Victrix causa diis placuit sed victa Catoni.

Lucan, Pharsalia, I.128.


1. Introduction

In her paper published in QIL as basis of this Symposium, concerning the decision of the ICJ on the existence of a dispute in the Marshall Islands case,[1] Professor Bonafé, a modern Cato, embraces with enthusiasm the losing side.[2] She supports the contentions of the Marshall Islands on that point, rejected by the International Court of Justice; and offers additional arguments, not considered by the Court, why their cause should have prevailed, at least to the extent that the case should not have been excluded ab initio. On the issue dealt with by the Court, the present writer leans rather to the opposite view; but a complete dialogue would require more time and space than is here available. Comment will here be offered simply on certain salient issues.

The decision was a finding solely on jurisdiction, by the upholding of one of the Respondent’s preliminary objections. A curiosity of the judgment is that it nowhere indicates on what basis jurisdiction was asserted and upheld (eg, a treaty provision, or an acceptance of jurisdiction by declaration under Article 36(2) of the Court’s Statute); the implied assumption is that whatever the textual basis, the objection would in any event be fatal, which appears to be correct, even though it appeared that in this field there might be legal implications from the nature of the jurisdictional instrument invoked.[3]

Central to the Court’s decision was the principle that it must be able to discern the objective existence of a dispute between the parties at the date on which the proceedings were commenced. The Court held further that for the dispute to have thus existed, the respondent must at that date have been aware that there was a dispute between itself and the applicant; and that in the instant case this was not so, with consequent dismissal of the Application. In the view of Professor Bonafé, this ruling was an innovation, such a requirement not having been applied in previous jurisprudence, and unjustified, or at least insufficiently justified.[4] This is the first question here examined.

Professor Bonafé, citing earlier jurisprudence, also considers that the Court should have held that the procedural defect, being easily remediable, should be treated in effect as having been remedied, and the case permitted to proceed. In this she is, in the opinion of the present writer, probably correct: the Court seems at the very least to have neglected the point.


2. Was the ‘awareness’ requirement in the Marshall Islands case an unprecedented restriction?

To establish the presence of a revirement de jurisprudence, it would be necessary to point to one or more cases in which the respondent did not have, or was not shown to have had, the requisite awareness, but the Court nevertheless found that a dispute existed.

Clearly relevant in this context, and duly discussed by Professor Bonafé, is the decision in the case of Alleged Violation of Sovereign Rights between Nicaragua and Colombia.[5] In that case, the Court discounted Colombia’s ‘surprise’ at the bringing of the case, since it ‘could not have misunderstood the position of Nicaragua over [their] differences’, i.e. it was aware of them. Nevertheless, Professor Bonafé takes the view that the Court ‘did not regard awareness [on the part of the Respondent] as a requirement for the determination of the existence of a dispute’, the basis for this conclusion being that ‘it was not mentioned in the general part of the judgment dedicated to the notion of dispute’.[6] This seems a rather simplistic, if not naïve, way to interpret an ICJ decision. The jurisprudence of the Court is to be derived from its effective application of the law; it often makes general statements of principles and rules in addition to applying them to the specific facts, and such statements are valuable even beyond the extent to which the case exemplifies their practical application. However, if a case demonstrates the working of a rule in a specific context, it is irrelevant for jurisprudential purposes whether any general statement of the law in the decision is widely enough drawn to entail that specific result. In any event, in the ‘general part’ referred to in the Sovereign Rights case, the Court cited the classic definition in South West Africa, quoted in the Georgia/Russia case, that ‘It must be shown that the claim of one party is positively opposed by the other’;[7] it is difficult to see how it is possible to ‘positively oppose’ a claim of the other party of which one is not yet aware.

Professor Bonafé then turns to the 1996 decision in the Bosnia/Serbia Genocide case, which again she does  not accept as a further precedent going to ‘awareness’. It is true that, as the Marshall Islands decision pointed out, in that case ‘the issues the Court focused on were not the date when the dispute arose but the proper subject-matter of that dispute, whether it fell within the scope of the relevant compromissory clause, and whether it “persist[ed]” at the date of the Court’s decision.’ After establishing that the parties to the case were parties to the Genocide Convention, the Court continued: ‘… it remains for the Court to verify whether there is a dispute between the parties’, but the sentence went on: ‘that falls within the scope of that Convention’. It did not say ‘whether there is a dispute between the parties, and if so whether it falls within the scope of that Convention’; but this would have been unnecessary since the first point, the mere existence of a dispute, was itself not disputed. It is clear from the verbatim record that, in this respect, the parties were arguing in the 1996 proceedings solely over whether there existed a dispute falling within the compromissory clause of the Genocide Convention, not whether there existed a dispute at all before the filing of the Application. Accordingly, the case is no authority for the proposition that, where this issue is raised, no demonstration of the existence of the dispute, no ‘awareness’, before the filing of the Application is required.

A similar approach is apparently adopted to the 2011 Judgment in the Georgia/Russia case, though Professor Bonafé concedes that ‘the Court tried to establish whether the applicant’s claims concerning racial discrimination came to the attention of the respondent before the institution of proceedings’. In her view however the fact is discounted on the basis that ‘this enquiry appeared to be strictly related to the fact that Georgia relied, as jurisdictional basis, on the compromissory clause of the [CERD] Convention’.[8] The logic here is obscure, unless the argument is that where a conventional jurisdictional basis is relied on, there must be pre-Application awareness of the dispute, but not, or not necessarily, when the claim is based on breach of convention, but the jurisdictional basis is optional-clause declarations, as is in the Marshall Islands case.[9] Professor Bonafé cites the point made by Judge Greenwood in his Separate Opinion, that where there exists a dispute as to compliance with a convention and at the same time a wider dispute, the Respondent needs to know that ‘a claim is being made against it regarding the interpretation or application of the Convention’[10]; but it is not clear how this supports her argument, since Judge Greenwood clearly had in mind awareness before the filing of the Application (as the Application itself would hardly fail to make the point clear). The same point made above applies here: if a State is entitled to know, before proceedings are instituted, that it is accused of breach of a specific convention, should it not also be entitled to know, also before proceedings instituted, that its conduct is, in some other way, called in question, ie that there is a matter in dispute between itself and the future Applicant?

Is there guidance to be obtained from consideration of the purpose for which the requirement under discussion exists? In Professor Bonafé’s view, ‘[t]he purpose of the existence of the dispute requirement is not to protect the parties from ‘surprise’ applications’, but ‘to safeguard the judicial function of the Court’, as ‘[t]he existence of the dispute makes the contentious function of the Court effective, as clarified in the Northern Cameroons case’.[11] In support of this view, Professor Bonafé invokes the rule of forum prorogatum. Since the consent of the respondent enables the Court to exercise its judicial function even if that State is unaware of the existence of the dispute before the filing of the Application, ‘what counts for the exercise of the Court’s jurisdiction is the objective determination’, that is, by the Court, ‘of the existence of the dispute. … The subjective perception that the parties may have before the application does not matter’.[12] The last sentence is correct; but the reason, ‘what counts’, is not any determination by the Court, but the consent of the respondent. The Court is only able to proceed because the respondent has waived the right to object on jurisdictional grounds: if one of those grounds was the absence of a pre-Application dispute, the waiver of objection signifies that it is now irrelevant whether, objectively, there was or was not such a dispute. There is no reason why the point should be examined and determined, ‘objectively’, by the Court at all.

The discussion turns on a false dichotomy. We may agree that the purpose of the ‘existence of the dispute’ rule is to protect the Court’s judicial function, but nevertheless consider that, for there to exist a dispute, the Respondent must be aware of the opposition of views; and this is not so that it may be protected against surprise Applications, but so that a dispute may be found to exist. A dispute between two parties, one of whom has no idea that there exists any such dispute, is a self-contradictory concept.[13]

Professor Bonafé notes that the Court stated that ‘If the Court had jurisdiction with regard to disputes resulting from exchanges in the proceedings before it, a respondent would be deprived of the opportunity to react to the claim made against its own conduct’,[14] and apparently reads (and refutes) this as the sole justification for the ‘awareness’ requirement.  But the Court immediately continued: by stating the ‘awareness’ rule as an additional point: ‘Furthermore, the rule that the dispute must in principle exist prior to the filing of the application would be subverted’.[15] Thus the possibility of reaction by the Respondent is not advanced as the justification for that rule.

The existence of a dispute may certainly be deduced from the conduct of the parties. To this end Professor Bonafé argues that, in the Marshall Islands case, ‘the Court merely turned a blind eye [to] the fact that the UK declined to co-operate with certain diplomatic initiatives, failed to initiate any disarmament negotiations, and replaced and modernized its nuclear weapons’, and the Court thus chose ‘to ignore the fact that the claim was positively opposed by the respondent’s conduct’.[16] But this means that the existence of a dispute was visible to the Marshall Islands; but what has to be established is that the United Kingdom was aware that the Marshall Islands regarded this conduct as a breach of their rights. In passing, we may also note that the obligation a breach of which the Marshall Islands were asserting was not an obligation to abandon a State’s own nuclear weapons, but an obligation to pursue disarmament negotiations.[17] Negotiations are, at least, two-sided: they cannot even be ‘initiated’ without some degree of co-operation from a partner or partners.[18] Accordingly, it would not have been evident to the UK that its conduct would necessarily be regarded by other parties to the Treaty as a breach.[19]

Linked with the question of the significance of conduct, there is also a suggestion that the ‘awareness’ requirement may involve inconsistency with the principle of equality of the parties, or of States. We are told that ‘powerful states can be in a position to invoke the responsibility of the future respondent at such an early stage’, but ‘weaker states could not be able to do so’; and the example offered is that of President Trump’s exchange with North Korea, threatening ‘fire and fury’, which would satisfy the Court’s requirement of ‘awareness’ if one of the two States brought proceedings against the other. But while South Korea, shall we say, could not credibly make the same threat, it is not the threat that matters: it is the attention drawn to the conflict of views, which South Korea could perfectly communicate without Trumpian braggadocio; thus the principle of equality is surely preserved.

The significance of the ‘awareness’ requirement should not be over-stated; but nor should it be trivialised. It is, first, an over-statement to suggest that that requirement ‘reflect[s] a conception of judicial settlement that is secondary to diplomatic settlement’ and ‘means that the ordinary means of dispute settlement is diplomatic settlement’.[20] The Court has not suggested that diplomatic negotiation is essential; merely that the view of the one party is opposed by the other, whatever the form in which that opposition may be discerned. The Court did not say that a complainant State must first – before filing an application to the Court ˗ try to settle the issue by diplomatic means: all that it said was that that State must, before litigating, make sure that the existence of its grievance, the possible future basis of such litigation, is known to the potential Respondent.

On the other hand, Professor Bonafé seems to underestimate the extent to which the filing of an application transforms the situation, if only because of the fact that proceedings before the Court, and thus the nature of the dispute (which may previously have been unrevealed), are known to the public from then onward.[21] There may be no right to prevent such revelation, but it cannot be said that the situation of the Respondent is unaffected by the commencement of proceedings.[22] That said, Professor Bonafé is clearly right to emphasize the distinction between ‘the awareness of the existence of the dispute and the awareness of the intention of the applicant to bring the case before the Court’.[23]

More generally, it is not apparent in what way the Court ‘rendered more difficult the establishment of a dispute’;[24] it merely took the logical view that a dispute has to be (at least) two-sided, thus each party must be aware that the other party disagrees with it.


3. Should the Court have rejected the UK objection on the basis that the defect could easily be remedied, as in the Croatia v Serbia case?

Professor Bonafé draws attention[25] to the 2008 decision in the Genocide Convention case, where the Court recognized the existence of a valid objection to jurisdiction, but held that it was ‘not in the interests of justice to compel the applicant to begin the proceedings anew – or to initiate fresh proceedings’, in which ‘the unmet condition would be fulfilled’.[26] In Professor Bonafé’s submission, the Court should have adopted a similar course in the Marshall Islands case. This was not suggested by the applicants in that case, since for them the question did not arise until the decision had been given;[27] nor was it suggested by any of the dissenting judges.[28] It is in fact striking that no judge seems to have raised the issue, so that the judgment found it unnecessary to mention it. In one sense the unmet condition would seem to be one ‘easily fulfilled’: indeed the existing proceedings would have created the necessary ‘awareness’ in the respondent.

The silence of the Court on the point makes it difficult to assess the thinking that underlay the decision.  It may be that a relevant issue was the nature of the defect in the proceedings in the cases being compared. In both the Genocide case and the Marshall Islands case it was jurisdictional, and the passage cited from the earlier case was responding to an argument that ‘an applicant’s deficiency’ in that respect ‘might be overcome in the course of proceedings, while that of a respondent may not’; thus by that date the issue had been, as it were, half-decided. In the Marshall Islands case, the problem was not merely one of jurisdictional status, or of capacity: if there was no dispute, then there was no case, within the framework of which the Court could go on to consider whether it was or was not ‘in the interests of the sound administration of justice to compel the applicant to begin the proceedings anew’.

In the one case, the Court was saying: ‘By the Application we were seised of an existing dispute, which was a valid act of seisin, but the Respondent had not agreed to our jurisdiction to decide it; it has subsequently given that agreement, so we can go ahead without the formality of the case being re-submitted.’ In the other, it would have to say: ‘The Application did not seise us of an existing dispute, so there was no valid act of seisin, so we had – and still have – no dispute before us.’ On that basis the case would have to be re-started. It is true that in the 2008 decision the Court considered that, although the defect in that case meant that it had not been ‘properly seised’ (in the words of an earlier decision), this did not mean that it ‘lack[ed] the competence to decide on its own jurisdiction’. But an endeavour to seise the Court of what is, at that moment, a non-existent dispute is surely more than a defect in seisin, but an ineffective act, a nullity.

Thus might the Court have argued in order to continue to reject the Marshall Islands’ application; but this does not excuse its apparent failure to consider the point, even if it did not apparently have the assistance of the parties and their counsel in recognizing its existence.



* Formerly Principal Legal Secretary, International Court of Justice and Professor of International Law, Graduate Institute of International Studies, Geneva.

[1] Obligations concerning Negations relating to the Cessation of the Nuclear Arms Race and to Nuclear Disarmament (Marshall Islands v United Kingdom, Marshall Islands v India, Marshall Islands v Pakistan) (Preliminary Objections) ICJ Judgment 5 October 2016, nyr (hereafter Marshall Islands v United Kingdom)

[2] Cato was not in this instance isolated: the judgment was only adopted by the President’s casting vote, the division of votes being 8 to 8; and it has earned the unanimous disapproval of an AJIL symposium (cited by BI Bonafé, ‘Establishing the Existence of a Dispute before the International Court of Justice: Drawbacks and Implications’ (2017) 45 QIL-Questions Intl L 3, fn 11). Some other contributions to this symposium have already been published, and these may well also agree with Professor Bonafé’s approach, but I have preferred to postpone reading them until I have completed and submitted my own view on the matter.

[3] See Marshall Islands v United Kingdom (n 1) para 38.

[4] ‘The Court does not depart from its settled jurisprudence unless it finds very particular reasons to do so’: Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Croatia v Serbia) (Preliminary Objections) Judgment [2008] ICJ Rep 435, para 71.

[5] Alleged Violation of Sovereign Rights and Maritime Spaces in the Caribbean Sea (Nicaragua v Colombia) (Preliminary Objections) Judgment [2016] ICJ Rep 3

[6] Bonafé (n 2) para 16.

[7] Marshall Islands v United Kingdom (n 1) para 50, citing [1962] ICJ Rep 328.

[8] Bonafé (n 2) para 17.

[9] Worthy of no more than a footnote is the following suggestion: ‘In the French version of the judgment references to the ‘connaissance’ of the respondent are to be found in paras 61, 87 and 104, whereas in the English version the only reference to the ‘awareness’ of the respondent is in para 87. This might confirm that for the Court in 2011 such awareness was not so essential in the determination of the existence of a dispute between the parties.’ (Bonafé (n 2) fn 28). The corresponding English expressions are ‘was known to’ (Marshall Islands v United Kingdom (n 1) para 61), ‘were plainly aware’ (ibid para 87) and ‘came to the attention of’ (ibid para 104).  To anyone acquainted with translation techniques in general, and specifically at the level of an ICJ judgment, the idea that these variations signalled a deliberate weakening of the argument advanced is wholly unconvincing.

[10] [2011] ICJ Rep 276 Separate Opinion Greenwood para 9.

[11] Bonafé (n 2) para 56, referring to [1963] ICJ Rep 38.

[12] Bonafé (n 2) para 57.

[13] An interesting parallel is to be found in the Court’s handling of the concept of ‘crystallization’ of a dispute, as a means of defining the moment at which a dispute, for the purposes of its resolution by the Court, comes into existence. This was in the context of a dispute over territorial sovereignty: before the moment of crystallization any activity by one party in relation to the dispute could be taken into account by the Court in support of that party’s claim, but after crystallization there would be in effect a presumption (though the Court does not use the term) that such acts were ‘undertaken for the purpose of improving the position of the Party that relies on them’ (Sovereignty over Pulau Ligitan and Pulau Sipadan (Indonesia/Malaysia) Judgment [2002] ICJ Rep 682, para 135, cited in Sovereignty over Pedra Blanca/Pulau Batu Puteh, Middle Rocks and South Ledge (Malaysia/Singapore) [2008] ICJ Rep 12, para 32). To seek to improve one’s position in a dispute implies awareness that there is a dispute; or contrariwise, there is no reason to be suspicious of such activity if the party acting was unaware that its legal position was disputed.

[14] Marshall Islands v United Kingdom (n 1) para 43.

[15] ibid, emphasis added.

[16] Bonafé (n 2) para 12.

[17] ‘Each of the Parties to the Treaty undertakes to pursue negotiations in good faith on effective measures relating to cessation of the nuclear arms race at an early date and to nuclear disarmament, and on a treaty on general and complete disarmament under strict and effective international control’: Treaty on the Non-Proliferation of Nuclear Weapons, art VI.

[18] ‘[O]ne cannot negotiate alone’: Marshall Islands v United Kingdom (n 1) Dissenting Opinion of Judge Crawford, para 33.

[19] Possibly the UK could, and should, have been doing more to kick-start such negotiations, to such an extent as to constitute a failure to meet its Treaty obligations; but the modernization of its own nuclear weapons, for example, could not be a breach of the Treaty so long as other States’ possession of them rendered that conduct appropriate as a matter of national security.

[20] Bonafé (n 2) para 59.

[21] A fact of which tactical advantage has been taken in the past, in the so-called ‘phony’ cases, eg the two Aerial Incident cases and the two Antarctica cases in 1956.

[22] As is apparently suggested in Section III.1 (paras 38-46).

[23] Bonafé (n 2) para 60; though the UK declaration, being clearly drafted ex abundanti cautela, may not in fact exemplify such confusion.

[24] ibid para 47.

[25] ibid para 41.

[26] Application of the Genocide Convention (n 4) para 85. This would seem not to be possible if the jurisdictional basis has in the meantime been disturbed, eg by the withdrawal or amendment of a declaration under art 36(2) of the Statute. (The UK did in fact close the stable door in this way, though rather belatedly, as Professor Bonafé notes (para. 61).

[27] A far-sighted counsel might have argued on the lines of ‘Even if you think that the UK was not aware of the existence of the dispute at the date of the Application, it is so aware now, so justice is satisfied’.

[28] Judge Robinson noted the decision in his dissenting opinion (para 54), but did not suggest that the Court should have taken the same course in the then current proceedings.