The Use of Force against Terrorists: A Rejoinder to Federico Sperotto and Kimberley N. Trapp

I am grateful to Dr Trapp and Mr Sperotto for their comments on my article ‘The Use of Force Against Terrorists’1 and to the Journal’s editors for permitting me to add a rejoinder. While addressed to very particular and very different aspects of the article, I believe the two comments help put the argument made in it into perspective. Mr Sperotto situates my survey of legal developments within the political and historical context. The increased resort to force against terrorists in his view illustrates a more general shift, described as the move ‘from a “Grotian” model founded on common rules and institutions consolidated in the UN Charter, towards a “Hobbesian” one, dominated by the obsession for security and some rules of prudence’.2 In this ‘Hobbesian’ state of affairs, states have re-appropriated ‘a right that they had lost as a result of the creation of the UN’.3 9/11 is viewed as the decisive catalyst for this process, and debates about anticipatory self-defence are used to illustrate its problems. I agree with many of Mr Sperotto's points and am grateful to him for providing a broader perspective to my legal analysis. That said, I believe his attempt to place the use of force against terrorists within the ‘Hobbes v. Grotius matrix’ calls for three qualifications – two of them points of detail, one of a more general relevance. (i) In my view, Mr Sperotto overstates the impact of 9/11 on the regime governing anti-terrorist force. Of course, the attacks on the twin towers have changed states’ perception of what terrorists are capable of, and influenced their interpretation of the jus ad bellum. However, contrary to Mr Sperotto's understanding I do not think that 9/11 ‘suddenly stopped’4 the trend towards a Grotian model. The restrictive analysis of the jus ad bellum had come under pressure long before. My article illustrates this by referring to the 1998 US attacks on Sudan and Afghanistan, or incursions into Iraq by Turkish and Iranian forces during the 1990s.5 It deliberately portrays 9/11 not as a ‘sudden reversal’, but as part of the international community's ongoing process of adapting the jus ad bellum to changing realities – a process which, in other fields, has led to the recognition of a right to rescue nationals and, during the 1950s–1980s, seemed to lead towards the acceptance of anti-colonialist force. When singling out one specific event (even one as decisive as 9/11), I believe one does not fully appreciate the continuous nature of the interpretative process. (ii) I was surprised by Mr Sperotto's focus on anticipatory self-defence. While agreeing that states are pushing for an expansive reading of self-defence, I do not think debates about anticipatory self-defence illustrate this trend very well. For the reasons set out in the article,6 the radical re-reading of self-defence into a right to use pre-emptive force is not likely to succeed. As for anticipatory self-defence proper, there has probably been a development, but it seems increasingly overshadowed by the gradual acceptance of the accumulation doctrine. Of the various challenges to the restrictive analysis, anticipatory self-defence to me seems the least dangerous. (iii) Finally, Mr Sperotto's main concern: Grotius and Hobbes. It is part of the fascination of both writers that nearly every development in international law can be described as a move from Hobbes to Grotius or vice versa – at least in Hedley Bull's influential analysis (which Mr Sperotto adopts), Hobbes and Grotius, together with Kant, provide the three relevant ‘competing traditions of thought’ which can explain ‘the history of the modern states system’.7 Labels such as ‘Grotian’ or ‘Hobbesian’ capture different understandings of the international system, and on that basis can be employed usefully. My only concern is that we ought not to employ them schematically. To avoid that risk, I would add two caveats to Mr Sperotto's description. The first may seem trivial: when comparing Grotius and Hobbes, it is crucial to underline that one juxtaposes their ‘approaches’, not their actual views of the law. To Grotius, the use of force against terrorists could have amounted to a ‘just war’, which could serve to punish opponents.8 His proposed regime governing military force was rule-based, and in that respect may have presented a great advance, but the rules had little in common with the ones enshrined in the UN Charter. We may herald the ‘Grotian moments’ of 20th century international law, but should be grateful that our world is no longer governed by Grotius’ jus ad bellum. Secondly, even when focusing on the spirit rather than the letter of Grotius, we should be prepared to accept that not each and every re-adjustment of the jus ad bellum necessarily leads to ‘a more anarchical world’.9 It is telling that in his comment, Mr Sperotto focuses on the unilateral use of force. The article suggests that international law has evolved in another respect, namely by permitting the use of force against terrorists within the UN's system of collective security.10 This aspect of developments, if anything, reinforces the role of ‘international institutions consolidated in the UN Charter’, which Mr Sperotto views as part of the ‘“Grotian” model’.11 But even when focusing on the unilateral use of force, it may well be that re-adjustments permitting ‘new’ uses of force are brought about not by ‘obsession for security’,12 but by the progressive realization that certain community goals require effective protection, including protection by forcible means. At this stage, one should not dismiss the possibility that freedom from terrorism may turn out to be one such goal, just as freedom from colonialism seemed to be to most states during the 1960s and 1970s. Both caveats I believe may help avoid misunderstandings caused by Grotian or Hobbesian labels. In her comment, Dr Kimberley N. Trapp focuses on a more specific aspect, namely questions of attribution. She is critical of my attempt to explain recent instances of anti-terrorist self-defence by modifying the regular standard of attribution, which in her view is conceptually problematic and has distorting effects on Article 51 of the UN Charter and on the law of state responsibility.13 Dr Trapp puts forward a differentiated approach: in her view, attribution in the strict sense remains necessary if a state seeks to defend itself against another state. If the response is directed against terrorists as such, attribution is not necessary; in this case, the requirement of necessity provides sufficient protection against abuse: ‘[i]f a state is complicit in its territory being used as a base of terrorist operations, then a use of defensive force in response to terrorist attacks by non-state actors from that state's territory is necessary, and the complicity provides the justification for the violation of the host state's territorial integrity’.14 This indeed is an alternative way of explaining recent practice. I do not think, however, that it is as different an explanation from my own approach as Dr Trapp suggests, or that it offers a more convincing explanation. (i) Disagreement on ‘modified attribution’ versus ‘necessity’ approaches overshadows the fact that Dr Trapp and I proceed from the same starting-point and reach essentially the same result. Unlike many other commentators, we both accept that since Article 2(4) of the UN Charter prohibits inter-state force, there must be an inter-state element to self-defence.15 As for the result, Dr Trapp shares my main argument – probably still a minority view, but gaining ground – that contemporary international law has come to recognize a right of self-defence against terrorist attacks even where these cannot be attributed to another state under the traditional tests. We differ when it comes to justifying that result. (ii) To continue with commonalities, I believe neither of the two alternative explanations we offer is conceptually fully convincing. I readily accept that when seeking to explain the new practice of anti-terrorist self-defence in terms of attribution, one is in danger of ‘muddying the conceptual waters’.16 What puzzles me is that Dr Trapp suggests her approach could avoid this problem. Of course, by rejecting complicity (or other lesser forms of involvement), she manages to preserve the ‘integrity’ of attribution in its narrow sense. But she can do so only because she is prepared to recognize a second category of self-defence, into which the problematic instances of anti-terrorist self-defence are being ‘outsourced’. Her approach breaks down the concept of self-defence (requiring an armed attack and permitting necessary and proportionate responses) into two. ‘Self defence version 1’ can be resorted to against armed attacks which are attributed to another state. In contrast, ‘self defence version 2’, more limited in scope, permits forcible responses against non-state attacks and is available whenever such measures are necessary. This presents a blend of existing, competing approaches: the ‘armed attack’ requirement is interpreted to mean neither ‘state attack’ nor ‘armed attack irrespective of its author’, but said to be context-specific. Its interpretation depends on the scope of the right exercised. As a result, self-defence is sometimes limited by necessity only (version 2), and sometimes by attribution and necessity (version 1). (iii) Perhaps our approaches can be visualized as attempts to enlarge a house which has become too small. We both agree that there is a need for more space. My approach (modifying attribution) may be likened to the addition of a new room. Dr Trapp (admitting a second version of self-defence) builds a new house alongside the old one. There is nothing wrong with that – but it is curious that, having re-designed the whole site, she should criticize others for interfering with the original building. Rather, I believe Dr Trapp would have to explain why there could be a new house (a new version of self-defence). On this point, she remains rather cautious. She argues that the non-committal approach adopted in DRC v. Uganda ‘perhaps suggest[s]’ a distinction between two versions of self-defence.17 Moreover, a ‘context-specific reading’ of Nicaragua is said to reveal that the Court's pronouncements in the case were about anti-state self-defence only.18 But the first argument is speculative and the second difficult to sustain: the Nicaragua Court expressly noted that it would ‘define the specific conditions … [of self-defence], in addition to … necessity and proportionality’ and then went on to discuss the required degree of state involvement without the slightest hint that this should depend on the scope of self-defence operation.19 Still less can any such limitation be found in the subsequent statement in the Wall opinion.20 Both pronouncements can of course be criticized for a variety of reasons, but I do not see how the distinction between two versions of self-defence, put forward by Dr Trapp, can be read into their broad language. More importantly, this distinction is difficult to bring in line with the wording and context of Article 51 of the UN Charter. Its main problems may be put in the form of four questions: (i) How can the differentiated interpretation of the ‘armed attack requirement’ – sometimes requiring attribution, sometimes not – be brought in line with the ordinary meaning of Article 51 of the Charter, which seems to treat ‘armed attack’ as one concept? (I accept of course that there are arguments for reading ‘armed attack’ to mean ‘armed attack by another state’, just as there are arguments supporting the broader ‘armed attack irrespective of its author’.21 But can both readings be maintained at the same time?) (ii) If self-defence version 2 permits defensive force if necessary, why is it then necessary to insist on attribution for self-defence version 1 – could this not also be solved by applying the necessity test? (iii) Why should Article 51 of the Charter draw a distinction between different forms of self-defence if both involve – as Dr Trapp recognizes – infringements of another state's territorial integrity and violate Article 2(4) of the Charter? (iv) Finally, more pragmatically than conceptually, how can the seemingly clear distinction between self-defence version 1 and self-defence version 2 (with the different legal standards it implies) be meaningfully applied to self-defence operations targeting terrorists operating from within state installations or in other ways integrated into state structures? I am sure these questions can be answered. But I believe that when answering them one is very likely to muddy the conceptual waters a lot more than by modifying standards of attribution. (iv) There is a second aspect to Dr Trapp's argument. She claims that her approach avoids ‘mischief’ to the law of self-defence and the law of state responsibility. But what form of mischief does she claim to avoid? With respect to self-defence proper, she provides just one hint. She is critical of using the modified attribution framework because of ‘all this implies about “who done it”’.22 But attribution is not crime fiction and states are no sleuths seeking to find the murderer; they seek justification for conduct which prima facie seems illegal. And in this respect, it seems Dr Trapp and I would require them to address the same issues: we would both admit necessary and proportionate measures of self-defence if the host state was complicit. Under one approach, victim states could avail themselves of the flexibility of a modified attribution standard; under the other, they could invoke self-defence version 2. I do not see how ‘who done it?’ questions should affect the debate. Dr Trapp has a stronger case when warning of unwanted effects that a modified attribution standard may produce in the field of responsibility. She is right to underline the distinction between responsibility for complicity and responsibility for wrongful, attributable conduct, which my article glosses over too hastily.23 But she is equally right to recognize that there is no necessary link between attribution for the purposes of Article 51 and attribution in a state responsibility context. A more careful use of terminology (e.g. stressing that complicity was a sufficient basis for attribution for the purposes of self-defence) can easily help avoid the alleged ‘distorting effect’. In any event, this effect (even if considered distorting) would be much more limited than Dr Trapp asserts. Contrary to what she seems to suggest, the ILC's provisions on attribution are not exhaustive. States are free to agree on stricter standards of attribution without thereby ‘collaps[ing] a primary rule into a secondary rule of state responsibility’.24 And they often do so – Article 91 of the first Additional Protocol to the Geneva Conventions is a prominent example. If they do, they deviate from a residual framework and should not be accused of ‘destabiliz[ing] conceptual clarity’.25 Nobody disputes the importance and relevance of the ILC's text. But the real ‘stroke of genius’26 (if any) of the text is its flexibility – it allows for diversity in the primary rules and provides for ‘unity light’ rather than strict uniformity.27 Nothing in the ILC's Articles would prevent the emergence of stricter standards of attribution in one specific field such as self-defence. (v) In her 2007 article in the International and Comparative Law Quarterly, Dr Trapp argued ‘that a middle ground can be found between the two extremes that either hold that terrorist attacks must be attributable [under strict Nicaragua standards] to a State before …, or posit that a right to use defensive force against non-State actors exists irrespective of the territorial State's non-involvement’.28 I entirely agree. Dr Trapp's comment presents one way of charting this middle ground; my article presents another. As often, those exploring middle grounds cannot lay claim to conceptual clarity. Dr Trapp rightly criticizes me for muddying conceptual waters by lowering the standard of attribution. It seems to me that when noting the splinter in someone else's eye, she may have overlooked the beam in her own. Yet when looked at from a distance – and I would wish to conclude on this note – our approaches have a lot more in common than might appear. They both seek to explain recent anti-terrorist practice within the framework of an inter-state concept of self-defence. They proceed from the same starting-point and reach similar results. Within the current debates about the function and scope of self-defence against non-state attacks, I would characterize them as part of the same strand of thinking.