipatory or pre-emptive self-defence, is also made,27 but this conclusion is not unassailable, as Section 2.2.4 will show.
The debate is familiar to any student of the modern ius ad bellum, but it may be beneficial to review the most important arguments and to ascertain their sound- ness. Most writers believe that an armed attack is required.28 The opposite view is
27 For example by Ian Brownlie. This argument does not detract from the strength or weakness of the de-coupling of the ‘presence’ requirement and the armed attack conditionality.
28 Ago (1982) supra note 9 at 65, 67 (para 114); Alexandrov (1996) supra note 20 at 95; Brownlie (1962) supra note 20 at 242; Brun-Otto Bryde, Self-defence, in: Rudolf Bernhardt (ed.), Encyclopedia of public international law (2000) Volume 4, 361–364 at 362; Georg Dahm, Das Verbot der Gewaltanwendung nach Art. 2(4) der UNO-Charta und die Selbsthilfe gegenỹber Vửlker- rechtsverletzungen, die keinen bewaffneten Angriff enthalten, 11 Jahrbuch für Internationales Recht (1962) [Festschrift für Rudolf Laun zu seinem achtzigsten Geburtstag] 48–72 at 51; Dinstein (2005) supra note 9 at 182; Michael Donner, Die Begrenzung bewaffneter Konflikte durch das moderne jus ad bellum, 33 Archiv des Vửlkerrechts (1995) 168–218 at 180; Terry D. Gill, The law of armed attack in the context of the Nicaragua case, 1 Hague Yearbook of International Law (1988) 30–58 at 35; Gray (2008) supra note 6 at 117–118, 128; Kaiyan Homi Kaikobad, Self- defence, enforcement action and the Gulf wars 1980–88 and 1990–91, 63 British Year Book of International Law 1992 (1993) 299–366 at 304–305 (‘self-defence is predicated upon and is the logical outcome of an armed attack’); Hans Kelsen, Collective security and collective self-defense under the Charter of the United Nations, 42 American Journal of International Law (1948) 783–796 at 791; Hans Kelsen, The law of the United Nations. A critical analysis of its funda- mental problems (1950) 797–798; Klaus Kersting, ‘Act of aggression’ und ‘armed attack’.
Anmerkungen zur Aggressionsdefinition der UN, 23 Neue Zeitschrift für Wehrrecht (1981) 130–143 at 136; Friedrich Klein, Der Begriff des ‘Angriffs’ in der UN-Satzung, in: Karl Carstens, Hans Peters (eds), Festschrift Herrmann Jahrreiss. Zu seinem siebzigsten Geburtstag – 19. August 1964 – gewidmet (1964) 163–188; Kunz (1947) supra note 19 at 877; O’Connell (2002) supra note 24 at 6; Albrecht Randelzhofer, Article 51, in: Bruno Simma (ed.), The Charter of the United Nations. A commentary (2nd ed. 2002) 788–806 at 793; Carsten Stahn, ‘Nicaragua is dead, long
Uncertainty in International Law 12
held by a minority of scholars in relatively dated publications29 and all opponents base their conclusion on the ‘black hole’ theory. There are additional arguments which we will discuss below, but in every case the linchpin of the argumentative train of thought is the ‘cut-out’ of Article 51’s terms. The reason why this theory is utilised will become clear during our review of the linguistic weapons employed in this battle.
The exchange of fire begins with the assertion that while Article 51 acknow- ledges the right to individual or collective self-defence ‘if an armed attack occurs’, that does not mean that ‘it must therefore follow that self-defence is valid only against an armed attack’.30 The thrust of this charge seems to be that it is a logical fallacy to conclude from the assertion of one possibility that other possibilities are excluded; in this case that the assertion of a right to self-defence in case of an armed attack cannot be interpreted to mean that other conditions may not justify self-defence.31 Stephen Schwebel demands of the Drafters that they should have written ‘if, and only if, an armed attack occurs’ if they had wanted to make it a necessary condition.32 According to him, the word ‘if’ does not indicate a neces- sary condition (and, perhaps, a logical equivalent),33 but rather one sufficient condi- tion34 for the lawful exercise of the right of self-defence. One might be able to reply to this argument that it seems odd that the Drafters would have included in the Charter only the most obvious case – the highest level of infringement – rather than define a threshold for the lawful exercise of defence.35 Is it not more likely that the Drafters, assuming arguendo that they did intend to make ‘armed attack’ a necessary condition, would have used the confident ‘if’, rather than the overly cautious ‘if, and only if’ one would perhaps use as a lawyer drawing up a private contract? Third, it might be ventured to argue that the assertion: ‘to
live Nicaragua’ – the right to self-defence under Art. 51 UN Charter and international terrorism, in: Christian Walter et al. (eds), Terrorism as a challenge for national and international law:
Security versus liberty? (2004) 827–877 at 840–841; Robert F. Teplitz, Taking assassination attempts seriously: did the United States violate international law in forcefully responding to the Iraqi plot to kill George Bush?, 28 Cornell International Law Journal (1995) 569–617 at 580–581;
Alfred Verdross, Bruno Simma, Universelles Vửlkerrecht. Theorie und Praxis (3rd ed. 1984) 288;
Wilhelm Wengler, Das vửlkerrechtliche Gewaltverbot (1967) 4; Luzius Wildhaber, Gewaltverbot und Selbstverteidigung, in: Wilfried Schaumann (ed.), Vửlkerrechtliches Gewaltverbot und Friedenssicherung (1971) 147–173 at 153. See also: Armed Activities (2005) supra note 26 at 223 (para 148).
29 Bowett (1958) supra note 11 at 187–188; Franzke (1965) supra note 12 at 133–134; McCormack (1991) supra note 12 at 35; McDougal and Feliciano (1961) supra note 12 at 232–241; Edward Miller, Self-defence, international law, and the six day war, 20 Israel Law Review (1985) 49–73 at 66; Schwebel (1973) supra note 3 at 479–480; Waldock (1953) supra note 12 at 496–498.
30 Bowett (1958) supra note 11 at 188 (emphasis added).
31 Franzke (1965) supra note 12 at 133–134; McDougal and Feliciano (1961) supra note 12 at 232;
Schwebel (1973) supra note 3 at 479–481; Waldock (1953) supra note 12 at 495.
32 Schwebel (1973) supra note 3 at 480; Nicaragua (1986) supra note 5, Dissenting Opinion Schwebel at 347–348.
33 Rather than merely A⇒B, this would mean that A≡B.
34 A→B.
35 Dinstein (2005) supra note 9 at 183–185.
2.2.1 Self-defence 13
construe Article 51 as containing a necessary condition would be a logical flaw’ is itself based on a logical flaw, because the inclusion of only one (sufficient) condi- tion in the text does not imply that there are other sufficient conditions. Indeed, the absence of such other conditions in a treaty means that, as a matter of treaty law, one condition is the only condition able to fulfil the requirement. Hence on this train of thought, the one condition is a necessary condition, evoking the Latin phrase expressio unius exclusio alterius est.
Edward Miller poses the question ‘whether the word “if” has a meaning of condition or hypothesis’.36 He erroneously believes that the French version of Article 51 uses the phrase ‘dans un cas ó’,37 concluding that ‘[t]he use here of the indefinite article is clearly suggestive of hypothesis.’38 The French version of Article 51, however, employs the French equivalent of ‘if’, namely ‘dans le cas ó’.
He himself admits, ‘[h]ad the drafters wished to imply condition, they would have used the form “dans le cas ó”.’39 For Ian Brownlie the French text is less equivocal than the English version.40
Thus, the wording of the first part of the first sentence of Article 51 is suf- ficiently clear to establish a logical condition.41 Another argument left with respect to the wording of Article 51 is to argue that ‘neither Article 51 nor any other word formula can have, apart from context, any single “clear and unambiguous” or
“popular, natural and ordinary” meaning that predetermines decision in infinitely varying particular controversies.’42 This is a valid argument, but an argument that is not persuasive within the boundaries of international legal scholarship. First, if one were to reduce the importance of legal text vis-à-vis some other influence one would give up something which is a norm and therefore does not need to be established as authoritative – unlike the norms of customary international law – but only requires interpretation within a given frame. That relative certainty is replaced by other influencing factors (what other factor but a text is as certain to express the meaning of the norms?) which are very uncertain. Second, one can argue that that argument is self-defeating. Why would one ‘believe’ Article 51 when it claims to receive the ‘inherent right’ of self-defence if words do not mean anything? Third, if one were to impute that the argument is that written norms (legal texts) demand interpretation and that a text does not predetermine every
36 Miller (1985) supra note 29 at 66.
37 Cf. also Higgins (1962) supra note 6 at 299: ‘It should be noted that the French text is considerably less restrictive, reading ‘dans un cas ó un Membre des Nations Unies est l’objet d’une aggression [sic] armée.’
38 Miller (1985) supra note 29 at 66.
39 Miller (1985) supra note 29 at 66.
40 Brownlie (1962) supra note 20 at 242.
41 Brownlie (1962) supra note 20 at 242; Dahm (1962) supra note 28 at 52: ‘unequivocal and internally consistent wording’ ‘eindeutiger und in sich sinvoller Wortlaut’; Claus Kreò, Gewaltverbot und Selbstverteidigungsrecht nach der Satzung der Vereinten Nationen bei staatlicher Verwicklung in Gewaltakte Privater (1995) 172; O’Connell (2002) supra note 24 at 13; Verdross and Simma (1984) supra note 28 at 288.
42 McDougal and Feliciano (1961) supra note 12 at 234.
Uncertainty in International Law 14
case, every future interpretation, then the author’s assertion will find universal consent – as long as it is within the the frame of possible meaning.43 To deny the phrase ‘if an armed attack occurs’ the role of logical connector would transcend the possible meanings that can be attached to it and would thus stand outside the norm (see Section 4.2).
At this point the battleground shifts to the preparatory works of Article 51 at the San Francisco conference. Timothy McCormack uses the travaux préparatoires as his main argument against the conditionality thesis.44 The main thrust of this argument is that nothing in the process of drafting shows that the Drafters wanted to restrict the pre-Charter right of self-defence.45 McCormack makes several arguments from the silence of the Drafters, e.g.:
(a) There was no discussion of the phrase at issue at San Francisco. ‘If there had been an intention to deliberately restrict anticipatory self-defence by the inclusion of these words, then surely that would have been commented on by at least some of the delegates.’46
(b) The French text and the English text contain several inconsistencies.
McCormack argues that there was no intention to restrict self-defence to cases of armed attack, because the English text was not reconciled with the French text. This was not done, because the phrase was not considered of sufficient importance.47
(c) ‘If Article 51 had been intended to determine the limits to the right of self- defence under the Charter, the Article ought to have mentioned the amount of force that is permissible. However, Article 51 is silent as to this important requirement.’48 In his mind, therefore, there was no intention to restrict.
The counter-argument is that nothing shows intent to restrict a pre-Charter right, because nothing exists which could show such intent. There is no discussion in the travaux préparatoires of the phrase ‘if an armed attack occurs’.49 Albrecht Randelzhofer states categorically that ‘nothing can be drawn from the travaux préparatoires, either in support of this interpretation or against it’.50 Indeed, one could argue that nothing in the preparatory works restricts privileging a pre- Charter right of self-defence, except the clear wording of the resultant paragraph.
43 Section 4.2.
44 McCormack (1991) supra note 12.
45 Bowett (1958) supra note 11 at 188; Hans-Georg Franzke, Die militọrische Abwehr von Angriffen auf Staatsangehửrige im Ausland – insbesondere ihre Zulọssigkeit nach der Satzung der Vereinten Nationen, 16 ệsterreichische Zeitschrift fỹr ửffentliches Recht (1966) 128–175 at 141; McDougal and Feliciano (1961) supra note 12 at 235–236; Waldock (1953) supra note 12 at 496–497.
46 McCormack (1991) supra note 12 at 35.
47 McCormack (1991) supra note 12 at 36.
48 McCormack (1991) supra note 12 at 37.
49 Brownlie (1962) supra note 20 at 242.
50 Randelzhofer (2002b) supra note 28 at 792 (MN 10).
2.2.1 Self-defence 15
Modern international law has shown a clear tendency to privilege text over original intent, as evidenced by Article 32 Vienna Convention on the Law of Treaties 1969 (VCLT).51
After reviewing the debate we are left with a clear picture. Even if we were to support the ‘inherent right’ doctrine (Section 2.1), it is not correct to assume that such a relationship could unmake the clear words of Article 51. The phrase ‘if an armed attack occurs’ is written in a treaty text and cannot be unwritten by these arguments. Mary Ellen O’Connell recently put it thus:
Even if earlier custom allowed preemptive self-defense, arguing that it persisted after 1945 for UN members requires privileging the word ‘inherent’ over the plain terms of Article 2(4) and the words ‘armed attack’ in Article 51. Indeed, it requires privileging one word over the whole structure and purpose of the UN Charter.52
One could not be clearer than the majority in Nicaragua:
In the case of individual self-defence, the exercise of this right is subject to the State concerned having been the victim of an armed attack. Reliance on collective self- defence of course does not remove the need for this.53
The ‘inherent right’ of self-defence may be framed as declaration, but it is only part of the Charter because Article 51 admits it. If that text had not included the
51 The International Court of Justice has become convinced that the VCLT essentially reflects customary law, in particular its rules on interpretation. It has repeatedly said so: Arbitral Award of 31 July 1989 (Guinea-Bissau v. Senegal), Judgment of 12 November 1991, ICJ Reports (1991) 53 at 70 (para 47); Territorial Dispute (Libyan Arab Jamahiriya v. Chad), Judgment of 13 February 1994, ICJ Reports (1994) 4 at 21–22 (para 41); Legality of the Use by a State of Nuclear Weapons in Armed Conflict, Advisory Opinion of 8 July 1996, ICJ Reports (1996) 66 at 75 (para 19). Cf. Santiago Torres Bernárdez, Interpretation of treaties by the International Court of Justice following the adoption of the 1969 Vienna Convention on the Law of Treaties, in: Gerhard Hafner et al. (eds), Liber amicorum Professor Ignaz Seidl-Hohenveldern – in honour of his 80th birthday (1998) 721–748.
52 O’Connell (2002) supra note 24 at 13.
53 Nicaragua (1986) supra note 5 at 103 (para 195). This understanding of the law was recently reiterated and reinforced by the Court in Oil Platforms (Islamic Republic of Iran v. United States of America), Merits, Judgment of 6 November 2003, ICJ Reports (2003) 161 at 186–187 (para 51):
‘Therefore, in order to establish that it was legally justified in attacking the Iranian platforms in exercise of the right of individual self-defence, the United States has to show that attacks had been made upon it for which Iran was responsible; and that those attacks were of such a nature as to be qualified as
“armed attacks” within the meaning of that expression in Article 51 of the United Nations Charter, and as understood in customary law on the use of force’ (emphasis added). Cf. Jửrg Kammerhofer, Oil’s Well That Ends Well? Critical comments on the merits judgment in the Oil Platforms case, 17 Leiden Journal of International Law (2004) 695–718; Dominic Raab, ‘Armed attack’ after the Oil Platforms case, 17 Leiden Journal of International Law (2004) 719–735. In a more recent advisory opinion, the Court again reaffirmed that the presence of an armed attack is a necessary condition for the lawful exercise of the right of self-defence. Thus, the ‘orthodox view’ is for the moment supported by the Court’s jurisprudence constante: Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory, Advisory Opinion of 9 June 2004, ICJ Reports (2004) 136 at 194 (para 139): ‘Article 51 of the Charter thus recognises the existence of an inherent right of self- defence in the case of armed attack by one State against another State’ (emphasis added).
Uncertainty in International Law 16
right, there would have been no right – in Charter law – to self-defence: a right does not exist unless it is a norm. In this case, one might argue that customary international law has established such a right. However one might view the inter- relationship of the sources of international law, it is tautological that the Charter alone governs Charter law, even if the Charter itself were to refer to customary international law (as it might have done in Article 51). Therefore, if the ‘inherent right’ of self-defence of some wide, pre-Charter, scope, without the need for an armed attack, had been recognised by Article 51, it would have only done so ‘if an armed attack occurs’. Short of denying that words have any meaning, saying that the Charter does not say what it does could be described as a rhetorical clutching at straws. It is not a matter of uncertainty whether the presence of an armed attack is required for self-defence to justify a threat or use of force. The dissent we hear is an echo of a different time and no longer debated today. Today’s support for an extensive view of self-defence instead specialises on adapting the term
‘armed attack’ and the precise definition of ‘armed attack’ is the modern battle- ground; indeed it is within the core meaning of ‘uncertainty’ in international law.
Within the frame of possible meanings the true questions of interpretation start.