I. INTRODUCTION
In the shadow of momentous events, the question of responsibility of complicit States has arisen more and more frequently in recent years. The best-known cases concern the implication of third party States in the US-led attack against Iraq in 2003, the roles of European States in the US policy of ‘extraordinary rendition’, and Serbia's involvement in the Srebrenica genocide.
The rules on complicity in international law are not sufficiently developed to allow for clear-cut determinations in all those cases.Footnote 1 It is clear, however, that some of these cases have created political dilemmas. For example, some EU States and Canada, while being close allies of the United States, have been unhappy with some US conduct, in particular with the US-led attack against Iraq (II.). These States nevertheless provided support for the US. What are the legal consequences of such support? Was it furnished solely for political reasons or did it also imply legal endorsement?Footnote 2 In the following section we will address two questions: First, we will consider the rules which determine whether their support renders ‘complicit’ States themselves responsible under international law (III.). Second, we will ask whether such support can simultaneously be counteracted on the law-making or law-interpretive level: can a State express relevant legal objections against the conduct (and its underlying legal assertion) of another State to which it simultaneously contributes aid? Or are such objections legally insignificant (IV.).
We should point out that our main concern is not whether the United States, or any other primary actor, has violated international legal obligations. We assume that in the situations we discuss this was at least arguably the case. Our main concern is the question of responsibility of third party States. This issue becomes more important as States increase and intensify relations in the process of globalization. Relations between States are often so dense that a broad and rigorous rule on complicity would require constant scrutiny by States on whether their conduct which is prima facie ‘neutral’ does not stray into ‘complicity’. What happens if States do not arrive at a clear conclusion with respect to the assessment of the lawfulness of their peers' conduct? Is a State required to withdraw its support if the conduct to which support is given might be unlawful? Or does the rule apply that ‘States should (…) be entitled to presume that other States will act lawfully’?Footnote 3 Would it be enough for the State to refrain from contributing aid or assistance only in those cases in which illegality is clearly established? Conceivable future scenarios might concern even more legally ambiguous situations than the US-led attack on Iraq. Such situations could arise, for example, in the further aftermath of Kosovo's declaration of independence, or with respect to measures to support Pakistan in its attempts to secure its nuclear weapons against unauthorized use,Footnote 4 or the decision whether or not to support an ally which is considering military action against militia or terrorist groups within the territory of a third State.Footnote 5
II. PRACTICAL SUPPORT AND POLITICO-LEGAL OBJECTIONS
When the United States launched their attack on Iraq in 2003 the US government approached its NATO allies for support. While certain allies refused to contribute fighting troops and expressed political objections against this attack, they mostly acceded to US requests for support on a technical level and refrained from expressing an opinion on the legality of the attack.Footnote 6
This is not the place to reopen the debate on the legality of the US-led Iraq invasion. It is probably fair to say that invocations of the right to self-defence were rejected by the overwhelming majority of States and authors as being clearly erroneous and unpersuasive.Footnote 7 On the other hand, the official justifications by the United States and the United Kingdom which alleged an authorization by the Security Council were taken more seriously. Although most authors ultimately did not consider them to be persuasive it was widely accepted that they constituted a serious legal argument.Footnote 8
We use the 2003 Iraq conflict as an illustration: In this case, Italy and Germany, among others, granted over-flight and landing rights for US planes involved in the attack on Iraq. German armed forces guarded US facilities in Germany during the offensive military operations. German warships involved in the anti-terrorist ‘Operation Enduring Freedom’ escorted US ships on their way to the Persian Gulf which subsequently participated in the attacks against Iraq.Footnote 9 Italy declared itself ‘non-belligerent’ which meant, according to the Italian Supreme Defence Council, that it abstained from direct participation in the conduct of the hostilities.Footnote 10
Some governments attempted to justify their support for the US by referring to obligations allegedly arising from the North Atlantic TreatyFootnote 11 or secret treaties of assistance.Footnote 12 Such attempts, however, can be easily dismissed. If the US-led attack against Iraq was indeed a violation of Article 2(4) of the UN Charter other treaty obligations cannot justify it (Article 103 of the UN Charter); if that was not the case, supporting US action needed no justification. It is likely that the invocation of such treaties was rather meant to elucidate the political reasons for providing the assistance in question, ie maintaining an important alliance in the face of disagreements elsewhere.
A more subtle approach was adopted by the Canadian government.Footnote 13 While it granted over-flight rights for US planes on their way to the theatre of operations, Prime Minister Jean Chrétien declared that ‘(w)e have always made clear that Canada would require the approval of the Security Council if we were to participate in a military campaign’. He also said that he had told the President of the United States that ‘Canada was not going to support a war on Iraq’ without Security Council authorisation. Furthermore, he pointed out that ‘American authorities were “very clearly aware of Canada's position”.Footnote 14
The statement of the Canadian Prime Minister can be interpreted as an expression of disagreement with the legal position of the United States concerning military action in Iraq, albeit this disagreement was couched in diplomatic language. However, it can also be interpreted as a statement indicating Canadian political preferences. Assuming that it was an expression of disagreement with the legal position of the United States, or at least a consciously ambiguous statement, the question arises whether the Canadian combination of factual support for the US action and carefully voiced disagreement with the legal position of the US has any legal significance. Is the statement of the Canadian Prime Minister relevant ‘practice’ or an expression of ‘opinio juris’ for the development of customary law and/or the interpretation of UN Security Council Resolutions? Or is it vitiated by the contrasting practical assistance offered by the granting of over-flight rights?
III. RESPONSIBILITY FOR COMPLICITY
A. Introductory Remarks on Primary and Secondary Rules on Complicity
Complicity is not a term of art in international lawFootnote 15 although it is frequently employed to describe situations in which one State furnishes aid or assistance to another State's wrongful conduct.Footnote 16 It was once used by the ILC in its deliberations on supportive States,Footnote 17 but was discarded in favour of the more neutral sounding concept of ‘aid or assistance’ which now appears in Article 16 of the ILC Articles on State responsibility.Footnote 18
The responsibility of complicit States can be engaged in two ways: either a complicit State is responsible for the violation of a primary rule to which it is bound itself, or it incurs derivative responsibility for its assistance to the internationally wrongful act of another State. In the case of the support given to US action in Iraq the most important primary rule which may have been violated is expressed in Article 3 lit. f) of the General Assembly's 1974 Definition of Aggression.Footnote 19 This rule provides that the ‘action of a State in allowing its territory, which it has placed at the disposal of another State, to be used by that State for perpetrating an act of aggression against a third State’ itself qualifies as an act of aggression'.Footnote 20 The possible basis for a derivative responsibility is contained in Article 16 of the ILC Articles on State Responsibility.
Is the distinction between a manifestly erroneous and an arguable justification in any way relevant for our case of complicity? At first sight this is counter-intuitive as it would seem to contradict the elementary binary code of the law,Footnote 21 namely that either a given behaviour is illegal, in which case the rules apply which attribute responsibility for complicit behaviour, or that a given behaviour is legal, in which case the same rules for the attribution of complicit behaviour do not apply. Thus, either the US violated the prohibition of the use of force, in which case Article 3 lit. f) of the Definition of Aggression applies, or the United States has not committed an aggression or used force contrary to Article 2 (4) of the UN Charter (because its behaviour was justified by Security Council Resolutions), in which case States which put their territory at the disposal of the US military have also not committed a wrongful act under international law. This simple analysis suggests that the uncertainty about the validity of the alleged justification of the main actor applies equally to its helper(s). We assume that this is indeed the case for the primary rule as it is articulated by Article 3 lit. f) of the Definition of Aggression. But what about the secondary and more general rule on complicity?
B. Article 16 of the ILC Articles on State Responsibility
Article 16 of the ILC Articles on State Responsibility provides that:
A State which aids or assists another State in the commission of an internationally wrongful act by the latter is responsible for doing so if: (a) That State does so with knowledge of the circumstances of the internationally wrongful act; and (b) The act would be internationally wrongful if committed by that State.
This provision raises a number of questions, in particular whether it actually reflects a rule of customary international law (1), whether only such ‘aid or assistance’ is relevant which crosses a certain minimum threshold (2), whether the knowledge element requires a certain intention to assist (3) and, finally, how Article 16 interacts with primary rules on complicity and the rules set forth by the regime of serious breaches of peremptory rules under general international law (4). We will come to the conclusion that the answers to all these questions, if taken together, militate in favour of a narrowly circumscribed norm of responsibility for providing aid and assistance which only applies if it is sufficiently clear that the act of the other State is indeed illegal.
1. Article 16 as customary international law?
The debates in the International Law Commission have left unresolved the question of whether Article 16 actually reflects a rule of customary international law. In its recent Genocide Convention Case, however, the International Court of Justice has declared this to be the case,Footnote 22 the first time that complicity was discussed by the Court. The case concerned the responsibility of Serbia and Montenegro for possible acts of genocide in Bosnia and Herzegovina. The Court did not apply Article 16 directly, since the case did not involve a situation of inter-state complicity but of aid or assistance to the Republika Srpska, a non-State actor. Non-State actors are generally beyond the purview of the ILC Articles on State responsibility. The Court, however, examined whether the principle underlying Article 16 applied to Article III, lit. e) of the Genocide Convention which establishes an obligation to punish complicity in genocide. The Court saw no reason why it could not resort to Article 16 by means of analogy. While this conclusion is plausible, the rather cursory treatment of Article 16 by the Court cannot replace a more thorough assessment of its basis and its interpretation.
Arguing in favour of the customary status of such a rule, both Special Rapporteurs Roberto Ago and James Crawford relied on ambiguous state practice.Footnote 23 This practice consists of protests against the supply of financial and military aid to a belligerent,Footnote 24 the permission of the use of a State's territory by another State to carry out an armed attack against a third State,Footnote 25 the circumvention of sanctions imposed by the UN Security Council,Footnote 26 and calls by the UN General Assembly to UN Member States to refrain from supplying arms and other assistance to countries with questionable human rights records.Footnote 27 Almost all of these examples can be interpreted as relating to primary rules rather than to a distinct secondary rule: For example, the supply of financial and military aid to a belligerent may be a violation of the laws of neutrality. In addition, the use of a State's territory by another State to carry out an armed attack is a violation of the primary obligation of a State not to knowingly allow its territory to be used to the detriment of other States.Footnote 28 It is true that the ILC has also made reference to specific primary rules which prohibit States to provide assistance for the commission of wrongful acts by other States in certain areas,Footnote 29 but this argument only raises the question of whether the norms referred to are exceptions to or whether they are expressive of a general rule. In other words, it is doubtful whether a secondary rule can be derived from a combination of mostly primary obligations. We would, however, not rule out this possibility since it is questionable whether a strict distinction between primary and secondary rules can always be drawn.Footnote 30
Whether the practice quoted by the ILC is sufficient to prove the existence of a secondary rule cannot be answered here. Governmental comments expressed during the ILC drafting process seem to lend more support to the existence of a customaryFootnote 31 rule than not.Footnote 32 While scholarly accounts are divided,Footnote 33 the existence of such a secondary rule seems at least to have been confirmed more recently than not. This is not only true with respect to the above-mentioned judgment of the International Court of Justice in the Genocide Convention Case Footnote 34 and its pleadings.Footnote 35 The Venice Commission of the Council of Europe has referred to Article 16 when determining the international legal obligations of Council of Europe Member States in respect of secret detention facilities and inter-state transport of prisoners.Footnote 36 The presence of French military advisors in Rwanda prior to the genocide in 1994 prompted the Rwandan government to establish an investigatory commission in order to show that France was complicit in the genocide.Footnote 37 The US State Department held an investigation into the possible Israeli use of US-made cluster bombs in the 2006 Lebanon conflictFootnote 38 which may have violated an agreement between the US and Israel prohibiting the unlawful use of such weapons. Article 16 was also considered and applied by the German Federal Constitutional CourtFootnote 39 and the German Federal Administrative Court.Footnote 40 In some cases governments have accused other States of being complicit in violations of international law. The Syrian Minister of Foreign Affairs has sent a note to the United Kingdom in which he accused the latter for violating international law by supporting the construction of the Ilisu dam project at the River Tigris in Turkey.Footnote 41 The Arab League also issued a formal protest to the UK Government in this regard.Footnote 42
This array of recent statements and practice suggests that there does indeed exist a secondary rule prohibiting aid or assistance to internationally wrongful conduct, although its exact scope remains unclear. It remains to be determined how far Article 16 is the proper articulation of this norm.
2. A minimum threshold for prohibited ‘aid or assistance’?
The ILC has not defined what constitutes relevant ‘aid or assistance’. It is theoretically conceivable that ‘aid or assistance’ comprises every act (or omission) which facilitates the commission of an internationally wrongful act by another State.Footnote 43 However, it is implausible that Article 16 should cover ‘aid or assistance’ which is only remotely or ‘indirectly’ related to an internationally wrongful act.Footnote 44 There should rather be some special nexus between the aid and the wrongful act. This can be demonstrated by the case of the export credit guarantees concerning the construction of the Ilisu dam on the river Tigris.Footnote 45 After the UK had dropped its plans to participate, Austria, Germany and Switzerland were asked to provide export credit guarantees to an international consortium that was charged with the project by the Turkish government.Footnote 46 The three governments initially decided to grant the export credit guarantee subject to a series of conditions obliging Turkey to abide by international legal and soft law standards.Footnote 47 One of the conditions set out in the decisions of the respective governments was that the neighbouring States will be adequately consulted throughout the construction of the dam, an issue which had previously been raised by Syria and the Arab League vis-à-vis the United Kingdom.Footnote 48 This conditionality, however, neither ensures that the three States do not violate international law by granting the guarantee nor that the project is now actually in conformity with international law.Footnote 49
We wonder whether the export credit guarantee is sufficiently linked to any (possible) violation of international law by Turkey so that the guarantor States in question would incur responsibility for assistance.Footnote 50 It is clear that Turkey ultimately benefits from the guarantee, even if the guarantee was formally given to private investors. It is not at all clear, however, whether Turkey actually commits an internationally wrongful act with respect to Syria and Iraq by constructing the dam. While Turkey must abide by the customary principles of fair and equitable use of shared natural resources and of good neighbourly behaviour the precise content of these principles is not easy to determine. Whereas the UN Convention on the Non-Navigational Use of WaterwaysFootnote 51 is fairly detailed as far as procedural obligations are concerned it does not seem to go beyond the Lac Lanoux Arbitration which puts an emphasis on the concept of ‘good faith’.Footnote 52 Turkey's compliance with duties of consultation is equally uncertain. In the Ilisu case, for example, it is also unclear to what extent an existing joint commission of the three States on the common use of the Tigris has dealt with the matterFootnote 53 and whether this would suffice for Turkey to have complied with its obligations to consult under international law.
Note that it is not only a factual uncertainty which characterizes the Ilisu dam case. It is also the unclear state of the law which complicates the assessment of the situation by supporting States. In any event, the case makes clear that it is not always possible to draw a sharp line between factual and legal uncertainty. In many cases, it will be the legal uncertainty which will enhance the factual uncertainty and vice versa. The Ilisu dam case points to a more general conclusion. It suggests that the risk of international responsibility should not be equal for, on the one hand, a State which aids another State to violate a clear international rule and, on the other hand, for a State which aids another State in a legally ambiguous situation.Footnote 54 In the second case, it makes more sense to make only the acting State bear the risk of international responsibility. Vaughan Lowe has maintained that States should normally be entitled to presume that other States will act lawfullyFootnote 55 and that to attribute the same risk of unlawfulness to the main actor and to the helper would mean to treat the assisting State as if it were acting jointly with the main actor.Footnote 56 While such a strict rule on the responsibility for ‘aid and assistance’ at first sight appears to be beneficial for the international rule of law—as it would claim to force States to be their ‘brother's keeper’ and to steer far away from the risk of being implied in illegal activity—it would at the same time discourage many typical and usually beneficial forms of international co-operation. We assume that, on balance, most States prefer an interpretation of the rule on the responsibility for ‘aid and assistance’ which leaves room for such typical forms of cooperation as long as the implication in illegal activity is not sufficiently clear.
International law provides a confirmation for this conclusion in a somewhat related context: It does not hold a State responsible for ‘instigating’ another State to commit an internationally wrongful act.Footnote 57 In this situation it is not normative imprecision, but rather the factual uncertainty whether the instigated State will eventually act upon the instigation (as well as the normative assumption that States typically carefully weigh their options and are not easily ‘instigated’ by other States), which justifies the attribution of responsibility solely to the acting State.Footnote 58 However, the common element is the assessment that, in a situation of uncertainty (whether factual or normative), it is the main actors who should bear the risk of responsibility for their decisions.
In fact, this interpretation of Article 16 is indirectly corroborated by the initial thoughts of Roberto Ago on complicity in the International Law Commission. When Ago presented his Seventh Report on State responsibility in 1978, the situations falling under the envisaged provision on complicity were described by him as follows:
〈There〉 … are the cases in which the existence of an internationally wrongful act unquestionably committed by a State, attributable to it as such and without the slightest doubt involving its international responsibility, is accompanied by the existence of participation by another State (…).Footnote 59
3. The knowledge element
Another important feature of Article 16 is the requirement that the assisting State have ‘knowledge of the circumstances of the internationally wrongful act’. This means, according to the ILC Commentary, that ‘the aid or assistance must be given with a view to facilitate the commission of the act and must actually do so’.Footnote 60 The Commentary thus narrows the textual meaning of Article 16 and goes against the general thrust of the ILC Articles which, in general, presuppose no distinct or separate requirement of fault or wrongful intent for an internationally wrongful act.Footnote 61 However, a requirement of fault or wrongful intent may be found in specific articles of the Draft or may be derived from primary norms which impose different standards of liability.Footnote 62 It thus appears that the ILC wanted Article 16 to be interpreted narrowly so that the ‘knowledge’ element turns into a requirement of wrongful intent.Footnote 63 Likewise, the International Court of Justice has, mutatis mutandis, adopted this approach in its recent Genocide Convention Case.Footnote 64 Here, the Court was faced with the special intent requirement from the Genocide Convention, namely the dolus specialis which is directed towards the elimination of a group under Article II of the Genocide Convention. The complicity rule in Article III, lit. e of the Genocide Convention having more of a criminal law origin, the Court used Article 16 by way of analogy.Footnote 65 Inquiring whether Serbia was responsible for complicity in genocide, the Court pointed out that
‘[T]here is no doubt that the conduct of an organ or a person furnishing aid or assistance to a perpetrator of the crime of genocide cannot be treated as complicity in genocide unless at the least that organ or person acted knowingly, that is to say, in particular, was aware of the specific intent (dolus specialis) of the principal perpetrator.’Footnote 66
If the analogy to Article 16 is supposed to be meaningful, this entails that knowledge of the circumstances of the wrongful act is required at the least (as the Court has put it) also with respect to Article 16. The words ‘at the least’ suggest that, as a general rule, more than mere knowledge is required.
Several governments have wondered about the meaning of the knowledge element of Article 16.Footnote 67 Ultimately, the same reasons which point in favour of a requirement of certainty for what should be considered relevant ‘aid or assistance’ also speak in favour of a narrow interpretation of the knowledge element of Article 16 in the sense of ‘wrongful intent’.Footnote 68 A State that has regularly given development aid to another State which now uses some of this aid to suppress a minority may have ‘knowledge of the circumstances of the internationally wrongful act’ but it certainly does not give ‘the aid or assistance … with a view to facilitate the commission of the act’. Here again, the interpretation should be led by the goal not to discourage many typical forms of international cooperation which, on the whole, have more beneficial than adverse effects.
Nevertheless, a caveat is appropriate here. The requirement of wrongful intent should not allow States to deny their responsibility for complicity in situations where internationally wrongful acts are manifestly being committed. When State A regularly exports military material to State B and it is obvious that State B is systematically violating human rights when repressing its ethnic minorities with the help of this material, State A should not be allowed to hide behind the position that it did not wish to support the commission of such wrongful acts. One can perhaps think of a due diligence obligationFootnote 69 of supporting States to assure that their support is not used for wrongful ends.Footnote 70 Such an approach is supported by State practice and could be based on an analogy with obligations of neutral States to control the export of weapons to belligerents.Footnote 71 A number of States have passed legislation on the international sale of weapons according to which these weapons may subsequently only be used for purposes in accordance with international law.Footnote 72 Such provisions raise the question whether they are sufficient to exclude responsibility for complicity in all cases.Footnote 73 In some cases therefore, a lack of intent can be offset by sufficient knowledge.
4. The systematic relationship between Article 16, the serious breaches regime and primary rules on complicity
While much speaks in favour of the existence of a rule of customary international law that prohibits the provision of ‘aid or assistance’ to the commission of an internationally wrongful act by another State, such a rule should be interpreted narrowly in order not to discourage the typically beneficial forms of inter-State cooperation.Footnote 74 Any provision on complicity and its interpretation should take into account the stability and the smooth running of the international system as a whole. Thomas Franck has highlighted that determining and defining a legal rule is crucial for its legitimacy and that indeterminate standards are likely to facilitate non-compliance with international law.Footnote 75 Hence, it should be established that the assisting State had not only knowledge of the circumstances of the internationally wrongful act but that it also provided its assistance ‘with a view to’ facilitate the commission of this act and that the act to which assistance is given consists of a violation of a sufficiently precise and clearly established rule within international law.
This conclusion is confirmed by the rules of the ILC Articles on ‘serious breaches of obligations under peremptory norms of general international law’ (Articles 40 and 41). Article 41 (2) provides for a separate duty of non-assistance but, in contrast to Article 16, it does not require intent or knowledge. Thus, as far as serious violations of jus cogens are concerned there is a stronger rule against complicity.Footnote 76 It is striking that the ILC has only lowered the standards of attribution in the limited field of serious breaches of peremptory rules. As the ILC has pointed out in its commentary, the lack of a subjective element in Article 41 (2) is motivated by the fact that ‘it is hardly conceivable that a State would not have notice of the commission of a serious breach by another State.’Footnote 77A contrario, this suggests that uncertainty about the commission of the main act is a relevant factor for the application of Article 16.
Another argument for a restrictive interpretation of Article 16 is the existing array of primary rules on complicit States which may impose stricter standards with respect to the required knowledge.Footnote 78 One example of a primary rule clearly requiring more than mere due diligence is the obligation of non-refoulement in Article 3 of the Convention against Torture.Footnote 79 Here, the threshold to engage the responsibility of complicit States is considerably lower since a serious risk that torture occurs is considered to be sufficient in order to trigger the applicability of the rule.Footnote 80 This provision is also an example of how jus cogens rules may impose themselves on the question of how much diligence is due by a State which engages in international cooperation.Footnote 81
The relationship between Article 16, the serious breaches regime and the primary rules on complicity can thus be conceived in the following way: Article 16 provides for the general rule on complicity. As such, it is characterized by the most restrictive interpretation among the three. The other two regimes for responsibility of complicit States go further: The serious breaches regime does so not only because of the importance which the international community attaches to the values protected therebyFootnote 82 but in particular because of the assumption that serious violations of peremptory norms are clearly identifiable.Footnote 83 Some primary rules go further because States have for different reasons committed themselves to a stricter regime. It is thus a systematic and functional relationship between three different regimes which suggests a restrictive interpretation of Article 16. This systematic view also suggests that the case of reasonable uncertainty about whether sufficient Security Council authorization exists for a particular use of force triggers the general and residual regime of Article 16. After all, the assumption underlying the serious breaches regime that ‘it is hardly conceivable that a State would not have notice of the commission of a serious breach by another State’Footnote 84 typically does not apply in such cases. Similar considerations may apply in situations in which only one of two peremptory norms can arguably be satisfied: In the face of genocide, for example, a State may choose to side with those who are using force to prevent and stop the genocide.Footnote 85 Should the assisting State refrain from doing so in order not to incur the risk of being held complicit in a violation of Article 2, para 4 of the Charter?
5. Conclusion
Applied to the US-led attack against Iraq and to States like Canada and Germany, the secondary rule on ‘aid or assistance’ requires more preconditions to be fulfilled than the pertinent primary rules. We have assumed that the primary rule as it is articulated by Article 3 lit. f) of the Definition of Aggres-sion implies that the uncertainty about the validity of the alleged justification of the main actor applies equally to its helper(s). This would be true for the granting of over-flight rights. However, as far as supportive acts which fall outside the scope of this rule are concerned (guarding of US facilities; escorting of US ships on the high seas) States like Canada and Germany could argue that the possible violation of international law by the US was not sufficiently established to justify a secondary, derivative responsibility. This argument could not rest on an alleged lack of precision of the prohibition of the use of force (eg in the light of the challenge by the ‘Bush Doctrine’ of pre-emptive self-defence), but rather on the arguability of the claim, as it has been put forward by the United Kingdom and the United States, that the Security Council had given sufficient authorization to use force against Iraq.Footnote 86 This line of reasoning would not require States like Canada and Germany to adopt the argument that there was indeed sufficient Security Council authorization for the US-led attack, but only to assert that the legal situation was sufficiently unclear to justify that the main actors, in this case the United States and its fighting coalition allies, incur sole responsibility (regardless of certain acts of support).Footnote 87 Another line of argument for States like Canada and Germany could be that they did not act ‘with a view to’ facilitate the attack, but rather as part of a continuous service within a long-standing alliance relationship which they did not want to put into jeopardy. However, a sensible counter-argument would be that the NATO alliance relationship does not require member States to do more than what the individual member State considers necessary (Article 5 of the NATO Treaty) and that the supportive acts were given specifically on the occasion of the Iraq invasion.Footnote 88
IV. THE EFFECT OF COMPLICITY ON INTERNATIONAL LAW-MAKING AND THE DETERMINATION OF RESPONSIBILITY
Regardless of the exact scope of the prohibition of complicity in international law, ‘aid or assistance’ may count as practice in the process of the formation of new customary rules, or as subsequent (interpretative) practice of the parties to a treaty in the sense of Article 31 para 3 lit. b) of the Vienna Convention on the Law of Treaties.Footnote 89 Does this mean that a State, by contributing aid or assistance to a given conduct, has conclusively expressed its position in the process of international law-making, or with respect to the legal assessment of a certain situation? Is complicity equivalent to an endorsement of the legal claims and policies put forth by the aided State?
This question must certainly be answered in the negative when the assisting State plausibly explains its conduct on a different or narrower rationale than the main acting State. If, for example, the United States had justified their attack against Iraq solely with reference to the concept of pre-emptive self-defence as embodied in the National Security Strategy of 2002Footnote 90 and the United Kingdom had rendered the assistance on the clear understanding that it (only) considered the US attack to be justified on the basis and within the limits of certain UN Security Council resolutions, the assistance by the UK would surely not have counted as State practice supporting the Bush administration's strategy of pre-emptive self-defence as a new development of international law.Footnote 91
But what if the assisting State declares, as Canada appears to have done in the Iraq case, that it considers the conduct of the acting State illegal, and nevertheless simultaneously renders assistance to it? Should such a declaration be considered to be without effect (because it is contradicted by the State's own action)? Or can the declaration somehow neutralize the contrary conduct? Or can the declaration even take precedence over the assisting conduct for the purpose of law-making or law-assessment? In other words, does international law give preference to actions of a State over its words? Or does it only take unequivocal contributions into account for its development? Is it conceivable that States are allowed to play contradictory roles depending on different functions with respect to international law?
A. The Role of Protest in International Law
The relationship between contradictory statements and practice has mostly been discussed with respect to the practice of one State and the conflicting statement of another.Footnote 92 The classical debate between writers claiming a preponderance of practice and authors claiming that statements are at least equally relevant need not be repeated here.Footnote 93 It is sufficient to look for those aspects in the classical discussion which are relevant to our study. What we face here is the problem of an apparent conflict between the practice and the statements of one and the same State. The State issues a mixed message. Its statement does not coincide with its actions.Footnote 94 Does there exist a general duty not to behave ambiguously? Such a rule would exclude the consideration of statements which are contradicted by conflicting practice. Before we turn to the question of how mixed messages must be interpreted in international law, a preliminary issue needs to be addressed: Is there a rule in international law which disqualifies ‘paper protests’?
Protests play a role in different areas of international law. For our purposes we should distinguish between three groups of cases. The first concerns diplomatic protests against alleged violations of international law, frequently in the context of diplomatic protection. Here, the protesting State only wishes to express its disagreement with a certain conduct. The second type concerns protests which aim at vindicating rights—such protests are frequently applied in the context of disputes over title to territory. The third type of case concerns protests by which a State wishes to influence the development of customary rules or the interpretation of a given rule of international law.
The first group can be left out of consideration here. With respect to the second group, mere paper protests do not seem to be counted as effective protests. In other words, in order to uphold a title over territory, the claim needs to be effective.Footnote 95 It is not sufficient to merely allege jurisdiction over a given territory.Footnote 96 However, what about the third group concerning the development of customary international law?
Anthony D'Amato has tried to play down the importance of a mere protest that is not backed up by physical acts. In his view, such a protest alone would not suffice to hinder the coming into existence of new customary rules which were, in turn, supported by physical acts.Footnote 97 A similar position seems to have been upheld by Judge Read in his Dissenting Opinion to the Anglo-Norwegian Fisheries Case:
Customary international law is the generalization of the practice of States. This cannot be established by citing cases where coastal States have made extensive claims, but have not maintained their claims by the actual assertion of sovereignty over trespassing ships. (…) The only convincing evidence of State practice is to be found in seizures, where the coastal State asserts its sovereignty over the waters in question by arresting a foreign ship and by maintaining its position in the course of diplomatic negotiation and international arbitration.Footnote 98
However, if compared with the other two groups of protests, the Fisheries Case lies at the faultline between groups two and three. The extension of a fisheries zone and the breadth of the territorial sea are similar to situations in which title to territory is at stake. Indeed, it has been noted that paper protests are only generally ruled out for situations in which specific claims are at stake.Footnote 99 It is indeed difficult to see how customary international law should develop in situations which do not involve specific claims against other States. This is corroborated by the approach of the Permanent International Court of Justice in the Lotus case. When the Court considered whether a rule existed prohibiting the exercise of jurisdiction over collisions at the high sea it pointed out that:
‘[I]t does not appear that States concerned have objected to criminal proceedings in respect of collision cases before the courts of a country other than that the flag of which was flown, or that they have made protests …’Footnote 100
Hence, the Court did not suggest that protests had to be backed up by physical acts. Any other approach would have been difficult to conceive. What should other States have done in order to support their protests? Should they have intervened by force in other States which resort to the type of criminal proceedings at stake? Michael Akehurst pointed to the consequences of such an approach; in his view, neglect for simple protests without physical acts had the potential to turn a ‘polite difference of opinion … into a major international dispute’ which would be ‘disastrous … for world public order.’Footnote 101 Especially if all States of the international community should have the chance of participating in the development of international law, it is ultimately necessary that ‘mere’ protests count in some way.Footnote 102
For these reasons, we conclude that protests which are not accompanied by physical acts do count in the process of international law-making.
B. The Interpretation of Mixed Messages
The preceding considerations do not resolve the question of how to interpret a mixed message, consisting of a complicit act and a simultaneously issued protest. Is it conceivable that a State is precluded from protesting against a legal development due to its factual implication therein? The rules on customary law-making and the interpretation of treaty rules do not provide us with clear-cut answers. Yet, the general principles of estoppel and acquiescence come to mind.
1. Acquiescence
Acquiescence means tacit consent.Footnote 103 In the Gulf of Maine case, the Chamber of the International Court of Justice defined acquiescence as being ‘equivalent to tacit recognition manifested by unilateral conduct which the other party may interpret as consent.’Footnote 104 A failure of a State to assert its right when that right is openly challenged means that the right is abandoned.Footnote 105 In our context, the core question is whether other States may interpret the complicit acts as consent although a protest was issued simultaneously. Considered alone, the support could very well be seen as consent but a simultaneously issued objection speaks against the assumption of ‘tacit’ consent. To the contrary, the objection would turn what could be seen as tacit consent into ambiguous support accompanied by objection. In the simple, yet striking words of Michael Akehurst: ‘Protest is the opposite of acquiescence’.Footnote 106 A situation of acquiescence is therefore not present in our situation, because the objection is aimed at eliminating the interpretation of the complicit act as legal endorsement. And since the principle of acquiescence is rooted in the idea of good faith,Footnote 107 what counts is whether legitimate expectations have been created on the side of other States. The simultaneously issued objection eliminates all possible reliance on a continued and coherent supportive behaviour. Under these circumstances States will not know for sure which position the State in question will adopt in the future, but at the same time they have no reason to rely on only one of the two available options. In other words, no legitimate expectations for a given legal development are created.
2. Estoppel
Just as the principle of acquiescence, estoppel is also rooted in the idea of good faith.Footnote 108 For some, the absence of acquiescence already excludes the possibility of estoppel. This position is, however, not shared by all authorities.Footnote 109Estoppel refers to a situation in which a State has been induced to undertake legally relevant action or abstain from it by relying on clear and unambiguous representations made by another State to the detriment of the former State.Footnote 110 The International Court of Justice has most clearly set out the requirements for a situation of estoppel in its judgment in the North Sea Continental Shelf Cases:
[O]nly the situation of estoppel could lend substance to this contention—that is to say if the Federal Republic were now precluded from denying applicability of the convention régime, by reason of past conduct, declarations, etc., which were not only clearly and consistently evinced acceptance of that régime, but also had caused Denmark or the Netherlands, in reliance on such conduct, detrimentally to change position or suffer some prejudice.Footnote 111
This statement speaks clearly against the position that estoppel could put some limits on the sending of mixed messages. The Court speaks of ‘clearly and consistently evinced acceptance’. A simultaneously issued objection hinders the emergence of a clear and consistent position.
3. ‘Protestatio facto contraria non valet’
Another rule which could militate against the consideration of a simultaneously issued protest might flow from the maxim protestatio facto contraria non valet which itself is an emanation of the general rule of venire contra factum proprium non valet. The maxim goes back to Roman and canon law and served to decide whether protest could be considered as valid. Its general thrust is that protest which contradicts the actions of the protesting person does not count since the factual behaviour is considered to be of more weight than the protest. Is this maxim a general principle of law in the sense of Article 38, para. 1 lit. c) of the Statute of the International Court of Justice? While the principle has been discussed in several—mostly civil lawFootnote 112—jurisdictions,Footnote 113 its exact meaning is unclear. At the international level, the 2004 Unidroit principles on international commercial contracts include a rule on inconsistent behaviour which closely resembles the international legal principle of estoppel.Footnote 114 It thus seems to be the case that the substance of the maxim is contained in other principles such as estoppel or good faith.
However, even if we would assume the contrary, the impact of the principle contained in the maxim would be limited. According to FrenchFootnote 115 and German scholarshipFootnote 116 the problem with the protestatio facto contraria-rule is that it is essentially circular. It only holds that protests are not to be taken into account to the extent that the protesting person does not have the legal power to decide over the effectiveness of the protest.Footnote 117 In French writings, a strong emphasis is put on the fact that sometimes protests which accompany actions are allowed to express a non-vouloir of one party.Footnote 118 As the maxim of protestatio facto contraria non valet in itself is ambiguous, French civil law has developed other rules on the requirement of consistent behaviour or on the obligation not to contradict oneself to the detriment of someone else.Footnote 119 With regard to these rules the same can be said with respect to the Unidroit principles: they embody more or less what is understood by acquiescence and estoppel in international law.
4. Good faith
Finally, as far as the principle of good faith is concerned it is questionable whether mixed messages can create legitimate expectations on the part of other States on which they can rely. The answer can only be negative. To use ‘good faith’ as a tool against the sending of mixed messages would mean that State practice would exclusively determine the formation of new customary international law.
5. Conclusion
If the traditional two-elements theory of customary international law is to be upheld, not all instances of practice can count in the same way for the development of customary international law. Otherwise the element of opinio iuris sive necessitatis would become meaningless in many cases. While the exact scope and importance of the psychological element of custom has from time to time been called into question,Footnote 120 its relevance is confirmed by Article 38, lit. b of the Statute of the International Court of Justice, the jurisprudence of the ICJFootnote 121 and by the vast majority of scholarly writings.Footnote 122 If we accept that a State must have an opportunity to declare that an aspect of its practice is not ‘evidence of a belief that this practice is rendered obligatory by the existence of a rule requiring it’,Footnote 123 we see no reason why this should not apply to the situation we are discussing here. Even those authors who generally question the impact of the opinio iuris of States on the development of customary international law recognize its importance in the context where ‘the circumstances are such as to suggest that the precedents should perhaps not count’.Footnote 124
Sir Hersch Lauterpacht has maintained that ‘the far-reaching effect of the failure to protest is not a mere artificiality of the law. It is an essential requirement of stability …; it is a precept of fair dealing inasmuch as it prevents states from playing fast and loose with situations affecting others.’Footnote 125 At first sight, this statement seems to postulate a duty of consistency and to speak against the sending of mixed messages (and their possible consideration afterwards). On closer inspection, however, the statement rather supports our analysis: Lauterpacht did not address the question of what is a protest, but only the effects of non-protest. In fact, Lauterpacht hinted at the possibility that States may distance themselves from their practice as not counting for the development of the law. When he discussed the role of opinio iuris he pleaded for a nuanced approach: As a general rule, the opinio iuris could be inferred from the material practice, however, ‘except when it is shown that the conduct in question was not accompanied by any such intention.’Footnote 126
We can therefore conclude that no rule or principle exists in international law which excludes the sending of mixed messages and their consideration in the law-making process.
C. The Dual Role of States
If there is no rule which excludes mixed messages we can address the question of their legal significance. It is, of course, tempting to postulate a rule according to which a State's practice must prevail over its contrary statements since practice is somehow more weighty.Footnote 127 Such a position overlooks, however, that only such practice with which a State identifies can be counted as relevant. This is clear, for example, in cases in which a State commits certain human rights violations but at the same time claims to be bound by human rights treaties. Similarly, the ICJ has affirmed in the Nicaragua Case that violations of a rule may actually strengthen it if the State in question explains its conduct by reference to accepted exceptions or justifications to the rule.Footnote 128 It is true that the ‘practice’ of the State violating human rights is usually unacknowledged, while Canada did acknowledge its granting of over-flight rights to the United States. It is, however, equally true that if a State acknowledges to have intentionally violated international law in a particular case (without claiming that it had a right to do so), this breach of the law will not necessarily be counted as State practice in support of a claim that the conduct should be counted as permitted.Footnote 129 If, for example, Canada had offered Iraq compensation for its granting of over-flight rights to the United States, its practice of granting over-flight rights would certainly not have counted as supporting an assertion of the United States to have been legally entitled to attack Iraq. Thus, in the case of mixed messages, it is not always actions of a State which prevails over its statements.
This approach can be explained when we distinguish the two different roles which States play in international law. States are both subjected to international law and create and authoritatively interpret it. This dual role is apparent in the process of the formation of new customary law. A well-known example is the Fisheries Jurisdiction Case: when Iceland claimed and enforced a fishery zone beyond its territorial waters it violated the applicable law of the sea.Footnote 130 At the same time and by the same act, however, it put forward a new legal claim which eventually crystallised into a new rule of international law. The two roles States play in international law have been conceptualised by Georges Scelle in his theory of the dédoublement fonctionnel. According to Scelle, States act not only in their own capacity but also as agents of the international community.Footnote 131 As long as the international community of States lacks its own proper organs, State organs fulfill different functions. This is especially true with respect to the process of international lawmaking. In Scelle's own words: ‘… Les gouvernants étatiques se donnent à la fois comme sujets de Droit international et comme seuls créateurs de Droit …’Footnote 132
The ongoing institutionalization of the international legal order certainly reduces the importance of Scelle's theory.Footnote 133 The proliferation of international courts and tribunals and the activities of international organisations take away certain areas of application of the dédoublement fonctionnel of States. Despite this development States are still the primary actors in the formation of customary and treaty law as well as with respect to the authoritative interpretation of international law. International law-making and interpretation in areas without compulsory dispute settlement procedures remain the crucial domains for the dédoublement fonctionnel of States.Footnote 134
The distinction between States' roles as lawmakers and as subjects of international law makes it possible to attribute the different parts of a mixed message to the different roles a State plays and thereby to disentangle and to clarify the mixed message. It is conceivable that the practice of a State is meant to belong to the sphere of implementation of the law, whereas its statement expresses the State's position on how the law should be developed or interpreted.
D. Ambiguity in Grey Areas
States may issue mixed messages. However, the evolution of customary and treaty law would risk going astray if it became too easy for States to dissociate themselves from their actions. Practice needs to remain the basis for customary international law. Statements should express the opinio juris of a State and include protest against unwanted legal developments. Does this call our preceding considerations into question?
We wish to argue for a nuanced approach. In most cases, State practice means what it appears to be, especially if it is unaccompanied by interpretive statements. Sometimes, however, practice does not (fully) express the legal position of the State concerned. States may see themselves under political pressure to break the law, or to embark on legally questionable conduct, or to support an undesired legal development. It is in the interest of the international community as a whole and the development of international law in general if practice which yields to such pressures does not too easily become the norm. The international legal system should resist changes of an overly volatile character. While the adaptability of international law to new challenges is similarly important, a legal system should be able to differentiate in this regard. States which express their reservations concerning novel interpretations of the law can play an important role in this regard. Of course, changes of the law can result from giving in to political pressure and not necessarily from the conviction of their inherent value. This may have been the case with certain changes in the law of self-defence after 11 September 2001.Footnote 135 However, in clearly controversial cases such as the attack against Iraq, the political goal of some States not to make matters worse should not preclude them from expressing their independent legal position, either with respect to the interpretation of the law (whether the attack against Iraq was justified) or with respect to the development of the law in general.Footnote 136 In certain circumstances, mixed messages are helpful for the proper development of international law.
However, mixed messages are not an appropriate tool in situations where the law is clear. This was the case with the core aspects of the so-called rendition programme of the United States. There can be no doubt that the inter-State transfer of persons outside regular legal procedures with a possibility that these persons become victims of torture in a third state violates international law.Footnote 137 Responsibility for complicity in these actions should be clearly affirmed where the other prerequisites are met.
There are, however, situations in which the legal situation is less clear. The US-led attack on Iraq in 2003 is a good example of such a case. While, in our opinion, the better arguments support the position that this attack was neither justified by the right of self-defence nor by an authorization of UN Security Council resolutions, some international lawyers have respectably argued that certain Security Council resolutions provided a justification for the attacks.Footnote 138 In such a grey area, there should be room for assisting States to both incur the risk of being held responsible for a violation of (primary rules of) international law, and, in their role as law creators, to throw their weight on the side of those who express their preference for a different interpretation or development of international law. Thus, in this limited area States may send mixed messages which reflect the different roles they play with respect to international law.
The sending of mixed messages may, however, come at a price. States which repeatedly express themselves in an equivocal manner are likely to lose credibility. If a State builds a reputation for being unreliable or equivocal, its actions will arguably become less influential over time. Therefore, States should in their own interest as far and as often as possible behave consistently. Only in situations which involve factual and legal uncertainties can it become reasonable for a State to issue a mixed message.
CONCLUSION
States which assist another State to commit an internationally wrongful act risk being held responsible, along with the offending State, on the basis of the applicable primary rules. Uncertainty as to whether such rules apply in a given situation fully translates to the assisting State. However, as far as responsibility on the basis of the secondary rule of Article 16 of the ILC Articles on State Responsibility is concerned, substantial uncertainty as to the legality of the main act excludes such responsibility.
Regardless of their international legal responsibility for their acts of assistance, States should formulate their position with respect to the development or authoritative interpretation of international law. In most cases the conduct of a State for the purpose of state responsibility and the conduct for the purpose of the development of the law coincide. However, in a legally ambiguous situation it is possible for a government to provide assistance in order not to risk a political alliance and at the same time to express its position that international law excludes or restricts the conduct of the acting State. Thus, in some situations States may preach the contrary of what they practice and reconcile their habitual forms of cooperation with an independent perspective on the development and interpretation of international law.
We wish to thank Eyal Benvenisti (University of Tel Aviv) and Andreas Paulus (University of Göttingen) for their valuable comments and criticism.
