MFK-Mendip Job ID: 9538BK--0017-2 2 - 227 Rev: 14-05-2003 PAGE: 1 TIME: 12:53 SIZE: 61,11 Area: JNLS OP: RB EJIL 14/2 chg202 EJIL 2003 * J.W. Goethe University, Frankfurt a.M.; Peace Research Institute Frankfurt. .............................................................................................................................................................. EJIL (2003), Vol. 14 No. 2, 227­240 ............................................................................................. Terrorism and the Legality of Pre-emptive Force Michael Bothe* Abstract The use of military force is only lawful if and to the extent that it comes under an accepted exception to the general rule prohibiting the use of force, i.e authorization by the Security Council and self-defence. Lawful self-defence requires the actual existence of an armed attack, or of a situation to be considered as equivalent to an armed attack. A threat may be so direct and overwhelming that one cannot require the victim to wait to act in self-defence until the attack has actually started. This principle of necessity and immediacy is still part of customary international law. The doctrine of pre-emptive strikes formulated in the recent US National Security Strategy proposes to adapt this concept to new perceived threats in a way that would constitute an unacceptable expansion of the right of anticipatory self-defence. Vagueness and the possibility of abuse of any broader definition requires maintaining the traditional strict approach. A change might result in the abolition of the prohibition of the use of force altogether. Opening up broader possibilities for anticipatory self-defence is not desirable. To face so-called new threats, recourse to the Security Council is preferable to unilateral use of force based on a doctrine of pre-emptive strikes. 1 Introduction: Current Scenarios The title of this article is formulated in general and abstract terms It is not intended to provide a legal opinion on a particular event or contingency. Nevertheless, if legal science is to make a meaningful contribution to the ongoing debate on the legitimate use of force in international relations, it must consider specific scenarios. As will be shown, certain variations regarding facts trigger different legal arguments. If these differences are brought to bear, the paper has indeed to refer to current policy debate and actual policy planning, as is documented, for instance, in the `National Security Strategy' submitted to Congress by President Bush on 17 September 2002. This is also the approach taken by Abraham Sofaer in his contribution to this symposium. The Strategy can be seen in the tradition of an older debate which received a new 228 EJIL 14 (2003), 227­240 1 Weller, `The Changing Environment for Forcible Responses to Non-traditional Threats', 92 ASIL Proceedings (1998) 177. impetus after the end of the East­West conflict: the perception of `new threats'1 and the discussion about effective and legitimate responses to these threats. The Strategy brings specific threats into sharp focus and pinpoints specific means of dealing with them, including but not limited to the use of military force. The scenario of the threats is twofold. The first scenario is non-state violence, i.e. terrorism -- whatever the exact definition of that phenomenon might be. The second is the existence of `rogue' states, which may harbour or at least support terrorists and/or which possess or acquire weapons of mass destruction (WMD) and are willing to use them without restraint. 2 The Point of Departure: The Illegality of Anticipatory Self-defence The point of departure of the legal debate is uncontroversial: the prohibition of the use of force is a valid norm of customary international law and is enshrined in the Charter of the United Nations. Thus, any specific use of force can be lawful only if it can be based on an exception to this rule which is valid as a matter of law. Leaving aside the question of intervention by invitation, there are two such exceptions: self-defence and authorization by a competent international organization. Furthermore, it is uncontroversial that lawful self-defence requires the existence of an armed attack. It is on the interpretation of that principle that there is indeed disagreement. For this principle is only the starting-point of the legal debate when it comes to the assessment of particular situations. There are at least three questions: What exactly is an armed attack? Could it be understood as including threats to use force? Are there, in addition, situations which, as a matter of law, are to be considered as equivalent to an armed attack? Who, in addition to the immediate victim, is entitled to use the right of self-defence? What is the point of reference for determining what is attack and what is self-defence? This is a question which arises where acts of violence occur in the context of an existing larger framework of tension or reciprocal use of force. We shall now examine these questions in this order. 3 The Definition of Armed Attack According to the rule of customary international law, which is formulated in Article 31 of the Vienna Convention on the Law of Treaties, the interpretation of the Charter Terrorism and the Legality of Pre-emptive Force 229 must start with the `ordinary meaning to be given to the terms of the treaty'. It is by this means that one must ascertain whether and to what extent an attack may include an action which is threatened, but which has not yet occurred. The usual method of ascertaining this `ordinary meaning' is by recourse to dictionaries. Thus, Webster's Encyclopedic Dictionary states: `an offensive military operation with the aim of overcoming the enemy . . .', or: `an offensive move in a performance or contest', and: `Attack . . . applies to the beginning of hostilities'. The Oxford Dictionary defines attack as: `violent attempt to hurt, overcome, defeat'. All of these definitions imply that attack is an actual action in the sense of a move forward. A threat of action does not yet constitute an attack in the ordinary meaning of the term. But Article 31 also refers to the ordinary meaning of the terms `in their context'. This means that other parts of the text have to analysed as to whether they shed some light on the meaning of a certain term. As to the meaning of the term `armed attack' in Article 51, one has first to take into account the text immediately connected to the term. The relevant sentence reads: `. . . an armed attack occurs'. It does not add `or threatens'. This also suggests that what is meant is an attack which actually happens, not one which is merely likely to happen. The next element of the context which needs to be considered is the formulation of the rule to which Article 51 constitutes the exception: Article 2(4) prohibits both the threat and the use of force. Thus, the Charter framers clearly were aware of the problem of the threat of force. It is hard to conceive that it was merely due to a drafting oversight that the notion of threat was only mentioned in the rule and not in the exception. Furthermore, it is appropriate to refer to Article 39, which defines those situations where the Security Council may take action under Chapter VII of the Charter, while Article 51 defines the situations where a state may unilaterally decide to use military force. In Article 39, three terms are used: `threat to the peace, breach of the peace, or act of aggression'. These notions together, quite obviously, cover a field which is broader than that of an armed attack within the meaning of Article 51. In this provision, too, the framers of the Charter did indeed consider the situation of a `threat', but the text clearly suggests that the power to authorize the use of force in the case of a mere threat lies with the Security Council alone. Such case remains below the threshold at which a state may decide to use force unilaterally. Finally, according to Article 31 of the Vienna Convention again, the meaning of the terms of a treaty must be ascertained `in the light of its object and purpose'. It is the object and purpose of the Charter of the United Nations to restrain the unilateral use of force. An interpretation which excludes the threat of an attack from the notion of an armed attack that triggers the right of self-defence is certainly compatible with, if not required by, this object and purpose. The conclusion is thus clear: `armed attack' in the sense of Article 51 is an actual armed attack, which happens (`occurs'), not one which is only threatened. This 230 EJIL 14 (2003), 227­240 2 Y. Dinstein, War, Aggression, and Self-Defence (1988), at 184; Randelzhofer, `Article 51', in B. Simma (ed.), The Charter of the United Nations. A Commentary (2nd ed., 2002) 803 (with many references, but citing no position to the contrary); O'Connell, `The Myth of Pre-emptive Self-Defence', Paper prepared in conjunction with the ASIL Presidential Task Force on Terrorism, at 8 et seq (also containing further references). 3 GA Res. A/3314 of 14 December 1974. 4 Military and Paramilitary Activities in and against Nicaragua, ICJ Reports (1986) 14, at 103. 5 Annex to Resolution 2625 (XXV). conclusion is shared by the overwhelming majority of legal doctrine, which clearly holds `anticipatory self-defence' to be unlawful.2 This being so, the use of force in response to a perceived threat of an attack can be considered as lawful only if it is possible to show that, for some reason, the interpretation of the term `armed attack' must be construed more broadly, using a broader notion of context, founding itself on more general considerations. If one considers strategies of legal reasoning which are used, in practice, to justify the use of force in particular cases, the approach based on an expansive interpretation of `armed attack' is very common. Two examples may be given before we turn to the question of anticipatory or pre-emptive self-defence. The first example is that of indirect attacks. The rule prohibiting the use of force applies to situations involving states; it is not, as a matter of principle, concerned with actors which are not subjects of international law. However, in situations whereby a state is actually involved, to a sufficient degree, in non-state violence, it is indeed accepted that this involvement is equivalent to an armed attack and may thus entail the same consequences as an armed attack, namely it can trigger the right of self-defence. To quote the definition of aggression adopted by the General Assembly in 1974:3 Any of the following acts qualify as an act of aggression: . . . (g) 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 the acts listed above (i.e. acts of transborder military force by state organs), or its substantial involvement therein. In its Nicaragua judgment, the ICJ accepted this text as being an expression of customary law and took it as a point of departure for the legal assessment in that particular case.4 On the basis of this construction, the intervention of the United States and its allies in Afghanistan after September 11 was, controversial details apart, justified as self-defence. This is an approach to legal reasoning which can be called a constructive armed attack, or a situation equivalent to an armed attack. The second example of this constructive armed attack approach is the so-called `blue water theory'. It has remained controversial and is now legal history. According to this theory, to quote the `Friendly Relations' Declaration of the General Assembly,5 [t]he territory of a colony or other non-self-governing territory has, under the Charter, a status separate and distinct from the territory of the State administering it . . . Terrorism and the Legality of Pre-emptive Force 231 6 Randelzhofer, supra note 2; O'Connell, supra note 2. 7 State Secretary Webster, British and Foreign State Papers 29 (1840­1841), at 1129, 1138. 8 Dinstein, supra note 2, at 182, 244; Meng, `The Caroline', in R. Bernhardt (ed.), Encyclopedia of Public International Law, vol. 1 (2000), at 537 et seq. 9 `National Security Strategy', at 15, available at www.whitehouse.gov/nsc/nss.html. The consequence of this construction is, then, drawn in the elaboration of the prohibition of the use of force: Every State has the duty to refrain from any forcible action which deprives peoples referred to in the elaboration of the principle of equal rights and self-determination of their right to self-determination and freedom and independence. The background of these formulations is the attempt, be it achieved or not, to construe the suppression of fights for decolonization as armed attacks which thereby trigger a right of (individual and collective) self-defence. These examples may suffice to show that the concept of constructive armed attack, or a situation equivalent to an armed attack, is not foreign to the arsenal of international legal reasoning. But whether such constructions have really become positive international law is another question. In this respect, one has to note that use of the consequence of the `blue water theory' to serve as a legal justification of counterforce was not really accepted by the Western powers at the relevant time. The International Court of Justice, in its Nicaragua judgment, while accepting the construction of an armed attack by involvement in non-state transborder violence, gave it a somewhat restrictive interpretation which, according to some authors, sheds some doubt on the legality of the US intervention in Afghanistan, a view which I do not share. Having these general considerations concerning the interpretation of the notion of armed attack in mind, we can now turn to the question of anticipatory or pre-emptive self-defence. Many authors acknowledge that a threat may be so direct and overwhelming that it is just not feasible to require the victim to wait to act in self-defence until the attack has actually started. In this case, a situation equivalent to an armed attack prevails. A formula expressing this idea and its limits, which is not uncontroversial,6 but still quoted with considerable agreement, is that pleaded by the United States in the Caroline case in 1841: There must be `a necessity of self-defence, instant, overwhelming, leaving no choice of means, and no moment for deliberation'.7 It is thus at least defensible that the principle of necessity and immediacy, as expressed in the Caroline formula, be considered as part of customary international law, even under the United Nations Charter.8 I submit that this is as far as pre-emptive self-defence possibly goes under current international law. It is interesting to note that President Bush's National Security Strategy seems to recognize this as being the law, at the same time distancing itself from it:9 Legal scholars and international jurists often conditioned the legitimacy of pre-emption on the existence of an imminent threat -- most often a visible mobilization of armies, navies and air forces preparing an attack. 232 EJIL 14 (2003), 227­240 This is a point of departure based on the Caroline formula. But the text continues: We must adapt the concept of imminent threat to the capabilities and objectives of today's adversaries. . . . The greater the threat, the greater the risk of inaction -- and the more compelling the case for taking anticipatory action to defend ourselves, even if uncertainty remains as to the time and place of the enemy's attack. It is hard to tell whether this is meant as an argument de lege lata or de lege ferenda. There is nothing in current state practice, case law or legal writing which would suggest that this broad, even overly broad construction of a situation equivalent to an armed attack is part of positive customary law today. The marriage between the precautionary principle, known from the field of national and international environ- mental law, and the right of self-defence is, to say the least, a novelty. In the field of environmental law, the precautionary principle means that action to protect the environment may or must be taken, even in the case of uncertainty about the danger. If transferred to the field of legitimization of the use of force, that rule would then read: in case of uncertainty, strike -- a somewhat weird conclusion. Whether or not the argument of the gravity of the threat is valid as a matter of fact, a good factual argument does not make new law. The lex ferenda aspect of the argument will be analysed below. De lege lata, however, the expansion of the right of anticipatory self-defence proposed in the National Security Strategy is not acceptable. There are voices that contradict this assertion. They maintain that a broader interpretation of admissible self-defence is still the law, that older conceptions of permissible self-defence have somehow survived the entry into force of the Charter of the United Nations. The argument takes as a point of departure an argument which constitutes a recourse to context, as already mentioned above. The particular context used for the interpretation is a further element in the text of Article 51, namely that it recognizes self-defence as an `inherent' right. They understand this word as a referral to previous concepts of self-defence predating the Charter. As a matter or textual interpretation, one can engage in a kind of intellectual game: Does the notion of an inherent right allow or even require a broader interpretation of the term `armed attack'? Or does the narrow notion of an armed attack, used without any qualification or addition, limit the concept of an inherent right? Both constructions are logically possible. But it is certainly not possible, without betraying basic professional rules of interpretation, to re-interpret the notion of self-defence, understood, according to the ordinary meaning as developed above, as a proportionate response to an actual attack, as meaning any use of force reasonable under the (threatening) circum- stances. Whether or not the Caroline formula was too restrictive as far as the law of the nineteenth century is concerned, whether or not a broader concept of the use of pre-emptive force based on natural law concepts was positive law in the nineteenth century, nothing in state practice suggests that such concept could have survived the entry into force of the Charter or has been revived since that date. Any other conclusion would treat both the Kellogg­Briand Pact and the Charter of the United Nations incorrectly as being irrelevant. This is true, although, admittedly, the concept of necessity as expressed in the Terrorism and the Legality of Pre-emptive Force 233 10 See Kirgis, `Pre-emptive Action to Forestall Terrorism', ASIL Newsletter, July/September 2002, at 5. 11 In this respect, it is useful to note the analysis of the threat posed by Iraq contained in a recent paper published by the UK Government, a paper designed to justify the US approach: whether there are any remaining stocks of BW or CW substances is uncertain, and it is probable that there are some production facilities that could be used (`Iraq's Weapons of Mass Destruction. The Assessment of the British Government', at 19 et seq, available at www.official-documents.co.uk/document/reps/iraq/cover.htm). No nuclear weaponry is currently available, but Iraq might acquire the ability to produce such weapons under certain conditions which may or may not be fulfilled in the future (ibid., at 26 et seq). All this is not the type of immediate threat which leaves `no moment for deliberation'. 12 It may be argued that this was the true justification for the British attack on the Caroline, which gave rise to the diplomatic correspondence from which the famous formula used by Webster stems. Caroline formula is not static.10 It is a concept which allows for some flexibility and adaptation to the circumstances of each particular case. But it certainly cannot be extended to a situation where there is uncertainty -- and time for many months of debate in national and international fora, including serious negotiations on ways of handling a dispute other than by the use of military force.11 4 The `Attacker': The Failed State and Non-state Force Another scenario related to that of `indirect' force, already mentioned above, has also to be analysed in this context, namely situations where armed force is used by non-state groups from a territory which is not effectively controlled by the (or a) government of the state in question. The question arises whether this could be considered a situation equivalent to an armed attack by that state. According to the `substantial involvement' formula reported above, this would probably not be the case. In the case of a failed state, there is no government to be `involved'. Or can an omission constitute an involvement? This appears somewhat doubtful. On the other hand, it is possible to argue that indeed an omission is equivalent to an action where there is a legal duty to act that is violated by the omission. On the basis of the assumption that there is a legal duty of every state to prevent transborder activities of terrorism originating from its territory, that construction may be considered valid. It is to be recognized that this may be the only way in which a state is able to protect itself against transborder force in a case where a right of self-defence would exist if that force were used by organs of the neighbouring state.12 This construction is in any event preferable to a thesis, which can sometimes be found in legal doctrine, to the effect that a `failed state' is not protected by the principle of the prohibition of the use of force. 5 The Victim and Others: Collective Self-defence The current debate about a pre-emptive strike against Iraq raises an additional problem as to who is entitled to exercise a right of self-defence Self-defence is the right of the victim of an attack. Although the prohibition of the use of force may be considered as an obligation valid erga omnes, the right of individual self-defence 234 EJIL 14 (2003), 227­240 13 A. de Hoogh, Obligations erga omnes and International Crimes (1996), at 327. 14 Missiles currently available could reach parts of Turkey and Saudi Arabia as well as Israel, Bahrain, Qatar, Georgia, Azerbaijan and Armenia. Missiles which might become available would also be able to reach larger parts of Turkey, Southern Russia and even some small Greek islands (`Iraq's Weapons of Mass Destruction', supra note 11, at 31). 15 Supra note 4, at 105. accrues only to the victim of the attack. In other words: no self-defence without a victim.13 Assuming pre-emptive self-defence were lawful, it would be the supposed or anticipated victim that possessed this right. Where the Caroline formula is applied, there can be no doubt about the identity of the victim. But in the case of the threat posed by Iraq, there is a serious problem. The danger that Iraq might attack the territory of the United States is far-fetched.14 Is, then, a right of anticipatory self-defence also claimed where the danger is that of an attack against a military presence abroad, e.g. the United States forces stationed in Saudi Arabia or Qatar? That would really overstrain the notion of a situation equivalent to an armed attack. In this situation, there remains the right of collective self-defence. If one assumes that Israel, Saudi Arabia, Turkey or Iran are threatened by the weapons of mass destruction in the possession of Iraq, a scenario which is not far-fetched, and that this threat constitutes a situation equivalent to an armed attack, a right of collective self-defence would also exist for the United States, if and to the extent that the other requirements of this right are fulfilled. According to the Nicaragua judgment of the ICJ, the right of collective self-defence requires a request by the victim possessing the right of individual self-defence.15 Thus, a pre-emptive strike by the United States against Iraq would require a request made by a state of the region which could be considered to be directly threatened by missiles in the possession of Iraq. 6 Self-defence within an Existing Armed Conflict The last question de lege lata which has to be examined in considering the legality of pre-emptive strikes is that of the point of reference for the determination of what constitutes an armed attack and what is defence. This problem is quite clear in the case of an armed conflict. Where an international armed conflict between two States exists, that point of reference is the beginning of the armed conflict. Once that conflict exists, the only question to be asked under the rules of the ius ad bellum, the prohibition of the use of force, is: who started the whole conflict? The individual military action undertaken within the framework of the conflict can only be judged in the light of the ius in bello, but not by the yardstick of the ius ad bellum independently from the question which party violated the ius ad bellum by starting the conflict. The consequences of this legal situation for the legality of military action which might be called pre-emptive are obvious. The question of the legality of anticipatory self-defence Terrorism and the Legality of Pre-emptive Force 235 16 The question may be asked whether in certain cases a conflict is limited to a certain theatre of war, or in a similar way, the expansion of the conflict beyond these bounds might constitute an action which should be judged on its own merits in the light of the prohibition of the use of force. 17 Dinstein, supra note 2, at 186. 18 Blum, `The Beirut Raid and the International Double Standard', 64 AJIL (1970) 73. 19 O'Brien, `Reprisals, Deterrence and Self-defence in Counterterror Operations', 30 Va.J. of International Law (1990) 421, at 426. 20 SC Res. 687 of 3 April 1991: `The Security Council . . . Conscious of the need to take the following measures acting under Chapter VII . . .' is simply irrelevant for assessing the legality of individual operations undertaken in the framework of an ongoing armed conflict.16 Israeli writers have quite often used the argument of the existence of an armed conflict between Israel and the Arab States to justify certain military actions undertaken by Israel. Examples are the destruction of an Iraqi nuclear reactor by Israel in 1981, otherwise unlawful pre-emptive self-defence,17 and the Beirut raid in 1968, otherwise an unlawful armed reprisal.18 In principle, an individual military action undertaken within the framework of an armed conflict cannot be singled out to be judged according to the yardstick of the ius ad bellum. But then the question arises whether there exists an ongoing armed conflict which covers the individual action. In relation to Israel, the Security Council has never accepted that argument and always judged individual Israeli actions in relation to each particular incident which did or did not trigger a right of self-defence for Israel.19 In relation to a preventive strike by the United States against Iraq, the question thus arises as to how their current relationship is to be characterized. The first question to be asked is whether the armed conflict which started with the invasion of Kuwait by Iraq still exists, meaning that the United States are still entitled to a right of collective self-defence for the victim Kuwait (and/or that the authorization to use force given by the Security Council in Resolution 678 of 29 November 1990 is still open as a basis for military action, a related but different question). It is submitted that the success of the alliance and the armistice resolution fundamentally changed the situation. The actual armed attack committed by Iraq was repelled, the sovereignty of the Kuwaiti authorities over their territory reinstalled. The armed conflict ended. A situation of self-defence no longer existed. It is true that the Security Council still considered that the situation came under Article 39 of the Charter and that the Council thus continued to act under Chapter VII of the Charter.20 But that does not mean that the armed conflict continued. The measures taken by the Security Council were typical for a post-conflict situation. Iraq repeatedly violated the provisions of the armistice resolution. But this does not mean that the armed conflict was reactivated. When the United States, and later the United Kingdom, indeed attacked Iraq to enforce the no-flight zones, the question whether they were justified arose anew. The question which arises as a consequence is whether there has been a continuous armed conflict since that resumption of military activities, meaning that the legality, under the ius ad bellum, of a new attack would not constitute a question separate from that of the legality of the earlier bombing. I submit that the various bombing periods 236 EJIL 14 (2003), 227­240 21 Supra note 9, at 15. constituted separate periods of armed conflict, each of which has to be assessed separately in the light of the ius ad bellum. The notion of continuous armed conflict is a dangerous one, open to abuse. It was argued that the authorization given earlier by the Security Council was still open as a basis for military action. This is a different question. Here we switch to the second type of legal justification of the use of force mentioned in the Introduction above. It would have been for the Security Council, not for individual states, to determine the consequences to be drawn from the breach of its resolutions. Such breach could not automatically re-establish the situation as it was before the resolutions. Thus, the resolution of April 1991 does not constitute a valid basis for strikes against Iraq taking place in 2002 or 2003. Another possibility might be to argue that an armed conflict broke out between Iraq and the United States because Iraq supported some of the terrorist activities of Al Qaeda. Whatever Iraq did to train or otherwise support Al Qaeda fighters, no evidence has been submitted to date that indicates that Iraq's involvement in the September 11 attacks was in any way comparable to that of the Taliban, which, as pointed out above, was the basis for the fact that the United States possessed a right of self-defence against Afghanistan. Thus, the construction of an armed conflict between the United States and Iraq cannot justify any pre-emptive strike by the former against the latter. 7 Lex ferenda As has been pointed out above, the legal reasoning developed in the National Security Strategy may also be understood as an advocacy de lege ferenda The question which has to be asked is how this change in the law could be achieved, and what type of modification of the law might be desirable. As an amendment to the UN Charter to modify Article 2(4) is out of the question, a change in customary law might be a way. A usual procedure to modify customary law is to break it and to accompany the breach by a new legal claim. This is, for instance, how the development was triggered which finally led to the recognition, also under customary law, of the concept of the exclusive economic zone in the Law of the Sea. Does President Bush's National Security Strategy constitute a step in this direction? This leads us to the further question of whether such development would be desirable. We have, thus, to ponder the pros and cons of changing the restrictive concept of anticipatory self-defence presented above. The case for a change of this restrictive concept of pre-emptive self-defence is made by the National Security Strategy: Given the goals of rogue states and terrorists, the United States can no longer solely rely on a reactive posture. The inability to deter a potential attacker, the immediacy of today's threats, and the magnitude of potential harm that could be caused by our adversaries' choice of weapons, do not permit that option. We cannot let our enemies strike first.21 To put it simply, the argument is: we cannot wait. Indeed, the traditional approach Terrorism and the Legality of Pre-emptive Force 237 22 Hailbronner, in D. Schindler and K. Hailbronner, `Die Grenzen des völkerrechtlichen Gewaltverbots', 26 Berichte der Deutschen Gesellschaft für Völkerrecht 49, at p. 80 et seq (1986). 23 See, for instance, SC Res. 487 of 19 June 1981 relating to the Israeli attack against the nuclear reactor in Bagdhad. has always had the drawback of depriving a potential victim of the possibility to choose the most advantageous moment to fight a danger which may be extreme.22 The argument, thus, is not new. It has been used by Israel to justify a number of incursions into the territory of its neighbours, and has been rejected by the Security Council.23 The prohibition of the use of force, including the prohibition of anticipatory self-defence, has developed in international practice and doctrine despite the awareness of this drawback. This should make us think twice before arguing for a change of the law. An essential argument for maintaining the restrictive concept is the problem of vagueness and the possibility of abuse. Any new rule to be created would have to give an adequate answer to the following questions: How to define and limit a possibly expanded right of self-defence? How serious must the threat be? Is possession of weapons of mass destruction enough? Who is threatened and who may attack? What about the possession of nuclear arms by India, Pakistan, North Korea and Israel? What precisely distinguishes them (if there is a difference), in legal terms, from Iraq? What does `harbouring' terrorists mean? There must be knowledge. But if there is, what kind of effort is a state required to make in order not to be considered as harbouring terrorists? Satisfactory answers to these questions are not at hand. All too easily, a standard of reasonableness boils down to subjectivity and speculation. The National Security Strategy seems to recognize the dilemma, in particular the risk of abuse: . . . nor should nations use pre-emption as a pretext for aggression. This sentence is followed, however, by a somewhat enigmatic postulate: Yet in an age where the enemies of civilization openly and actively seek the world's most destructive technologies, the United States cannot remain idle while dangers gather. Does this sequence imply a differentiation between (other) `nations' and the United States? Is the demand, thus, a special law for the hegemon? Does the strategy, in its attempt to change the law of self-defence, also want to abolish the principle of sovereign equality? If one looks seriously at the problems caused by the argument de lege ferenda, more questions than answers appear. Any expansion of the right of pre-emptive self-defence beyond the Caroline formula is just not workable. That being so, an additional problem of legal policy arises, namely the vulnerability of the prohibition of the use of force. The impossibility of placing any legal limit on the exception means that the validity of the prohibition of the use of force itself will be in jeopardy. A look back into history can clarify the problem. In the early times of modern international law, in the era of Grotius, the theory of `just war', of bellum iustum applied: a war was lawful when fought for a just purpose and by just means. 238 EJIL 14 (2003), 227­240 International practice of the eighteenth and nineteenth centuries had to accept that this rule was not viable. It was impossible to determine in any particular case whose cause was just and whose not. As a result, the rule of bellum iustum, which at the outset was understood as a legal restraint on war, turned into the opposite: international law became indifferent as to resort to war. Until the end of the First World War, resort to war was not forbidden, except in the case of a specific treaty (for instance, in the case of Belgian neutrality). Thus, the attempt to create a rule which is unable to give a workable definition of permissible force might end in the abolition of the prohibition of the use of force altogether, as previously occurred. This would mean destroying one of the most important and salutary cultural and political achievements of the twentieth century. That danger is all the more real as the rule prohibiting the use of force is particularly vulnerable for another reason as well. This rule was not really developed by state practice. There has never been a consistent practice of abstention from the use of force. What changed after the First World War was the reaction of relevant actors against the use of force. An opinio iuris slowly emerged that war was unlawful, and official reactions to the use of force changed accordingly. This change was brought about by a more fundamental societal change, a change of social value judgments. Shocked by the immense human suffering caused by the First World War, society was no longer ready to accept war as a positive phenomenon as it had done before, when the pacifism that existed was rather ridiculed by the societal establishment of the day. Briefly: the rule prohibiting war, and later the use of military force in general, is the product of a change in social value judgments. It lies in the very nature of such a process that it is reversible. True, we are a long away from a re-militarization of the value system of our civil society. But if we want to maintain international law as a restraint on the use of military force, we should very carefully watch any attempt on the part of opinion leaders to argue that military force is anything other than an evil that has to be avoided. The lessons of history are telling. If we revert to such broad concepts, such as the just war concept, to justify military force we are stepping on a slippery slope, one which would make us slide back into the nineteenth century when war was not illegal. It was exactly this concern about the slippery slope which prompted the Kennedy Administration during the Cuban missile crisis not to rely on the self-defence argument in order to legally justify the pre-emptive action it took, namely the so-called `quarantine'. Abram Chayes, then Legal Adviser to the State Department, explains this choice in the following words: In retrospect . . . I think the central difficulty with the Article 51 argument was that it seemed to trivialize the whole effort at legal justification. No doubt the phrase `armed attack' must be construed broadly enough to permit some anticipatory response. But it is a very different matter to expand it to include threatening deployments or demonstrations that do not have imminent attack as their purpose or probable outcome. To accept that reading is to make the occasion for forceful response essentially a question for unilateral national decision that would not only be formally unreviewable, but not subject to intelligent criticism, either, . . . Whenever a nation believed that interests, which in the heat and pressure of a crisis it is prepared to Terrorism and the Legality of Pre-emptive Force 239 24 A. Chayes, The Cuban Missile Crisis (1974), at 63 et seq. 25 The question whether this was a sufficient justification is a different matter. characterize as vital, were threatened, its use of force in response would become permissible . . . In this sense, I believe that an Article 51 defence would have signalled that the United States did not take the legal issues involved very seriously, that in its view the situation was to be governed by national discretion, not international law.24 The conclusion, thus, should be clear: a change in the law to the effect of opening up broader possibilities for anticipatory self-defence is not desirable. The objection which can be raised against the restrictive approach (de lege lata or de lege ferenda) on the use of military force is that it is `unrealistic', that in order to be meaningful it should authorize the use of force which is `reasonable' under the circumstances -- and outlaw only `unreasonable' use of force. But reasonableness and proportionality are concepts which are difficult to operationalize in the context of a decentralized system. They open the door to arbitrariness and subjectivity. It may be that the risk of violation is higher if the rule is very restrictive. Much will then depend on the reaction of other actors in individual cases. It may well be that the international community, reshaping the opinio iuris, will one day accept some instances of pre-emptive use of force. But this, it is submitted, is a much safer approach to the interpretation and development of the ius ad bellum than loosening any real restraint by boiling it down to a rule of reason -- a self-destructive mechanism for the prohibition of the use of force. 8 The Solution of the Dilemma: The Multilateral Option It is traditional wisdom of legal theory that where substantive law cannot bring about a sufficient degree of legal certainty, procedural rules must be used to obtain results which are socially or politically acceptable. This wisdom can and should be applied to the problem under discussion. We are facing a situation where an established rule, i.e. the prohibition of the use of force, including anticipatory self-defence, is challenged. But it is difficult, if not impossible, to create a new rule which is able to accommodate legitimate concerns without opening the door to unacceptable abuse. Practicable substantive legal restraints on the use of pre-emptive force are not readily available. On the one hand, there are serious doubts about the wisdom of the traditional rule which strictly limits anticipatory self-defence, but on the other hand, loosening these limits is also not acceptable, as no workable limits can be conceived as a substitute for the old ones. It is in this situation that the need for accepted procedures legitimizing the use of force arises. That was the Kennedy approach in the Cuban missile crisis when he sought and obtained OAS approval.25 The claim of better knowledge, better morals or the like does not create sufficient legitimization in the international system. This procedure does not need to be invented -- it already exists. This is, as rightly pointed out by the UN Secretary-General, the authorization by the Security Council. It is, in my view, one of the essential flaws of the National Security Strategy that the 240 EJIL 14 (2003), 227­240 United Nations and its system of collective security are not even mentioned, let alone examined as a viable option of security policy. The argument which is often heard against recourse to the Security Council as a source of legitimization is that it is all too often blockaded. There is, indeed, a certain balance of power in the Council, a system of checks and balances, designed with some degree of political wisdom by the drafters of the Charter. As a consequence, a majority led by one of the permanent members of the Council does not necessarily have its way. This is a leverage which all permanent members of the Council have used to their advantage as they thought appropriate. But this is not what can appropriately be called a blockade. Negotiations leading to a reasonable result are not impossible, as they were during the times of the `automatic' veto which characterized the era of the Cold War. In this new situation, recourse to a Security Council mandate is the only acceptable solution, both as a matter of law and policy, where, in the light of threats of terrorism and use of weapons of mass destruction by irresponsible governments, military action which cannot be construed as constituting self-defence seems to be required.