Jump to Content Jump to Main Navigation
Treaties and Subsequent Practice edited by Nolte, Georg

Part 1 Doctrinal Aspects, 2 Subsequent Agreements and Subsequent Practice within and outside the Vienna Convention

Luigi Crema

From: Treaties and Subsequent Practice

Edited By: Georg Nolte

From: Oxford Public International Law (http://opil.ouplaw.com). (c) Oxford University Press, 2015. All Rights Reserved.date: 21 May 2019

Treaties, interpretation — Vienna Convention on the Law of Treaties — Travaux préparatoires — Good faith

(p. 13) Subsequent Agreements and Subsequent Practice within and outside the Vienna Convention1

1.  The Painting and its Frame

The interpretation of treaties is a topic that has already attracted extensive and meticulous attention from legal scholars.2 However, in such a broad subject, like a wide oil landscape, some details appear sharp in the foreground, while others are more vaguely sketched and call for fresh comment. The title of this article focuses attention on one such detail: the role of subsequent agreements and practice in treaty interpretation; but before turning to this topic, a few words must be dedicated to better delineating the frame of the whole picture.

The scope and subjects of interpretation can vary; we see interpretation as proposed by legal scholars; interpretation as a matter of everyday life of international actors, and above all of states; and interpretation as it takes place in Tribunals. So, following the agenda of the International Law Commission on the topic of Subsequent Agreements and Subsequent Practice in Relation to the Interpretation of Treaties,3 I will focus the spotlight on interpretive reasoning in judicial decisions. Secondly, I will exclusively (p. 14) consider decisions by the International Court of Justice (ICJ) and by ad-hoc arbitral tribunals, in which, on the one hand, the presence of a dispute, the factual circumstances of the case, the special agreement, the petitum and the pleadings of the parties contaminate a pure philological approach to the text, and, on the other hand, judges and arbitrators are not regular participants in the functioning of a special legal regime.

Thirdly, I will comment on recent decisions only, handed down after the approval of the Vienna Convention on the Law of Treaties (VCLT, Vienna Convention).4

Fourthly, amongst these decisions I will devote special attention to what are perceived today as good interpretations, in particular on those characterized by a structured application of the Vienna rules, and on those characterized by a more liberal and broad, but consistent, approach.

2.  A Close-up View: The Vienna Convention

Even if the codification of rules on the interpretation of treaties seemed almost impossible at the time of their discussion,5 arts 31 to 33 of the Vienna Convention have so far obtained widespread, pervasive success: they have been used in (almost) every international jurisdiction;6 they have been declared customary and, thus, used in cases (p. 15) involving states not parties to the Vienna Convention;7 and they have been used to interpret almost every kind of international act,8 including both the antiquated and the contemporary.9 Even in those decisions where they are not applied, a Vienna-like language is often assumed, as in the award of the Eritrea-Ethiopia Boundary Commission.10 It seems that the menace of proliferation of international regimes and jurisdictions, as opposed to killing the articles, has fortified them in their role as unifiers of international legal reasoning. Thus, I want address them, and their analysis, before moving on.

Subsequent agreements and practice are mentioned in para 3 of art 31, which reads:

[For the purpose of the interpretation of a treaty] There shall be taken into account, together with the context:

  1. (a)  any subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions;

  2. (b)  any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation;

  3. (c)  any relevant rules of international law applicable in the relations between the parties.

This paragraph covers certain relevant interpretive factors, more or less determinative,11 such as authentic interpretation (explicit or tacit, of the original intentions, or of those that emerge later), and other concurrent obligations. A literal reading (p. 16) shows a certain amount of overlap between the three provisions of this paragraph; they progress from the mandatory interpretations at subpara (a) to the broader, residual, statement at subpara (c).12 While (a) clearly focuses on every subsequent authentic interpretation made by the parties (ie subsequent agreements that are explicitly interpretive of a previous treaty), the interpretation and the effects of subpara (b) are less clear.

3.  Searching for the Right Perspective: Subsequent Practice in the Vienna Convention

The only explicit reference to subsequent practice in the Vienna Convention is made at art 31(3)(b), ‘… any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’.13 This chapter will commence with this point, addressing those decisions that interpreted and applied it.

The ICJ has interpreted this disposition in two ways, one literal and the other much more liberal. According to the former, arts 31 and 32 establish a structured way of interpreting a treaty, and the references to subsequent practice as a primary means of interpretation are all captured under the clear and precise wording of art 31(3)(b). (p. 17) Thus, the search for the meaning of an expression begins with a multi-faceted analysis of the text—grammatical (punctuation, use of conjunctions), logical (eiusdem generis, etc), linguistic (references to the common language—dictionaries—or to the language of art—precedents), systemic (treaty as a whole). It then considers the object and purpose, followed by the context, and then the subsequent practice of the parties. The purpose of the latter is two-fold: to show the original intentions of the Parties, and, generally, to show the emergence of a new understanding of them. Lastly, regard is had to the travaux préparatoires and to other supplementary means, even if only to confirm an already clear meaning.14

In the Kasikili/Sedudu Island case the Court specified under which conditions art 31(3)(b) is applied:

To establish [the practice requested by art 31(3)(b)], at least two criteria would have to be satisfied: first, that the occupation of the Island by the Masubia was linked to a belief on the part of the Caprivi authorities that the boundary laid down by the 1890 Treaty followed the southern channel of the Chobe; and, second, that the Bechuanaland authorities were fully aware of and accepted this as a confirmation of the Treaty boundary.15

This decision gives a very strict and literal—and correct—interpretation of art 31(3)(b), focusing on the three concepts of parties, application of the treaty, and agreement contained therein. The reference to these terms implies that subsequent practice can be considered when the parties to a treaty, through their authorities, engage in common conduct in the application of the treaty, and when such action is conducted wilfully and with awareness (belief, fully aware) of the consequences of their actions.

In short, if the Vienna Rules are applied in a structured way to the interpretation of treaties, subsequent practice only has any relevance when the aforementioned conditions are met. Anything falling outside of those precise conditions can only have a subsidiary interpretive role under the broad provisions of art 32, supplementary means of interpretation, ‘Recourse may be had to supplementary means of interpretation, including [but not limited to], its preparatory works and the circumstances of its conclusion’. This possibility is explicitly affirmed in the Kasikili/Sedudu Island case,16 in the decision on the Sovereignty over Pulau Ligitan and Pulau Sipadan,17 and in several other decisions of the ICJ.18 All these decisions point in the same direction, providing (p. 18) for, in more or less clear terms, depending on the case—a secondary, limited interpretive role for subsequent practice.

Article 31(3)(b) has also been interpreted in a second, less literal, way. Just a few years before the Kasikili/Sedudu Island decision, the ICJ stated in the advisory opinion on the Use of Nuclear Weapons in Armed Conflict:

[…] the imperatives associated with the effective performance of its functions, as well as its own practice, are all elements which may deserve special attention when the time comes to interpret these constituent treaties.19

The Court, in the same paragraph continued placing under art 31(3)(b) the use of the practice of the organization for the interpretation of its constituent treaty. However, this is a completely different, more liberal, interpretation of art 31(3)(b): there is no mention of the necessity of an agreement, and full awareness of the parties is also not required. Even if subpara (b) of art 31(3) is relatively clear in stating that the practice has to establish the agreement of the parties, liberal evocation of this disposition has allowed the Court to consider the practice of the UN organs to interpret the statute of an international organization, ie of entities that are not the subject of the statute, but rather created by it. This practice, according to the interpretation of art 31(3)(b) laid out in Kasikili/Sedudu Island, would not have been considered.

4.  Initial Attempts to Find the Focal Point

Why this apparent disparity in the interpretation of art 31(3)(b)? One can surmise that in reality these two findings are analogous since in the second hypothesis the practice of the international organization became relevant because member states did not oppose it, thus acquiescing to it. A similar argument is present in the Namibia advisory opinion of 1971.20 However, in the advisory opinion on the Use of Nuclear Weapons in Armed Conflict this was not the case: in the latter, the Court considered the subsequent practice of different organs of the organization in interpreting the treaty, but without making any reference either to the conduct of the member states at meetings of the party states, or to the lack of protests by the member states;21 additionally, the court did not uphold that the original intention of the parties was to grant the organization a certain level of discretion to ameliorate the statute. With none of these arguments even raised, the practice of the organization in the application of a treaty seems to have been decisive, and not the lack of reaction by the parties. Another ICJ decision, the Mazilu(p. 19) advisory opinion, is clear in confirming this point. It provides the following:

In practice … the United Nations has had occasion to entrust missions […] to persons not having the status of United Nations officials […] In all these cases, the practice of the United Nations shows that the persons so appointed, and in particular the members of these committees and commissions, have been regarded as experts on missions within the meaning of Section 22.22

The acquiescence argument is not used, and in fact it does not seem sufficient to bridge together the two different interpretations of art 31(3)(b) shown above.

Alternatively, this difference could be explained by the different nature of the treaty under interpretation: one decision dealt with boundaries, and states have a keen interest in treaties which stabilize the delineations of their territories;23 the other dealt with the statute governing an organization, which can improve or modify its way of functioning over time. Even if international law provides for a single set of rules of interpretation, the nature of the interpreted treaty may call for different approaches in its interpretation, and for different thresholds in the establishment of interpretive values of subsequent practice. The fact that different kinds of treaties call for different interpretive approaches was introduced by the ICJ itself, with regard to the statutes of international organizations.24 In 1992, in the case between El-Salvador and Honduras, when art 31(3)(b) was mentioned and analysed by the ICJ for the first time, the Court did not consider the practice of the parties because of the special character of agreements establishing the jurisdiction of the court.25

This explanation is more convincing than the previous one and allows us to reach the first conclusion: the wording of art 31(3)(b) is not suitable for the interpretation of every kind of treaty. However, the ICJ’s two interpretations of art 31(3)(b) raise many other relevant issues that deserve further attention, which this first conclusion fails to fully resolve.

5.  Looking for the Right Position: Casting Some Glances at Other Canvases Nearby

In order to attain flexibility, it might be possible to stretch art 31(3)(b) when needed to the point that its letter is disregarded, as the advisory opinion on the Legality of the (p. 20) Use of Nuclear Weapons in Armed Conflict has suggested. More often, the ICJ has simply preferred to consider subsequent practice without qualifying it under any disposition of the Vienna Convention. For example, in the Jan Mayen case of 1993, the ICJ, in interpreting a treaty of 1965, referred to practices (including other international treaties on similar matters) that happened after the signature of the treaty under interpretation. The Court affirmed generically that ‘it is also appropriate to take into account, for purposes of interpretation of the 1965 Agreement, the subsequent practice of the Parties’.26 Other similar examples can be seen in the Land and Maritime Boundary between Cameroon and Nigeria case,27 and in the Armed Activities in Congo case of 2005. In the latter decision the ICJ addressed the interpretation of a protocol between Congo and Uganda, and without engaging in any analysis of the text, merely affirmed:

The two parties agreed that their respective armies would ‘co-operate in order to insure security and peace along the common border’. […] The Court believes that both the absence of any objection to the presence of Ugandan troops in the DRC in the preceding months, and the practices subsequent to the signing of the Protocol, support the view that the continued presence of Ugandan troops would be permitted by the DRC by virtue of the Protocol.28

From all these cases it is clear that subsequent practice can play more roles than the one specifically envisaged by art 31(3)(b). In the aforementioned case, the practice on the ground proved the existence of (or constituted) the agreement of the parties as the effective territorial scope of the treaty.29 In the Jan Mayen case of 1993, the reference to other legal instruments served as an aid in understanding the meaning generally intended by the parties, and therefore also most likely the one used in the treaty in question (considering a practice not established ‘in the application of the treaty’). In the advisory opinions quoted in the previous sections, practice was indicative of the interpretation of the statute by an international organization through the conduct of its organs (rather than that of the parties). These cases illustrate different uses of different subsequent practices and none of them refer to art 31(3)(b). The kind of practice considered changes according to what has to be clarified: the will of the parties with respect to a specific treaty; the meaning generally intended by the parties; or a possible meaning of a treaty emerging in practice, found by reference to those applying it. Regarding the last one, in the 1986 arbitration on the Filleting within the Gulf of St Lawrence, even a unilateral practice was considered relevant.30 While art 31(3)(b) codifies only the first usage, subsequent practice can be used—and has been used—also in these other interpretive dynamics.

(p. 21) It should be noted that subsequent practice is not just one of many means of interpretation, but has also been used by judges and arbitrators to illuminate the means of interpretation themselves: for example, it helped to clarify the object and purpose of a treaty,31 and to interpret in a systemic way the treaty as a whole.32 Finally, the role of subsequent practice is not only interpreting a treaty; certain decisions have found that the subsequent conduct of the parties had actually modified the agreement.33 This last point raises a discussion on its own merits, but this lies outside the scope of this chapter.

Taking a look back in time reveals that international jurisprudence has supported a broader way to consider subsequent practice for the purposes of interpreting a treaty in two distinct periods: first in the early 20th century (see, among many, the Affaire de l’impôt japonais of 1905);34 and secondly during the years of codification of the law of treaties (see, among many, the opinion on the Air transport services agreement (USA/Italy) of 1965).35 It was only during the years between the First World War and the 1950s that a voluntaristic vision of international law attributed a much more limited role to subsequent practice, especially by proposing that it is only relevant when it occurred at the time of the conclusion of the treaty.36

The same pattern emerges from the previous works of the ILC on the law of treaties and in other codifications of the customary rules on the interpretation of treaties: the second paragraph of art 71 of the Third Report on the Law of Treaties of 1964 prepared by Waldock contained a less structured reference to subsequent practice than the final version we know;37 the latter came out later, in the reorganization of the articles. Other documents, like the Resolution on the interpretation of treaties of the Institut de droit (p. 22) international of 195638 and the US restatement of foreign relations law, contain a more general reference to subsequent practice.39

6.  Still Looking for the Right Position: Casting Some Glances at Other Details on the Canvas (Art 31(1), (4), and Dynamic Interpretation)

Even without stretching and misquoting art 31(3)(b), the important functions of subsequent practice can be inferred from the Vienna Convention: especially from the first paragraph of art 31, ‘A treaty shall be interpreted in good faith’, and from the fourth paragraph, ‘A special meaning shall be given to a term if it is established that the parties so intended’. This last disposition refers in particular to a special meaning of the terms at the moment of the conclusion of the treaty.40 The subsequent practice can show the meaning intended by the parties at the moment of the conclusion of the agreement, however, apparently no cases have referred to subsequent practice under art 31(4). Similarly, there are not decisions that rely on subsequent practice through a reference to the good faith principle (art 31(1)), even if an undisputed practice can be seen as part of the continuous shaping of the treaty obligations performed over time in good faith.41

This leads to a partial overlap with the so-called dynamic interpretation, another area in which subsequent practice is relevant.42 An interpretation is dynamic when a term or an expression takes on a different meaning than the one originally agreed upon by the parties. Such can be due to several reasons: the legal meaning of a term, or its meaning in the common language, changing after the time of the conclusion of the agreement (evolutive interpretation, as the ICJ labels it).43 They can also be provoked by systemic factors (like a change in the legal framework or in circumstances),44 or by new considerations related to the object and purpose.45 All of these changes are determined by developments independent of the intentions of the parties.46

The evolution of a term, however, can also be determined by a change in the parties’ intentions. Even in interpreting a multilateral treaty of universal scope, the ICJ has stressed that the aim of interpreting treaties is to ascertain the intentions of the (p. 23) parties.47 The exclusive reference to a changed meaning, to the object and purpose of a treaty, and/or to changed legal framework and circumstances can result in the imposition of an external evaluation of this evolution upon the parties. To also take the subsequent practice of the parties into account in the assessment of this process allows for a consideration of how the parties themselves, that is, the subjects primarily involved in the application of the treaty, have faced and managed this change. It allows the intentions of the parties and a dynamic vision of international law to be taken together, avoiding the risk of arbitrary interpretations of a text, but also breaking up the common, static couplet of loyalty to the parties’ intentions and strict textualism.

This question recently emerged in the Navigational and Related Rights case, and the Court, instead of relying on the parties’ subsequent practice (as certain judges’ individual opinions suggested), used the legal fiction of original intention of the parties to adopt an evolutive concept.48 An example of an assessment of the evolution of a treaty through subsequent practice can be found in the ICJ’s advisory opinion on the Construction of a Wall in Palestine. In that case the Court referred to the practice of both the Security Council and the General Assembly to show how the meaning of art 12 of the Charter had evolved.49 Compared to the former case, the latter does not presume an original will to adopt an evolutive concept, but the latter interpretation is objectified in the sense that it is grounded on the practice, and on the practice of particular subjects—those involved in the application of the treaty: once again, as in the aforementioned opinion on the Use of Nuclear Weapons in Armed Conflict, it is not grounded on the practice of the parties.

In these cases, it is clear that subsequent practice can both display the original meaning of an agreement, and also demonstrate how intentions (and meanings) evolve and are crystallized in real time, revealing, in the midst of controversy, the existence of an ‘undisputed meaning’.

7.  Stepping Back and Enlarging the View

The difference between the Kasikili/Sedudu Island decision and the advisory opinion on Nuclear Weapons, aside from showing the inadequacy of art 31(3)(b) in regulating the interpretations of every kind of treaty, opens on another fundamental question: in the interpretation of treaties, is what matters only the subjects of the treaty, as the reference to the parties in art 31(3)(b) suggests, or also the treaty itself, the object of interpretation, and what relates to it?

To focus interpretation on the agreement between the parties would mean focusing on the practices, beliefs, and actions of the state parties, those that have the power to modify and interpret the treaty. This is the most common approach, and also the most (p. 24) consonant with the wording of art 31(3)(b). It is worth noting that one arbitral tribunal took a radically different and even stricter position on this question: in the award on the Protection of the Rhine from Pollution, the Tribunal specified that what really matters for the interpretation of a multilateral treaty are practices between the parties to the dispute only (and not between the parties to the treaty).50 Such a restrictive attitude privileges an approach more focused on the resolution of the dispute between the parties than on the interpretation of a multilateral treaty; it implicitly refers to the validity of inter se interpretive practices, a possibility allowed under international law,51 but can be criticized as it excludes many other potential factors relevant to interpretation, including what other parties (that have the same interpretive power of those involved in the dispute) to the same multilateral treaty have said and done.

On the other hand, considering the practice of the organs of an international organization means that the interpretation is centred on the text to be interpreted, and that conduct undertaken by any subject of international law (a state party, another state, the organs of an international organization), in the application of the treaty can also be relevant,52 that is by those more involved in its application. That was the case in the opinions on Mazilu, on the Use of Nuclear Weapons in Armed Conflict, and on The Construction of a Wall in Palestine.53 However, this approach can lead to further consequences: in the arbitration on the Bank for International Settlements,54 the tribunal interpreted a statute by making reference to recommendations and memoranda internal to the organization, as well as the reports of the private banks that had dealt with the interpreted provisions.55 The relevant inquiry was not on the parties and their conduct, but on the treaty and its application, even by private entities.

In all of these cases, the interpretation is centred on the text to be interpreted, and on the conduct undertaken by any entity that applied the treaty, whoever the involved actor may be. While this argument cannot be pushed too far, and all these assessments must be considered in the particular context and circumstances of the given case, it remains attractive to the extent that, by permitting consideration of practices by the subjects most involved in the application of a treaty, it allows much additional concrete evidence to be taken into account and recognizes the interpretive potential of subsequent practice, in any way qualified. The relative weight of this conduct must of course be judged by the interpreters, and assessed though the principle of good faith, a principle that is at the heart of treaty interpretation and at the base of subsequent practice (this relative weight, one might also note, is provided for in the non-mandatory (p. 25) language at art 31(3) ‘to take into account’).56 To avoid considering such practices at all because of the strict wording of art 31(3)(b) appears to be unjustifiable.

8.  A Blurred Figure: Subsequent MoUs

The last point on which I would like to touch is the way in which a particular kind of practice, namely non-binding agreements, is considered by jurisprudence.57 In non-binding agreements (memoranda of understanding, MoU) both the convergence of the parties and the fact that the parties did want to agree on something not binding are clear. The conclusions about their interpretive value are varied. In the case on Navigational and Related Rights, the ICJ did not take a position on the interpretive value of a memorandum of understanding, while Judge Skotnikov, in his dissenting opinion, and Guillaume, in his declaration, discussed its interpretive value, and labelled it as subsequent practice.58

The arbitral practice is richer. In the arbitration on the Protection of the Rhine from Pollution, both parties stressed that the pourparlers among the parties to the protocol were evidence of disagreement among them, but the arbitrators declined to pronounce on this point.59 In the Heathrow Airport Arbitration, an arbitral tribunal stated that an MoU is ‘consensual subsequent practice of the Parties, and as such is certainly available to the Tribunal as an aid to the interpretation of Bermuda 2 and, in particular, to clarify the meaning to be attributed to expressions used in the Treaty and to resolve any ambiguities’.60 In the Iron Rhine award, an MoU signed in 2000 was used to interpret two treaties from the 19th century, in a way that is both evolutive and systemic:

[T]he March 2000 MoU is equally not treated as legally irrelevant. Principles of good faith and reasonableness lead to the conclusion that the principles and procedures laid down in the March 2000 MoU remain to be interpreted and implemented in good faith and will provide useful guidelines to what the Parties have been prepared to consider as compatible with their rights under art XII of the 1839 Treaty of Separation and the Iron Rhine Treaty.61

Afterwards, both in the Heathrow Arbitration and in the Iron Rhine case, the arbitrators cautiously stressed that the final interpretation was based on other points of law, and not just on the memoranda.62

(p. 26) In the Abyei Arbitration, the Arbitral Tribunal referred a memorandum and other agreements under art 31(3)(b), and then explained that, in any case, they can be qualified under the other rules mentioned at subpara (c).63

This question does not seem determinative: the interpretive value of an MoU mainly depends on its content. A memorandum can be an understanding explicitly dedicated to the interpretation of a treaty, it can be part of the initiatives related to it, or just an element that clarifies the legal relationships of the parties on a certain point. Thus, the interpretive weight of such an understanding does not depend on its form, rather, as already observed (section 6 above), on the relationship it bears to the interpreted agreement, and on the purpose for which it is intended.

9.  How Not to be Unreasonable

My conclusions are mostly already scattered throughout the previous sections. To sum up, the wording of art 31(3)(a) is very clear, and is dedicated to a circumstance (a subsequent agreement explicitly interpretive of the previous one) that does not create any confusion. In dealing with subsequent practice, however, things are much more complicated. It is reductive to assert that the only possible use of subsequent practice falls under art 31(3)(b). Case law shows us that subsequent conduct has broader interpretive potential: not only when the practice culminates in an agreement; not only when the practice originates in the application of the treaty; not only when the practice is that of the parties.

The first conclusion that has been highlighted is that the way in which art 31(3)(b) is formulated does not fit the interpretation of every kind of treaty (see section 4 above).

A second conclusion is that subsequent practice serves several functions in dealing with a treaty. Which practice can be considered depends not on a rule, but rather on the question it is used for: a concordant practice of the parties in applying a treaty is a form of authentic interpretation, and can even modify a treaty. But in order to ascertain a meaning, even subsequent practice relating to other agreements has been considered (therefore not in the application of the treaty at stake). In choosing from possible meanings, even a unilateral practice can be useful. Subsequent practice has been used also to clarify the means of interpretation themselves (what the object and purpose of a treaty are; how to interpret, in a systemic way, the treaty as a whole) (for all these points see section 5 above).

Thirdly, what practice can be considered also depends upon which moment in time is relevant for knowing the intention of the parties: whether the purpose is to ascertain the original intent at the moment of the treaty signing, or to ascertain the emergence of an undisputed meaning (section 6 above).

Fourthly, the practice of those who are more involved in the application of a treaty, those who apply the treaty, depending on the circumstances, can be relevant also, not merely the practice of those who have the power to formally amend the treaty, that is the parties to the treaty (see section 7 above).

(p. 27) No consistent attempt has been made by the ICJ or by arbitrators to frame these different uses of practice under other dispositions of the Vienna Convention, such as art 31(1) with the reference to good faith, or art 31(4) with the reference to a special meaning intended by the Parties. The potential of subsequent practice has only been properly considered by either misinterpreting art 31(3)(b), or by not mentioning the Vienna Convention at all. In this way, it has been possible to include an interpretive practice without strictly showing an agreement, or demonstrating that the practices originated in the application of a treaty (see sections 5 to 7 above).

However, these techniques entail a risk: arbitrators and judges might feel free to do what they want with these rules, as long as they invoke the high threshold envisaged by art 31(3)(b) when they do not want to consider a subsequent practice, and do not mention the Vienna Convention when they do want to consider any subsequent practice.

The articles on the interpretation of treaties are special rules: they bind both states and judges in their functions. If misapplied, they could become ‘unwieldy instruments instead of the flexible aids which are required’.64 Indeed, some good decisions have been the direct result of not having properly applied those rules.

Nevertheless, these articles can afford to be somewhat tamed—or better, perhaps, opportunely amended. They primarily serve two purposes: first, they touch upon all the potentially relevant factors that a good, complete interpretive reasoning has to take into account; in this, they also amount to a useful set of questions on which both the parties have to take position, facilitating their exchange of views during the proceedings. Yet the Articles may turn out to be a useless chaotic function: we know all the things that we must put into the equation, but we cannot predict in any way the result that it will come from their combination. However, considering the second purpose of the articles, the Vienna rules on the interpretation of treaties also provide a few useful guidelines protecting against any wholly arbitrary interpretive constructions. Certainly the interpretation of a treaty is never a completely free exercise: for example, considerations of logic, on the nature of the treaty and on the circumstances of the case always play an important role. However, the drafters of the Vienna Convention also made some further choices that provide some guidance in approaching a text. The principle explicitly placed at the base of the Vienna rules on the interpretation of treaties, the principle that is the substance of these articles and informs them, is good faith that, together with a reference to the text at the beginning of art 31, implies the primacy of the intentions of the parties. This is a first guideline. Secondly, the subordinate place given to the preparatory works separates interpretation from narrow originalism, opening the interpretation of a treaty to an assessment of the intentions of the parties and of the law in light of its purpose, in the progression of time, and any variations of law and circumstances.65

It seems that the actual wording of art 31(3)(b) can be criticized under both purposes of the Vienna rules: first, in being strict, as mentioned above, it fails to correctly codify (p. 28) important expressions of the good faith principle. It can also be criticized with regard to the second function of the articles: by restricting the potential interpretive role of subsequent practice it separates the parties (and, in general, those most involved in the life of a treaty) from the management of new circumstances. It is not surprising, therefore, that several decisions have made broad use of subsequent practice in interpreting treaties, disregarding art 31(3)(b), but at the same time, paradoxically, taking an interpretive approach that is closer to the principles underlying interpretation as expressed in the Vienna Convention; in fact, to base the rules of interpretation on the good faith principle, with a clear separation of textualism from originalism (the division between arts 31 and 32), while at the same time limiting the interpretive relevance of subsequent practice, is contradictory.


1  The ideas here expressed were originally thought, researched, and developed in the preparation of a doctoral dissertation defended in May 2009 in Milan. They are explored at length in L Crema, ‘La Prassi Successiva tra Interpretazione e Modificazione dei Trattati’ (PhD thesis, University of Geneva/Università degli Studi di Milano 2009). I would like to thank Robert Kolb and Tullio Treves, who advised my dissertation, and Eirik Bjorge, George Letsas, Georg Nolte, and Cesare Pitea for the insightful observations expressed in the preparation of this chapter. My heartfelt thanks also to Stefan Raffeiner, who from Bozen to Berlin, and passing through Rome, established a truly European bridge of ideas that overcame linguistic barriers.

Associate Researcher, Università degli Studi di Milano.

2  To list only a handful of recent works which treat the topic in general: R Kolb, Interprétation et Création du Droit International (Bruylant 2006); CF de Casadevante Romani, Sovereignty and Interpretation of International Norms (Springer 2007); U Linderfalk, On the Interpretation of Treaties (Springer 2007); R Gardiner, Treaty Interpretation (OUP 2008); A Orakhelashvili, The Interpretation of Acts and Rules in Public International Law (OUP 2008); M Fitzmaurice, O Elias, and P Merkouris (eds), Treaty Interpretation and the Vienna Convention on the Law of the Treaties: 30 Years On (Nijhoff 2010); to them can be added other works dedicated to the interpretive reasoning of specific international jurisdictions such as I van Damme, Treaty Interpretation by the WTO Appellate Body (OUP 2009); G Letsas, A Theory of Interpretation of the European Convention on Human Rights (OUP 2009), a great number of articles (see, among many, the ‘Symposium’ published in the third issue of the (2010) 11 EJIL 507, as well the special number of the (2011) 115 RGDIP 291) and the sections dedicated to the interpretation in the works on the law of treaties: O Corten and P Klein, Les Conventions de Vienne sur le Droit des Traités: Commentaire Article par Article (Bruylant 2006); M Villiger, Commentary on the 1969 Vienna Convention on the Law of Treaties (Nijhoff 2009); E Cannizzaro (ed) The Law of Treaties Beyond the Vienna Convention (OUP 2011).

3  UNGA ‘ILC Report on the work of its sixty-first session’ (2009) GAOR 64th Session, Supp No 10 UN Doc A/64/10 355. The ILC included the topic in its programme of work in 2008, establishing a Study Group on the topic of Treaties over Time, cf the UNGA ‘Report of the ILC on the work of its sixtieth session’ (5 May to 6 June and 7 July to 8 August 2008) (2008) UN Doc A/63/10 Chap XII(A)(4) paras 351–2; UNGA Res 63/123 (11 December 2008) UN Doc A/RES/63/123. The Commission, at its 64th session in 2012, renamed the topic Subsequent Agreements and Subsequent Practice in Relation to the Interpretation of Treaties and changed the format of the work from the framework of a study group to that of a Special Rapporteur.

4  To date, the VCLT has 111 Parties, but the rules of interpretation have been applied also to states that have not signed the Convention, like France, and to those that have not ratified it, such as the United States and Iran.

5  As stated in Vienna by R Ago, President of the Conference (and previously a member of the ILC during its work on the law of treaties), on the occasion of the articles’ approval: ‘[T]he Conference had successfully disposed of the most controversial and difficult subject in the whole field of the law of treaties, the question of the interpretation of treaties. The section on interpretation had been condensed into a few formulas which had been adopted unanimously by the Conference. When the section had first come before the International Law Commission, many had felt that it might be unwise for the Commission to embark on a codification of so difficult a subject. He himself had taken a more optimistic view and was most grateful to the Conference for having proved him right’, ‘United Nations Conference on the Law of Treaties, Official Records, Second session’ (9 April–22 May 1969) (7 May 1969) UN Doc A/Conf.39/11/Add.1 59 para 77.

6  There are few exceptions: I did not find references to them in the cases decided by the EFTA Court and by the SADC Tribunal. The Vienna rules on interpretation were declared customary in many other jurisdictions, see Golder v United Kingdom (1975) European Court of Human Rights Series A No 18, 14, paras 29–30; Dispute between Argentina and Chile concerning the Beagle Channel (1977) 21 RIAA 53, para 15; ‘Other treaties’ subject to the advisory jurisdiction of the Court (Art 64 American Convention on Human Rights), Advisory Opinion OC-1/82, Inter-American Court of Human Rights Series A No 1 (24 September 1982) 7, para 33; Arbitral Award of 31 July 1989 (Guinea-Bissau v Senegal) (Judgment) [1991] ICJ Rep 53, para 48; Opinion 1/91 [1991] ECR I-060101, para 14; WTO, United States: Standards for Reformulated and Conventional Gasoline—Report of the Panel (29 January 1996) WT/DS2/R para 6.7; Aplicación de medidas de salvaguardia sobre productos textiles (Res 861/99) del Ministerio de economia y obras y servicios publicos (Brasil v Argentina), 10 March 2000, MERCOSUR DSB, Laudo III, 3, <http://www.mercosur.int> (accessed 14 November 2012); Prof Peter Anyang’ Nyong’o and Others v Attorney General of Kenya and Others (Abdirahin Haitha Abdi and Others Interveners), 30 March 2007, East Africa Court of Justice, 1/2006, 10–11, <http://www.eacj.org/docs/judgements/EACJ_Reference_No_1_2006.pdf> (accessed 14 November 2012); Responsibilities and obligations of States sponsoring persons and entities with respect to activities in the Area (Advisory Opinion of 1 February 2011) ITLOS Case No 17, 20–1, paras 57–8, <http://www.itlos.org/fileadmin/itlos/documents/cases/case_no_17/adv_op_010211.pdf> (accessed 14 November 2012); Ethyl Corporations v Canada (Jurisdiction), 24 June 1998, 27, para 52, <http://www.naftaclaims.com/Disputes/Canada/EthylCorp/EthylCorpAwardOnJurisdiction.pdf> (accessed 21 November 2012); Prosecutor v Dusko Tadic (Decision on the Prosecutor’s motion requesting protective measures for victims and witnesses) IT-94-1-T (10 August 1995) para 18; Prosecutor v Bagosora et al (Decision on the Admissibility of the Prosecutor’s Appeal from the Decision of a Confirming Judge Dismissing an Indictment against Théoneste Bagosora and 28 Others) ICTR 98-37-A (8 June 1998) 690, paras 28–9; Situation in the Democratic Republic of the Congo, Prosecutor v Dyilo (Pre-Trial Decision on the Confirmation of Charge) ICC-01/04-01/06 (29 January 2007) Annotated Leading Cases ICT 2005–07, 704–05, paras 276–85.

7  For example, the rules are regularly applied to France, that did not sign the VCLT, Case concerning the audit of accounts between the Netherlands and France in application of the Protocol of 25 September 1991 Additional to the Convention for the Protection of the Rhine from Pollution by Chlorides of 3 December 1976 (France/The Netherlands) (2004) 25 RIAA 267.

8  In this sense it is explicit the Laguna del Desierto Arbitration (Argentina/Chile) (Award), 21 October 1994, ILR 113, 194, 70. On the VCLT and the interpretation of Security Council resolution, see MC Wood, ‘The Interpretation of Security Council Resolutions’ in (1998) 2 Max Planck YB UN L 73 and ICJ in Accordance with International Law of the Unilateral Declaration of Independence in Respect of Kosovo (Advisory Opinion) [2010] ICJ Rep 403, para 94. On the VCLT and the interpretation of regulations enacted by international organization, see Responsibilities and obligations of States sponsoring persons and entities (n 6) paras 59–60. On the VCLT and the interpretation of state contracts, Eurotunnel Arbitration (The Channel Tunnel Group Ltd and France-Manche SA v UK and France), Partial award, 30 January 2007, ILR 132, 34, para 92. On the VCLT and the interpretation of an agreement between an international organization and a state, see Firma Brita GmbH v Hauptzollamt Hamburg-Hafen, 25 February 2010, C-386/08, ECR 2010 I-01289. On the VCLT and the interpretation of agreements between states and liberation movements, see Delimitation of the Abyei Area (Sudan/Sudan People’s Liberation Movement-Army), Final Award, 22 July 2009, ILM 48, 1357, para 572.

9  See, inter alia, Award in the Arbitration regarding the Iron Rhine (‘Ijzeren Rijn’) Railway between the Kingdom of Belgium and the Kingdom of the Netherlands (2005) 27 RIAA 35, para 45: ‘It is now well established that the provisions on interpretation of treaties contained in Articles 31 and 32 of the Convention reflect pre-existing customary international law, and thus may be … applied to treaties concluded before the entering into force of the Vienna Convention in 1980’, emphasis added. A confront between this excerpt and the intervention of Roberto Ago at the Vienna Conference (n 5), is completely striking.

10  Decision regarding Delimitation of the Border between Eritrea and Ethiopia (2002) 25 RIAA 83, para 3.4.

11  The works of the ILC on this point are discordant. In 1964, the commentary on the article said: ‘The term “take account of” is used rather than “be subject to” or any similar term because, if the rule is formulated as one of interpretation, it seems better, at any rate in sub-paragraphs (a) [later rule of customary law] and (b) [later agreement], to use words that leave open the results of the interpretation’ (1690) ILC Yearbook vol 1964, 2. That expression is not commented on the report of the Commission to the General Assembly of 1966, which, on the contrary, affirmed: ‘[I]t is only logic which suggests that the elements in paragraph 3—a subsequent agreement regarding the interpretation, subsequent practice establishing the understanding of the parties regarding the interpretation and relevant rules of international law applicable in the relations between the parties—should follow and not precede the elements in the previous paragraphs. The logical consideration which suggests this is that these elements are extrinsic to the text. But these three elements are all of an obligatory character and by their very nature could not be considered to be norms of interpretation in any way inferior to those which precede them’ (1691) ILC Yearbook vol 1966, 2, emphasis added.

12  A very generic reference to other rules at subpara (c) already includes every subsequent agreement (explicit or tacit) mentioned at subparas (a) and (b), but while these two dispositions have a specific temporal connotation (subsequent), subpara (c) extends also to relevant rules precedent to the treaty under scrutiny, as in the case Certain Questions of Mutual Assistance in Criminal Matters (Djibouti v France) (Judgment) [2008] ICJ Rep 177, paras 112–14, in particular at para 113. Moreover, subpara (a) includes subpara (b)—a subsequent practice such as the one described at subpara (b), that is, a tacit interpretive agreement, also falls under the generic expression subsequent interpretive agreement used at the subpara (a); a subsequent agreement is such independently by its form, MK Yassem, ‘L’interprétation des traités d’après la Convention de Vienne’, Rec des Cours (1976) 3 vol 151, 45; Gardiner, Treaty Interpretation (n 2) 219, 222.

13  Specifically on this question, C de Visscher, Problèmes d’Interprétation Judiciaire en Droit International Public (Pedone 1963) 121ff; JP Cot, ‘La conduite subséquente des parties a un traité’ (1966) 70 RGDIP 632; O Casanovas La Rosa, ‘La modificatiòn de los acuerdos internacionales por la pràtica posterior’ (1968) 11 Rev Esp de Derecho Internac 330; I Voïcu, De l’Interprétation Authentique des Traités Internationaux (Pedone 1968); E Menzione, ‘La modificazione dei trattati per via della condotta successiva delle parti’ (1971) 25 Dir Internaz 440; GP McGinley, ‘Practice as a Guide to Treaty Interpretation’ (1985) 9 Fletcher Forum 211; F Capotorti, ‘Sul valore della prassi applicativa dei trattati secondo la Convenzione di Vienna’ in Le Droit International à l’Heure de sa Codification. Etudes en l’Honneur de Roberto Ago, vol 3 (Giuffrè 1987) 196ff; G Distefano, ‘La pratique subséquente des Etats parties à un traité’ (1994) 40 AFDI 41; R Kolb, ‘La modification d’un traité par la pratique subséquente des parties’ in (2004) 14 RSDIE/SZIER 9. After the insertion of the topic ‘Treaties over time’ at the ILC (n 3), this topic knows now a new spring, J Arato, ‘Subsequent Practice and Evolutive Interpretation: Techniques of Treaty Interpretation over Time and Their Diverse Consequences’ (2010) 9 LPICT 443; AM Feldman, ‘Evolving Treaty Obligations: A Proposal for Analyzing Subsequent Practice Derived from WTO Dispute Settlement’ (2009) 41 NYU J Intl L & Pol 655; H Fox, ‘Art 31(3)(a) and (b) of the Vienna Convention and the Kasikili/Sedudu Island Case’ in Fitzmaurice, Elias, and Merkouris, Treaty Interpretation and the Vienna Convention on the Law of the Treaties (n 2) 59ff; A Roberts, ‘Power and Persuasion in Investment Treaty Interpretation: The Dual Role of States’ in (2010) 104 AJIL 179.

14  An example of such a way of applying arts 31–2 can be seen in Sovereignty over Pulau Ligitan and Pulau Sipadan (Indonesia/Malaysia) (Judgment) [2002] ICJ Rep 625. Linderfalk, On the Interpretation of Treaties (n 2) proposes a very dogmatic interpretation and way of applying these Articles.

15  Kasikili/Sedudu Island (Botswana/Namibia) (Judgment) [1999] ICJ Rep 1045, para 74.

16  ‘[T]he Court is bound to note that … surveys carried out on the ground identified the channel of the Chobe to the north and west as the “main channel” … The Court finds that these facts, while not constituting subsequent practice by the parties in the interpretation of the 1890 Treaty, nevertheless support the conclusions which it has reached by interpreting Article III, paragraph 2, of the 1890 Treaty in accordance with the ordinary meaning to be given to its terms’, Kasikili/Sedudu Island (n 15) para 80, emphasis added.

17  Sovereignty over Pulau Ligitan and Pulau Sipadan (n 14) para 80.

18  Legality of the Use by a State of Nuclear Weapons in Armed Conflict (Advisory Opinion) [1996] ICJ Rep 66, para 27; Pulp Mills on the River Uruguay (Argentina v Uruguay) (Request for the Indication of Provisional Measures: Order) [2006] ICJ Rep 113, para 53: in this order on precautionary measures the question of subsequent practice was addressed by Uruguay, and was mentioned by the Court, although this interpretive argument was only cursorily recalled by the Court in its final conclusions. See also Dispute Regarding Navigational and Related Rights (Costa Rica v Nicaragua) (Judgment) [2009] ICJ Rep 213, para 56, in which a translation made in 1887 of a treaty of 1858 was considered a support of an interpretation already established.

19  Legality of Nuclear Weapons in Armed Conflict (n 18) para 19. The relevant practice considered by the ICJ consisted of some resolutions of the WHO Assembly, a report of the General Director and a report of the Management Group, Legality of Nuclear Weapons in Armed Conflict (n 18) para 27.

20  Legal Consequences for States of the Continued Presence of South Africa in Namibia (South West Africa) notwithstanding Security Council Resolution 276 (1970) (Advisory Opinion) [1971] ICJ Rep 16, para 22: ‘This procedure followed by the Security Council […] has been generally accepted by Members of the United Nations and evidences a general practice of that Organization’.

21  The relevant practice considered by the ICJ was composed also by a report of the General Director and a report of the Management Group, Legality of Nuclear Weapons in Armed Conflict (n 18) para 27.

22  Applicability of Article VI, Section 22, of the Convention on the Privileges and Immunities of the United Nations (Advisory Opinion) [1989] ICJ Rep 177, para 48.

23  See, mutatis mutandis, Territorial and Maritime Dispute between Nicaragua and Honduras in the Caribbean Sea (Nicaragua v Honduras) (Judgment) [2007] ICJ Rep 659, para 253. The interplay between treaty and effectivités, and the need for stability of boundaries, call for special regard. The leading case on decolonization, Frontier Dispute (Burkina Faso/Republic of Mali) (Judgment) [1986] ICJ Rep 554, in drawing the relationship between title and effectivités, followed a pattern similar to the rules on interpretation of the VCLT, para 63. The decisions of the Eritrea/Ethiopia Boundary Commission went further, envisaging a possible modifying effect of subsequent effectivités, Eritrea and Ethiopia Boundary Decision (n 10) 110–12, paras 3.6–3.12.

24  See Certain Expenses of the United Nations (Article 17, paragraph 2, of the Charter) (Advisory Opinion) [1962] ICJ Rep 151, 157.

25  The Court, to interpret the Special Agreement (24 May 1986) and a reference that it had to the General Peace Agreement between the same states (30 October 1980), analysed the consistent conduct of the parties related to that second treaty. Because according to the Court the special character of an agreement establishing jurisdiction calls for a strict, textual, interpretation, that practice was not considered, and art 31(3)(b) was not applied, Land, Island and Maritime Frontier Dispute (El Salvador/Honduras: Nicaragua intervening) (Judgment) [1992] ICJ Rep 351, paras 376–80.

26  Maritime Delimitation in the Area between Greenland and Jan Mayen (Denmark v Norway) (Judgment) [1993], ICJ Rep 38, para 28.

27  Land and Maritime Boundary between Cameroon and Nigeria (Cameroon v Nigeria) (Preliminary Objections) [1998] ICJ Rep 275, paras 61–6.

28  Armed Activities on the Territory of the Congo (Democratic Republic of the Congo v Uganda) (Judgment) [2005] ICJ Rep 168, para 46.

29  See also the similar Cameroon and Nigeria Boundary Dispute (n 27) para 67.

30  Case concerning filleting within the Gulf of St Lawrence between Canada and France (1986) 19 RIAA 225, paras 33–4, referring for purposes of interpretation to Canadian conduct. On unilateral practice, see Cot, ‘La conduite’ (n 13) 645; W Karl, Vertrag und spätere Praxis im Völkerrecht (Springer 1983) 143; more recently see Kolb, Interprétation et Création (n 2) 488.

31  Oil Platforms (Iran v USA) (Preliminary Objection) [1996] ICJ Rep 803, paras 29–30.

32  Constitution of the Maritime Safety Committee of the Inter-Governmental Maritime Consultative Organization, (Advisory Opinion) [1960] ICJ Rep 167–9.

33  See the historical Affaire de l’impôt japonais sur les bâtiments (Japon/Allemagne, France et Grand Bretagne) (1905) 11 RIAA 51–4; Case concerning the interpretation of the air transport services agreement between the United States of America and France, signed at Paris on 27 March 1946 (France/USA) (1963) 16 RIAA 5, 60–2; more recently Eritrea/Ethiopia Boundary Commission (n 10) 110–12, paras 3.6–3.12.

34  Affaire de l’impôt japonais (n 33) 51–4; see also Samoan Claims (Germany, Great Britain, USA) (1902) 9 RIAA 15, 25; Affaire relativeàl’Interprétation du traitéde commerce conclu entre l’Italie et la Suisse le 13 juillet 1904 (Italie/Suisse) (1911) 11 RIAA 257, 262, and the comments of R Kolb, La Bonne Foi en Droit International Public (PUF 2000) 275. The Chamizal Case (Mexico v United States) (1911) 5 AJIL 782, 805; Affaire de l’indemnitérusse (Russie/Turquie) (1912) 11 RIAA 421, 431.

35  Interpretation of the Air Transport Services Agreement (Italy/USA), Advisory Opinion (1965) 16 RIAA 75, 99–100. See also Case concerning the Temple of Preah Vihear (Cambodia v Thailand) (Merits) [1962] ICJ Rep 6, 32–3; Interpretation of the Air Transport Services Agreement (France/USA) (1963) RIAA 5, 60–2.

36  Treaty of Neuilly‚Article 179‚Annex‚Paragraph 4 (Interpretation) PCIJ Rep Series A No 3, 9; Interpretation of Article 3‚Paragraph 2‚of the Treaty of Lausanne (Advisory Opinion) PCIJ Rep Series B No 12, 24; see the comments of CC Hyde, International Law Chiefly as Interpreted and Applied by the United States (2nd edn, Little Brown 1945) 1497; Cot, La conduite (n 13) 639–41; and the intervention of ILC (1964) ILC Yearbook vol 1, 286.

37  ‘Reference may be made to other evidence or indications of the intentions of the parties and, in particular, to the preparatory work of the treaty, the circumstances surrounding its conclusion and the subsequent practice of parties in relation to the treaty, for the purpose of—a) confirming the meaning of a term […] b) determining the meaning of a term […] c) establishing the special meaning of a term’, commented in Third Report on the Law of Treaties (Waldock) (1964) ILC Yearbook vol 2, 52–62.

38  (1956) Annu de I’Inst de Droit Intl 364–5, art 2: ‘1. Dans le cas d’un différend porté devant une juridiction internationale il incombera au tribunal, en tenant compte des dispositions de l’article premier, d’apprécier si, et dans quelle mesure, il y a lieu d’utiliser d’autres moyens d’interprétation. 2. Parmi ces moyens légitimes d’interpréter se trouvent: […] b) La pratique suivie dans l’application effective du traité’.

39  L Henkin (ed), Restatement of the Law, Third, The Foreign Relations Law of the United States (Revised), vol 1 (American Law Institute 1987) 196–204.

40  Draft Articles on the Law of Treaties (1696) ILC Yearbook vol 1966, 2.

41  Kolb, Bonne Foi (n 34) 275.

42  On this topic, see M Fitzmaurice, ‘Dynamic (Evolutive) Interpretation of Treaties (Part I)’ (2008) 21 Hague YB Intl L 101, and ‘Dynamic (Evolutive) Interpretation of Treaties (Part II)’ (2009) 22 Hague YB Intl L 3, and the remarkable reflections of G Distefano, ‘L’interprétation évolutive de la norme internationale’ (2011) 115 RGDIP 373.

43  Navigational and Related Rights (n 18) paras 63–71.

44  Advisory Opinion on Namibia (n 20) para 53.

45  Iron Rhine (n 9) 72–4, paras 79–81.

46  Distefano, ‘L’interprétation évolutive’ (n 42) distinguishes between Interprétationévolutive endogène and exogène.

47  See the interpretation of the Fourth Geneva convention, Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory (Advisory Opinion) [2004] ICJ Rep 136, paras 94–5. In an arbitral tribunal, about a multilateral treaty, see Protection of the Rhine from Pollution (n 7) 295–6, para 62.

48  Navigational and Related Rights (n 18); see on the one hand the opinion of the majority, paras 64–71 and on the other hand the dissenting opinion of Judge Skotnikov, Navigational and Related Rights (n 18) (Separate Opinion of Judge Skotnikov) [2009] ICJ Rep 283, paras 9–10 and Navigational and Related Rights (n 18) (Declaration of Judge ad-hoc Guillaume) [2009] ICJ Rep 290, para 16.

49  Advisory Opinion on a Wall in Palestine (n 47) paras 27–8.

50  Protection of the Rhine from Pollution (n 7) 312, para 101 (see also para 137).

51  Possibility that can be inferred also from art 41 VCLT, Agreements to modify multilateral treaties between certain of the parties only.

52  Judge Spender, in his separate opinion, vigorously dissented from this treatment, Certain Expenses (n 24) (Separate Opinion of Judge Sir Percy Spender) [1962] ICJ Rep 182, 189–92, commenting the position endorsed by the Court at Certain Expenses (n 24) 157–61.

53  See Privileges and Immunities of the United Nations (n 22) paras 18, 47.

54  Permanent Court of Arbitration, Dr Horst Reineccius (Claim N 1), First Eagle SoGen Funds, Inc (Claim N 2), Mr Pierre Mathieu and La sociétéde concours hippique de La Châtre (Claim N 3) v Bank for International SettlementsPartial Award (22 November 2002) 89, para 193 <http://www.pca-cpa.org/showpage.asp?pag_id=1157> (accessed 29 November 2012).

55  Permanent Court of Arbitration (n 54) paras 172ff, especially at paras 193–4.

56  See n 11.

57  Another grey area involves the interpretive role of subsequent agreements on different matters signed between the same parties, and has already been explored, see Kolb, Interprétation et Création (n 2) 463–7 (treaties as part of the context latior sensu) and Gardiner (n 2) 281–3.

58  See Navigational and Related Rights (Separate Opinion of Judge Skotnikov) (n 48) para 9 and Navigational and Related Rights (Declaration of Judge ad-hoc Guillaume) (n 48) para 16.

59  Cf the positions of the parties, Protection of the Rhine from Pollution (n 7) 283, 289, paras 38, 48.

60  United States-United Kingdom Arbitration concerning Heathrow Airport User Charges (USA/UK), Award on the First Question (1992) 24 RIAA 3, para 6.7.

61  Iron Rhine (n 9) 98 para 157.

62  Heathrow Arbitration (n 60) 131, Ch 6, para 6.7: ‘The MoU is therefore available to the Tribunal as a potentially important aid to interpretation but is not a source of independent legal rights and duties capable of enforcement in the present Arbitration’. Iron Rhine (n 9) 98, para 157: ‘The respective allocation of costs … will depend not upon any undertakings given in the March 2000 MoU, but on other legal considerations (including what the Parties have thought reasonable during their negotiations in connection with the March 2000 MoU)’.

63  Cf Abyei Arbitration (n 8) 1369–70, paras 650–5.

64  I Brownlie, Principles of Public International Law (7th edn, OUP 2008) 602.

65  For this reading of the Vienna rules on the interpretation of treaties, see Délimitation de la frontière maritime entre la Guinée et la Guinée-Bissau (Guinea/Guinea Bissau) (1985) 19 RIAA 149, paras 66–84, especially para 68.