1. Introduction
In their article in a recent issue of this journal,Footnote 1 Jörn Griebel and Milan Plücken discuss the state responsibility aspects of the judgment of the International Court of Justice (ICJ) in the Genocide case,Footnote 2 and find the judgment wanting on several counts. In their view, the Court committed several grievous methodological errors. It erred when applying the Articles on State Responsibility (ASR) of the International Law Commission (ILC)Footnote 3 and the customary law of state responsibility, since it (supposedly) did not consider the rule set out in Article 8 ASR, which makes states responsible for acts committed under their instructions, direction, or control, to be a proper rule of attribution. It erred when it held that the complete dependence and control test that it had previously articulated in the Nicaragua caseFootnote 4 was the only rule capable of assigning de facto organ status to non-state actors. In Griebel and Plücken's view, the Court erred finally when it failed to take into account recent developments in state practice that point to more flexible rules of attribution.
Much of the critique that Griebel and Plücken make against the Court is reasonable and insightful. This is especially so with their policy arguments, to which I shall turn last. However, much of their criticism is based on what can only be termed their own idiosyncratic reading of the Genocide judgment. Many of the Court's supposed arguments that they proceed to demolish with aplomb are in fact nothing but straw men they themselves have created. At times their analysis also suffers because of their failure to appreciate the broader context of the Genocide case, the arguments of the parties, and the evidence that they actually presented to the Court.
The alleged faults of the ICJ's judgment that Griebel and Plücken dissect and identify are methodological in nature. It is precisely because of their potentially broad implications that I believe that a response to Griebel and Plücken would be helpful, as these questions will undoubtedly be rehearsed over and over again. Indeed, on 12 August 2008 Georgia filed an application with the Court alleging that Russia is responsible for years of violations of the Convention on the Elimination of All Forms of Racial Discrimination (CERD).Footnote 5 As with the Genocide case, this case will also largely turn (if it proceeds to the merits) on Russia's responsibility for the acts of what are prima facie non-state actors – the separatist Georgian entities of Abkhazia and South Ossetia, and certain paramilitary groups. Questions of state responsibility for acts of non-state actors are thus alive both before the Court and outside it, and the Genocide judgment, as the Court's most recent foray into the issue, is a crucially important precedent.
In addressing Griebel and Plücken's critique of the judgment, I shall proceed in the order of their own argument. First to be discussed is Article 8 ASR.
2. Attribution under Article 8 ASR and de facto organs
In the Genocide judgment the ICJ first established that, according to evidence presented to it, the only instance of genocide committed during the conflict in Bosnia and Herzegovina was the July 1995 massacre of some 8,000 Bosnian Muslim men and boys in Srebrenica at the hands of the Bosnian Serb army. The Court then proceeded to establish whether Serbia was responsible for the acts of the Bosnian Serbs in Srebrenica, and did so by using a two-step approach.
It first posed the question whether the Srebrenica genocide was perpetrated by organs of Serbia, and responded to that question in the negative. In the Court's view, none of the perpetrators of the genocide enjoyed organ status under Serbian internal law, per Article 4(1) ASR.Footnote 6 That, however, did not exhaust the Court's enquiry into the possible organ status of the genocidaires, as it remained to be established whether the Republika Srpska and its armed forces were Serbian organs de facto:
The first issue raised by this argument is whether it is possible in principle to attribute to a State conduct of persons – or groups of persons – who while they do not have the legal status of State organs, in fact act under such strict control by the State that they must be treated as its organs for purposes of the necessary attribution leading to the State's responsibility for an internationally wrongful act. The Court has in fact already addressed this question, and given an answer to it in principle, in its Judgment of 27 June 1986 in the case concerning Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States of America) (Merits, Judgment, I.C.J. Reports 1986, pp. 62–64). In paragraph 109 of that Judgment the Court stated that it had to
‘determine . . . whether or not the relationship of the contras to the United States Government was so much one of dependence on the one side and control on the other that it would be right to equate the contras, for legal purposes, with an organ of the United States Government, or as acting on behalf of that Government’ (p. 62).Footnote 7
Thus it was to be determined whether the relationship between Serbia on the one hand, and the Bosnian Serb entity, the Republika Srpska, on the other, was so much of (complete) dependence on one side and control on the other as to render the Bosnian Serbs de facto organs of Serbia. The Court found that under the available evidence the requirements of this test were not met, since the Bosnian Serbs retained a significant amount of autonomy from Serbia.Footnote 8
After finding that the Bosnian Serbs could not be considered to have been organs of Serbia either de facto or de jure, the Court went on to the second step of its approach, by examining whether the specific operation in Srebrenica was conducted under the direction or the ‘effective control’ of Serbia (this test being previously formulated in para. 115 of Nicaragua), pursuant to Article 8 ASR. It responded to this question in the negative as well, finding that, indeed, there was no evidence that Serbia even knew of the Srebrenica genocide while it was under way, let alone that it was controlling it.Footnote 9
So far so good, but now we come to Griebel and Plücken's most fervent, and least justified, criticism of the Court. In their view, by the way in which it applied the effective control test, the Court abolished Article 8 ASR as a proper rule of attribution, indeed it ‘did not regard [it] as an attribution rule at all’.Footnote 10 How so? Griebel and Plücken quote that portion of paragraph 397 of the Court's judgment stating that
An affirmative answer to this question [that of whether the requirements of Art. 8 ILC Articles are fulfilled] would in no way imply that the perpetrators should be characterized as organs of the FRY [Federal Republic of Yugoslavia], or equated with such organs. It would only mean that the FRY's international responsibility would be incurred owing to the conduct of those of its own organs which gave the instructions or exercised the control resulting in the commission of acts in breach of its international obligations.Footnote 11
In their view,
The plain reading of the wording is that the ICJ is of the opinion that the responsibility incurred by the state under Article 8 of the ILC Articles flows from the conduct of the state's organs in giving the instructions or exercising the control in question, as opposed to the action of the instructed or controlled entities. Considering that the function of the attribution rules is to attribute to the state the conduct of persons who have acted against international law, the Court's refusal to consider the persons acting under such instructions or control as de facto organs, and its foundation of responsibility in Article 8 situations on the wrongfulness of the state organs’ instructions or control, entirely stripped Article 8 of the ILC Articles of its character as an attribution rule.Footnote 12
Griebel and Plücken then spend several pages of their article explaining (quite correctly) that Article 8 ASR is in fact an attribution rule, and that any different conception of Article 8 would be entirely mistaken.Footnote 13 As they interpret the Genocide judgment, the Court thought of Article 8 as some sort of primary rule, as a separate wrongful act of directing or controlling the perpetrators of genocide, distinct from the commission of genocide as such.Footnote 14 They express their surprise that so far none of the academic commentators of the Court's judgment have spotted this quite glaring error in the Court's reasoning.Footnote 15 Griebel and Plücken are further dismayed by the fact that none of the Court's judges, at least five of whom were members of the International Law Commission (ILC) who participated in its work on the ASR,Footnote 16 criticized the Court's approach to Article 8 ASR.Footnote 17 In their minds, the only explanation for the judges’ silence on the matter is that they did not ‘fully realize’ that the Court was actually abolishing Article 8 ASR as a rule of attribution.Footnote 18
Unfortunately, Griebel and Plücken's analysis, substantively correct for the most part as it is, is nonetheless predicated upon two fallacies, the first being purely one of interpretation. Contrary to Griebel and Plücken, the Court never actually said that it thought of Article 8 ASR as anything other than a rule of attribution. Indeed, in the introductory paragraph in which it explains the steps of its analysis of Serbia's responsibility, the Court says that
Having thus explained the interrelationship among the three issues set out above (paragraph 379), the Court will now proceed to consider the first of them. This is the question whether the massacres committed at Srebrenica during the period in question, which constitute the crime of genocide within the meaning of Articles II and III, paragraph (a), of the Convention, are attributable, in whole or in part, to the Respondent. This question has in fact two aspects, which the Court must consider separately. First, it should be ascertained whether the acts committed at Srebrenica were perpetrated by organs of the Respondent, i.e., by persons or entities whose conduct is necessarily attributable to it, because they are in fact the instruments of its action. Next, if the preceding question is answered in the negative, it should be ascertained whether the acts in question were committed by persons who, while not organs of the Respondent, did nevertheless act on the instructions of, or under the direction or control of, the Respondent.Footnote 19
Moreover, the heading of the section of the judgment that contains paragraph 397 from which Griebel and Plücken construe their entire understanding of the judgment (which is in fact located on the same page), reads, ‘(4) The question of attribution of the Srebrenica genocide to the Respondent on the basis of direction or control’.Footnote 20 On the very next page, the Court says explicitly that
The rules for attributing alleged internationally wrongful conduct to a State do not vary with the nature of the wrongful act in question in the absence of a clearly expressed lex specialis. Genocide will be considered as attributable to a State if and to the extent that the physical acts constitutive of genocide that have been committed by organs or persons other than the State's own agents were carried out, wholly or in part, on the instructions or directions of the State, or under its effective control. This is the state of customary international law, as reflected in the ILC Articles on State Responsibility.Footnote 21
The Court then goes on to say that
[A] State is responsible only for its own conduct, that is to say the conduct of persons acting, on whatever basis, on its behalf. That is true of acts carried out by its official organs, and also by persons or entities which are not formally recognized as official organs under internal law but which must nevertheless be equated with State organs because they are in a relationship of complete dependence on the State. Apart from these cases, a State's responsibility can be incurred for acts committed by persons or groups of persons – neither State organs nor to be equated with such organs – only if, assuming those acts to be internationally wrongful, they are attributable to it under the rule of customary international law reflected in Article 8 cited above.Footnote 22
How the judgment of the Court, which should, like any legal instrument, be read reasonably, as a single whole, so as to avoid manifestly absurd results, could be interpreted as saying anything other than Article 8 ASR is a rule of attribution (let alone plainly so, as Griebel and Plücken doFootnote 23) is beyond the present author. Article 8 ASR is obviously a rule of attribution, beyond the shadow of a doubt, and the Court says nothing to the contrary.
But this brings us to Griebel and Plücken's second, much more profound, fallacy. Namely, although the Court did not say that Article 8 ASR was not a rule of attribution, it did say that Article 8 could not assign de facto organ status to a non-state actor which the state did not control completely.Footnote 24 However, in Griebel and Plücken's mind, attribution and organ status are apparently one and the same – if the conduct of an actor is attributable to a state, then it by definition must be a state organ, either de jure or de facto. Thus they say, for example, that ‘the Court's refusal to consider the persons acting under such instructions or control as de facto organs, and its foundation of responsibility in Article 8 situations on the wrongfulness of the state organs’ instructions or control, entirely stripped Article 8 of the ILC Articles of its character as an attribution rule’,Footnote 25 and that ‘one can – contrary to the Court – see Article 8 of the ILC Articles regarding questions of terminology as an attribution rule for de facto organs.’Footnote 26 Indeed, in their view, such a conception of Article 8 is supported both by the ILC and by unanimous academic authority.Footnote 27
As one of the authors cited in evidence of this apparent unanimity,Footnote 28 I can only speak for myself in saying that I emphatically do not share Griebel and Plücken's view that any actor whose conduct is attributable to a state under the customary rules of responsibility codified in the ASR is ipso facto one of the state's organs. There is consensus only on the (again completely obvious) point that Article 8 ASR is a rule of attribution. In its commentary to the ASR, the ILC actually never even uses the term ‘de facto organs’, so it is hard to see how Griebel and Plücken can invoke it in support of their position that attribution under Article 8 equals de facto organ status. Indeed, the ILC reserves the use of the term ‘organ’ solely to attribution under Article 4 ASR, and never uses it in the context of attribution under Articles 8–11 ASR.Footnote 29 The ILC moreover clearly does not view the category of organs to be identical with all persons whose acts are attributable to a state. Thus, for example, it speaks of the general rule which is that ‘the only conduct attributed to the State at the international level is that of its organs of government, or of others who have acted under the direction, instigation or control of those organs, i.e., as agents of the State’.Footnote 30
The notion of de facto organs is of course certainly used in the literature, often in a wider sense than that given to the notion by the ICJ. Even more often, however, authors who use the term do not define it, as if its meaning were self-evident. Neither do Griebel and Plücken, for that matter, although of course for them the categories of attribution and organ status appear to be interchangeable. I myself am probably equally guilty of the sin of failing to define clearly the concepts that I am using.Footnote 31 I can thus only explain my views on the issue. De facto organs (i) are entities that are (obviously) not de jure organs; but (ii) nonetheless, by virtue of the strength of the connection between them and the state, the rules that apply to de jure organs apply to them as well, even if they do not possess organ status under the state's domestic law. This is indeed, in my opinion, how the Court itself sees the concept. In its view, the acts of a state's organs are ‘necessarily attributable to it, because they are in fact the instruments of its action’.Footnote 32 In other words, it is the identity of the actor as a state organ that alone suffices for attribution to occur.Footnote 33 No other connection of the state to the act in question is needed.
By contrast, acts of other persons may be attributable to the state, but not necessarily. For the attribution to a state of an act of non-state organs to occur, a further link between the state and that actor is required – its mere identity is not enough. That is indeed the whole point of the attribution rules articulated in Articles 8–11 ASR, which, in the words of the ILC, ‘deal with certain additional cases where conduct, not that of a State organ or entity, is nonetheless attributed to the State in international law’.Footnote 34 It is precisely because the conduct is not that of a state organ, either de jure or de facto, that the law requires, for example, proof of a state's instructions, or its control over a specific act, or its positive acknowledgement of an act as its own.
Griebel and Plücken are of course free to define the term ‘de facto organ’ as they see fit, even so as to encompass all persons whose acts are attributable to the state,Footnote 35 but they cannot fault the Court for simply operating under a different definition. Thus, when the Court holds that, under Article 8 ASR, a state's ‘responsibility would be incurred owing to the conduct of those of its own organs which gave the instructions or exercised the control resulting in the commission of acts in breach of its international obligations’,Footnote 36 it is not saying that the acts in question would not be attributable to the state, or that the conduct of the state's organ in issuing their instructions or exercising their control would amount to a distinct wrongful act, as Griebel and Plücken seem to understand it. The Court is merely explaining the reason why the state is being held responsible, that is, what the state has done that renders it responsible for the wrongful act of a non-state entity – the rationale behind the rule attributing to a state an act that was not committed by one of its organs.Footnote 37 Likewise, under Article 11 ASR, an act can be attributed to a state if and to the extent that it acknowledges and adopts it as its own, even if it would otherwise not be imputable to it. That does not mean that the wrongful act is the state's acknowledgement as such.
The distinction made above between state organs (de jure or de facto) and other actors whose acts are attributable to the state is not merely theoretical or a matter of semantics, but has great practical consequences. These consequences are less in the applicable rules of responsibility, once responsibility is in fact established. Serbia's responsibility for the Srebrenica genocide, and its obligation to provide reparation, would have been completely the same whether the genocide was imputed to it under a de facto organ theory or under Article 8 ASR. The only purely legal difference is that the rule on the attribution of ultra vires acts in Article 7 ASRFootnote 38 by its own terms applies only to a state's organs or persons empowered by it to exercise elements of government authority. That rule conceptually cannot and does not apply to the attribution of acts by non-state organs under Articles 8–11 ASR, which are not based solely on the identity of the actor, but require positive proof of state control, instructions, or the adoption of the specific act that is being attributed. In the words of the ILC,
As formulated, article 7 only applies to the conduct of an organ of a State or of an entity empowered to exercise elements of the governmental authority, i.e. only to those cases of attribution covered by articles 4, 5 and 6. Problems of unauthorized conduct by other persons, groups or entities give rise to distinct problems, which are dealt with separately under articles 8, 9 and 10.Footnote 39
The greatest practical difference between attribution under a de facto organ theory and attribution under Articles 8–11 ASR lies not in the fact that Article 7 ASR applies to the former but not to the latter, but in the steps that one needs to actually take to prove, under the applicable evidentiary standard, the facts necessary for attribution to occur. As stated above, one does not have to prove state instructions or control over a specific wrongful act committed by an organ. All that is necessary is to prove that the persons acted in their official capacity. By contrast, non-state actors under Articles 8–11 ASR do not have an official capacity. For attribution of their acts to the state, it is necessary to prove, for instance, that the state actually instructed them or controlled them in the commission of the specific act concerned,Footnote 40 or later adopted the act as its own.
To translate this abstract discussion to the facts of the Genocide case, except for the Scorpions paramilitary unit, whose relationship with Serbia neither the Bosnian party nor the Court proprio motu explored fully,Footnote 41 Bosnia was unable to provide the Court with any evidence that Serbia was directly involved in the Srebrenica operation. It simply could not prove, and certainly not to the (justifiably) high evidentiary threshold that the Court had imposed, that Serbia directed or controlled the operation or even knew about the genocide in Srebrenica while it was under way. The only way that the genocide could have been attributed to Serbia was under a de facto organ theory – that is, that the Republika Srpska and its army were, as a general matter, organs of Serbia, and would thus have engaged its responsibility even in the absence of evidence of Serbia's involvement in the specific operation. Bosnia indeed provided the Court with significant evidence of Serbia's control and influence over the Republika Srpska at a general level.Footnote 42 That evidence, however, did not satisfy the rigorous complete dependence and control test that the Court employed for establishing de facto organ status. I shall now turn to Griebel and Plücken's critique of that test.
3. De facto organs and the complete control test
As we have seen above, in Griebel and Plücken's view the complete dependence and control test that the Court employed in Nicaragua and the Genocide case for determining de facto organ status is defective almost by definition, since they consider any actor whose acts are attributable to a state to be a de facto organ thereof. That, however, is neither the position of the Court nor that of the ILC, and hence Griebel and Plücken's critique simply misses the mark. Yet some parts of their critique of the complete control test have more bite. They argue that the test is so stringent that it could hardly ever be successfully applied, that it is almost inconceivable that it could ever be satisfied without the effective control test being satisfied at the same time, and that it is therefore redundant, or at the very least of negligible practical importance.Footnote 43 They doubt, moreover, that the complete control test can fit within Article 4(2) ASR, as did the Court in the Genocide judgment, since in their view only Article 8 ASR can govern control-based attribution.Footnote 44 Let us now examine each of these arguments in turn.
First, to see whether the complete control test is overly demanding it is necessary to see why the test is as strict as it is. As the Court explains it,
[P]ersons, groups of persons or entities may, for purposes of international responsibility, be equated with State organs even if that status does not follow from internal law, provided that in fact the persons, groups or entities act in ‘complete dependence’ on the State, of which they are ultimately merely the instrument. In such a case, it is appropriate to look beyond legal status alone, in order to grasp the reality of the relationship between the person taking action, and the State to which he is so closely attached as to appear to be nothing more than its agent: any other solution would allow States to escape their international responsibility by choosing to act through persons or entities whose supposed independence would be purely fictitious. . . . However, so to equate persons or entities with State organs when they do not have that status under internal law must be exceptional, for it requires proof of a particularly great degree of State control over them, a relationship which the Court's [Nicaragua] Judgment quoted above expressly described as ‘complete dependence’.Footnote 45
Thus it is the exceptional nature of assigning organ status to an entity which does not have such status under a state's domestic law that warrants a strict test. Doctrinally, the purpose of the test is to prevent states from avoiding responsibility merely by changing the provisions of their own domestic law.Footnote 46 As explained above, the reason why a connection between a state and a de facto organ must be intense is that the mere identity of the actor as a state organ suffices for attribution to occur.
To that Griebel and Plücken respond that it is hard to envisage the complete control test being satisfied without a simultaneous proof that the specific acts of the entity in question were at the same time effectively controlled by the state. This is so because it is necessary to prove that an organ acted in an official capacity for attribution to occur (something that is not at all in doubt, and that both the ICJ and the ILC naturally accept).Footnote 47 According to Griebel and Plücken, attribution based solely on general considerations of an actor's identity as a de jure or de facto state organ is questionable:
The general relationship which exists based on a formal link is in itself not enough; it must be reflected in the action itself. The person must have acted within its capacity, which can only be examined focusing specifically on the act itself.Footnote 48
In their view, ‘[d]etermining attribution solely on the basis of the general relationship between the state and the person or group and ignoring the relevance of this link concerning the act in question is conceptually doubtful’.Footnote 49
What Griebel and Plücken do not seem to realize is that there is a world of difference between proving that a person acted in the official capacity of a state organ and proving that the conduct in question was effectively controlled by the state. To take the facts of the Srebrenica genocide as an example, a Bosnian Serb soldier engaging in combat, massacring civilians, and raping Bosnian Muslim women is easily distinguished from, say, a soldier who in private molests the members of his own family. There is no doubt that the soldiers of the Bosnian Serb army who perpetrated the Srebrenica genocide were part of an organized military force of the Bosnian Serb pseudo-state, were acting on the orders of their commanders, and were hence not engaging in some sort of purely private conduct. Proving that Serbia actually controlled the commission of the genocide is another matter altogether. Likewise, were the Scorpions, for example, proved to have been de jure or de facto members of the Serbian (secret) police, that would have been the end of the attribution enquiry. There would have been no need for Bosnia to prove that the Serbian leadership instructed, directed, or controlled the Scorpions in their actions in Srebrenica.
Griebel and Plücken are thus simply mistaken when they say that a test focusing on the general relationship between a state and a non-state actor, rather than on the specific act that is being attributed, is ‘conceptually doubtful’.Footnote 50 Indeed, the overall control test that was announced by the International Criminal Tribunal for the former Yugoslavia (ICTY) Appeals Chamber in Tadić and that has garnered its share of supporters – among them Judge Al-Khasawneh in his dissenting opinion in the Genocide case,Footnote 51 and also seemingly, but not unequivocally, Griebel and PlückenFootnote 52 – is exactly like the complete control test in that it looks at a state's control over an actor at a general level. Where the two tests differ is in the intensity of the connection required. Thus, pursuant to the complete control test, it is necessary to establish that the relationship between a state and a non-state actor is one of complete ‘dependence on the one side and control on the other’, so that the non-state actor is in fact a mere instrument of the state, from which it has no real independence or autonomy.Footnote 53 By contrast, to satisfy the overall control test, it must be proven that the state
wields overall control over the group, not only by equipping and financing the group, but also by coordinating or helping in the general planning of its military activity. . . . [I]t is not necessary that, in addition, the State should also issue, either to the head or to members of the group, instructions for the commission of specific acts contrary to international law.Footnote 54
Hence, like the complete control test, the overall control test requires no proof of direct involvement of the state in the specific act that is being attributed – the overall control is not control over the act, but over the actor, an organized and hierarchically structured group, at a general level.Footnote 55 That, of course, makes perfect sense. One either controls the commission of a specific act effectively, or does not control it at all.
Reasonable persons can of course disagree as to whether identity- or organ-status-based attribution should require a higher or lower degree of control. However, one cannot support the overall control test but at the same time doubt the appropriateness or the utility of the complete control test because of its generality, since generality is the single greatest virtue of both of these tests. As to whether the complete control test is so stringent that it is of negligible practical importance, as Griebel and Plücken say,Footnote 56 we should abstain from passing such judgement easily without first examining the Court's assessment of the actual facts of the Genocide case. In the opinion of the Court, in regard of Serbia's responsibility for the Srebrenica genocide,
At the relevant time, July 1995, neither the Republika Srpska nor the VRS [Vojska Republike Srpske, the military forces of the Republika Srpska] could be regarded as mere instruments through which the FRY was acting, and as lacking any real autonomy. While the political, military and logistical relations between the federal authorities in Belgrade and the authorities in Pale, between the Yugoslav army and the VRS, had been strong and close in previous years (see paragraph 238 above), and these ties undoubtedly remained powerful, they were, at least at the relevant time, not such that the Bosnian Serbs’ political and military organizations should be equated with organs of the FRY. It is even true that differences over strategic options emerged at the time between Yugoslav authorities and Bosnian Serb leaders; at the very least, these are evidence that the latter had some qualified, but real, margin of independence. Nor, notwithstanding the very important support given by the Respondent to the Republika Srpska, without which it could not have ‘conduct[ed] its crucial or most significant military and paramilitary activities’ (I.C.J. Reports 1986, p. 63, para. 111), did this signify a total dependence of the Republika Srpska upon the Respondent.Footnote 57
It is crucial to recall again that because the case was jurisdictionally strictly confined to genocide, and because the only instance of genocide during the whole Bosnian conflict that the Court could establish was Srebrenica, the Court's analysis of Serbia's responsibility was confined solely to the Srebrenica events in July 1995. However, the amount of control that the authorities of Serbia exercised over the Bosnian Serbs was not equal for the duration of the entire conflict. Indeed, it lessened over time, as rifts between the Serbian leadership and that of the Bosnian Serbs became more and more pronounced. In 1992, which was the bloodiest year of the war by far, the Bosnian Serb campaign of ethnic cleansing was at its peak, with more than half of all civilian casualties of the entire conflict occurring in just a couple of months of that year.Footnote 58 That is when Belgrade's control over the Bosnian Serbs was at its highest, and when the former Yugoslav National Army (JNA) was still in effect operating in Bosnia. The paragraph from the Court's judgment quoted above is filled with caveats that were designed precisely to address the fluctuating nature of Serbia's control over the Bosnian Serbs, and to limit the Court's holding to one point in time, July 1995. The Court all but hinted at the possibility that its demanding complete control test would indeed have been met for events taking place in 1992, which unfortunately fell outside its jurisdictional purview. In other words, far from being made up of judges who do not ‘fully realize’ what they are doing, the Court was ever so careful, and the complete control test was hardly of negligible import. That test could rear its head again in the recently filed Georgia v. Russia case, where the Court will have to assess Russia's responsibility for the acts of Abkhazian and South Ossetian separatist entities.
Griebel and Plücken's last piece of purely doctrinal criticism is perhaps their most persuasive. Namely, they argue that in its work on the ASR, the ILC never considered control-based tests of attribution when it formulated Article 4, dealing with attribution by virtue of organ status.Footnote 59 Contrary to the Court, which placed the complete control test for establishing de facto organ status within the framework of Article 4 ASR,Footnote 60 Griebel and Plücken argue that such a test can only be subsumed under Article 8 ASR.Footnote 61
Griebel and Plücken may be right when they say that the ILC never intended Article 4(2) ASR's definition of state organs as persons or entities that include, but are not exhausted, in those persons or entities which enjoy such status under domestic law, to have as broad a scope as was given to it by the Court in the Genocide case. As for the Nicaragua complete control test, although the correct interpretation of Nicaragua as setting out two, instead of one, attribution tests was espoused by Judge McDonald in Tadić, it was certainly missed by the ILC in its work on state responsibility. The complete control test was likewise conflated with the effective control test by no less eminent a jurist than Antonio Cassese and the ICTY Appeals Chamber over which he presided in Tadić, probably in no small part due to the ICJ's less than clear drafting in Nicaragua.Footnote 62 Nonetheless, the Court is the authoritative interpreter of its own jurisprudence. It is true that, in the words of one eminent commentator, ‘Crawford's rules rock’,Footnote 63 but that does not mean that the ASR are an infallible exposition or the sum total of the law on state responsibility. If the Court had had the opportunity to clarify its holding in Nicaragua before the drafting of the ASR was completed, as it later did in the Genocide case, the ILC would undoubtedly have taken that into account. Although the ILC might not have envisaged a broader conception of de facto organs that would describe, for example, puppet states such as Republika Srpska, the text of Article 4(2) ASR is wide enough to encompass such a notion, as are indeed the ILC's commentaries.Footnote 64 It is worth noting that none of the former ILC members sitting on the Court's bench saw anything incompatible between the complete control test for the purpose of establishing de facto organ status and Article 4 ASR.
Last to be examined are Griebel and Plücken's policy and de lege ferenda arguments, to which I now turn.
4. Are the Court's two tests good enough for the modern world?
The most important question raised by Griebel and Plücken is not whether the Court's dual approach to attribution through the complete control and the effective control tests is doctrinally sound, but whether these tests are what the international community, faced with present-day problems, truly needs. There is, of course, much room here for reasonable disagreement, but I would like to comment on what are in my view some problematic aspects of Griebel and Plücken's policy arguments.
First, they express their regret that in its judgment the Court took no account of possible developments in the law of state responsibility subsequent to the 11 September 2001 terrorist attacks.Footnote 65 Namely, a possibility was raised in the literature that the response of the international community to these attacks, specifically the practically unanimous belief that the US invasion of Afghanistan was a lawful exercise of self-defence, indicated the birth of a new rule that a state which harbours terrorists – that is, allows them to operate from within its territory unimpeded – is responsible for any and all acts of such groups.Footnote 66 Indeed, President Bush issued a declaration stating such a doctrine in so many words.Footnote 67 This argument is predicated on an interpretation of Article 51 of the UN Charter as requiring not only that an armed attack be committed against a state, but also that that attack be attributable to another state against which the attacked state is to respond. Since the US invasion of Afghanistan was considered to be a lawful exercise of self-defence, so the argument goes, and since Article 51 requires attribution, the state of Afghanistan must have been held responsible by the international community for the 11 September attacks. As these attacks could not have been attributed to Afghanistan under any of the rules of attribution articulated in the ASR, a new rule of attribution – the harbouring of terrorists – must have emerged.
This nice syllogism is what Griebel and Plücken have in mind when they express their regret at the Court's failure to address it. This nice syllogism is also unfortunately troubled by a potentially flawed premise, which is that Article 51 of the Charter actually requires attribution of an armed attack to a state. The text of Article 51 simply does not do so – it speaks only of an armed attack, not of an armed attack by or attributable to a state. It could just as easily be read to allow action in self-defence against a state which is either unwilling or unable to prevent armed attacks by non-state actors operating from its territory, as at the time Afghanistan certainly was.Footnote 68 This debate on the two competing interpretations of Article 51 is indeed one of the major controversies of the contemporary jus ad bellum, frequently recurring in practice (as with the recent actions of Israel against Hezbollah in Lebanon, Turkey against the Kurdistan Workers’ Party (PKK) in Iraq, or Colombia against the Revolutionary Armed Forces of Colombia (FARC) in EcuadorFootnote 69), which is far from settled either wayFootnote 70 and which Griebel and Plücken fail even to mention, even though resolving this controversy of the jus ad bellum is logically a necessary prerequisite for deciding on harbouring as a rule of attribution.
In other words, far from it being clear that the 11 September attacks redefined the rules of attribution, it could be that they had no impact whatsoever on the law of state responsibility. On the other hand, it could also be that we are currently in some sort of legal limbo, waiting for the affirmation of a new rule of attribution – the Bush doctrine on harbouring – through subsequent state practice. It would not be the first such time.Footnote 71 In any case, whether harbouring has emerged as a new rule of attribution is of course a fascinating topic for academic enquiry, or even for moot court competitions.Footnote 72 But this is still a far cry from accepting that the ICJ should be condemned for failing to address the issue at this point in time. It should be recalled that, broadly speaking, the ICJ has always had two kinds of judge – those who think that the Court's role is the settlement of the dispute put before it by the parties in each particular case, and those who believe that the Court should also use the opportunities presented to it to go beyond the immediate concerns of the parties in a concrete case and give wide-ranging pronouncements on current issues of international law.Footnote 73 But not even the most ‘activist’ ICJ judge (to use that most blunt of terms) would have gone as far as Griebel and Plücken wanted the Court to go. Some controversies are simply better left to another day.
Griebel and Plücken do indeed say that the Court might have refrained from discussing these questions simply because it was applying the law of state responsibility as it stood when the Srebrenica genocide was committed, not as it might stand today in the post-11 September world.Footnote 74 But there is a further, and even simpler, explanation. Interesting though the issue of harbouring as a potential rule of attribution might be, it was never raised before the Court by the parties. Even Bosnia thought it would have gone too far had it argued that Serbia was responsible for the Srebrenica genocide solely on the account of harbouring genocidaires such as Mladić or the Scorpions. Instead, Bosnia properly argued that such conduct by Serbia would lead (as it did) to its responsibility for failing to prevent and punish genocide. It is of course true that jura novit curia, but the Court cannot have been expected just to go out on a limb and pronounce on issues that were not even hinted at, let alone argued before it, and to do so simply in order to satisfy the curiosity of academic commentators. Indeed, it is precisely the Court's silence on the matter that can allow state practice to develop unimpeded and perhaps eventually produce new tests of attribution.
Griebel and Plücken finally echo some arguments of Antonio CasseseFootnote 75 by saying that, as a matter of policy, the Court's two rigid tests of attribution allow states to easily avoid responsibility by using non-state actors for their more sordid endeavours.Footnote 76 In Griebel and Plücken's view, the international community's need for less restrictive rules of attribution is evident, as is the need for such rules’ deterrent effect on states.Footnote 77 While Cassese sees the Tadić overall control test as an appropriate response to such concerns,Footnote 78 Griebel and Plücken advocate an even looser standard, which would make states responsible for any act in whose preparation or execution they had a ‘substantial involvement’. They further posit that either harbouring or their own substantial involvement test might have evolved to lex specialis rules of attribution applicable only to terrorism.Footnote 79
Again, on these questions of policy there is much room for reasonable disagreement. It must be said, however, that, like Cassese, Griebel and Plücken pose a false dichotomy between a state's responsibility for everything and anything that a non-state actor does and no responsibility at all. It is true, for example, that under the ICJ's Nicaragua/Genocide paradigm the state of Afghanistan could not have been held responsible for the 11 September attacks, as it neither had complete control over al Qaeda at a general level nor instructed or effectively controlled al Qaeda in the commission of the terrorist attacks. (Incidentally, Afghanistan could not be held responsible for the 11 September attacks even under Cassese's overall control test – if anybody was doing the controlling, it was al Qaeda. It is only under a harbouring theory that attribution could have occurred.) That does not mean that Afghanistan bore no responsibility whatsoever. It was responsible for separate internationally wrongful acts of supporting a terrorist group and allowing it to use its territory to cause harm to other states.Footnote 80 Indeed, its failure to exercise due diligence to prevent armed attacks emanating from its territory resulted in an invasion in self-defence by the attacked state. Should we truly have considered Afghanistan to have been responsible for the 11 September attacks in the same way as if they had been committed by, say, special forces of the Afghan army? I think not. The lesson from this example, and the ICJ's finding of no state responsibility for the commission of genocide by Serbia, but of a breach of the primary obligation to prevent and punish genocide, is that attribution is not everything. States with substantial involvement in the wrongful acts of non-state entities will often be ‘caught’ by other primary rules of international law.
Griebel and Plücken's false dichotomy is reinforced by the warped jurisdictional perspective of the Genocide case, which again was in the end not about the totality of the Bosnian conflict, but solely about the July 1995 genocide in Srebrenica. As a substantive matter, Serbia was responsible for numerous wrongful acts, such as intervention in and the unlawful use of force against Bosnia, the support and financing of Bosnian Serb forces, the sending of paramilitaries, and so forth. Unfortunately, all of these wrongful acts fell outside the ICJ's jurisdiction, which was confined solely to violations of the Genocide Convention.Footnote 81 Even with regard to the Srebrenica genocide, although Serbia did avoid responsibility for the commission of the genocide, it was indeed found responsible for the distinct wrongful acts of failing to prevent and punish the genocide. To this Griebel and Plücken could, of course, object that there is little use in the separate obligation to prevent genocide if the only consequence of a breach of that obligation is a declaratory judgment of an international court, however eminent. There would be some truth in that objection. However, it should again be borne in mind that the declaratory remedy in the Genocide case, flawed and unsatisfactory as it was, was precisely what counsel for Bosnia itself asked for.Footnote 82 It should not be taken as a general statement by the Court that a state's failure to fulfil its due diligence obligations requires no more reparation than some sort of purely symbolic satisfaction.
As a broader matter, Griebel and Plücken's policy argument just begs the question whether it is indeed the purpose of the secondary rules of state responsibility to deter, say, state support of terrorism or, as Griebel and Plücken put it, prevent states from ‘conspiring with private groups for the purpose of violating other state's rights’.Footnote 83 In my mind at least, that is what primary rules are for, such as the one that states must not allow their territory to be used to harm other states. Even if deterrence of this sort is seen as the purpose of the rules of state responsibility, it is entirely questionable whether the deterrent effect would be greater if all acts of a non-state actor were to be attributed to a state that supports it, no matter how tenuous the connection, as opposed to the current state of the law where a state is instead held responsible for its own acts proper. Whether this supposed deterrent effect is worth fragmenting a general framework by fashioning lex specialis rules of attribution, as Griebel and Plücken seem to argue, is not at all clear.Footnote 84 It is equally debatable whether Griebel and Plücken's proposed rules of responsibility describe today's realities better than the rules articulated by the ILC and the ICJ. Far from accounting for the increasing importance of non-state actors in the modern world, Griebel and Plücken actually diminish the political importance and autonomy of these actors by fictitiously turning them all into state agents.
At any rate, the outcome of this policy debate lies in the years and decades to come. For my part, I can express my doubts that Griebel and Plücken's approach will withstand the test of time any better than the Genocide judgment. It is beyond question that the world faces serious challenges. These challenges, however, are better dealt with, both as a matter of law and as a matter of policy, by adapting the relevant primary rules. It is the jus ad bellum which should address the use of force by non-state actors, not the law of state responsibility. Stretching the rules of attribution is likewise a poor remedy for situations where the ICJ is possessed only of a limited jurisdiction and is unable to deal with the totality of a dispute, as in the Genocide case. The law of state responsibility can change, like any other part of international law. For the time being, however, I see no evidence that the law as it stands is incapable of dealing with contemporary concerns, be they related to global terrorism or otherwise.
5. Conclusion
When it comes to what Griebel and Plücken see as the Genocide judgment's greatest flaw – that is, its application of Article 8 ASR – it can only be said that they have criticized not the Court's judgment itself, but their own caricature of that judgment. Is theirs a possible reading of the judgment? Perhaps. But is it the best and the most sensible? Most assuredly not. Without in any way playing the role of the Court's apologist, I can thus only express my regret that Griebel and Plücken failed to do justice to the Genocide judgment, despite the many important issues of principle that they have raised. That is not to say, of course, that the Court's judgment is free from any serious flaws. It is far from perfect, partly due to the passivity of the Court in regard of certain factual questions and the remedy that it awarded, and partly because of the litigation strategies of the two parties, Bosnia in particular. There are enough things to criticize in the judgment as it actually is, without having to conjure up supposedly glaring errors out of thin air.
With regard to the broader question posed by Griebel and Plücken, whether the Court's exposition of the law of responsibility meets the present needs of the international community, that question must be put in a proper perspective. There is ultimately no dispute that a state will be responsible for all the (official) acts of an entity under its complete control, or for any specific act committed under its effective control. It is logically for those who advocate looser tests of attribution to prove conclusively that state practice and opinio juris support such rules.Footnote 85 Again, international law is not a static thing but an ongoing process. The rules of attribution are, like all its rules, liable to change, but it is yet to be demonstrated that the rules articulated by the ILC and the ICJ have actually changed or have proved to be inadequate. In that regard, much of the current critique of the supposedly great stringency of the rules of attribution is in fact directed at other controversies of international law, such as those pertaining to the jus ad bellum. As argued above, these problems should be addressed on their own merits, not through the medium of state responsibility.
