‘Relevant Rules’ as Normative Environment: Harmony vs Cacophony in the ITLOS Advisory Opinion on Climate Change

On 21 May 2024, the International Tribunal for the Law of the Sea (ITLOS) delivered its much anticipated Advisory Opinion on Climate Change. Other blog-posts have already dealt with various aspects of the Advisory Opinion (see, for instance, here and here), including the Tribunal’s approach to interpreting the United Nations Convention on the Law of the Sea (UNCLOS) (here, here and here). This post zeroes in on one particular interpretative issue, and its wider ramifications for the development of international law, namely the Tribunal’s approach to Article 31(3)(c) of the Vienna Convention on the Law of Treaties (VCLT) (which enshrines the principle of systemic integration) in connection with the interpretation of UNCLOS. Although ITLOS did not elaborate in detail on its approach, as can be seen from its entire analysis, the Tribunal has demonstrated a clear and principled choice with respect to the content and application of Article 31(3)(c) VCLT and its customary counterpart.

Points of Entry of External Rules

ITLOS devotes Section V of its Advisory Opinion to ‘Interpretation of the Convention and the relationship between the Convention and external rules’ (paras. 128-137). It notes, in particular, three points of entry, through which ‘external rules’ may inform the interpretation of UNCLOS provisions. These are: i) rules of reference, i.e., when UNCLOS provisions refer explicitly to external rules; ii) Article 237 UNCLOS, which ‘reflects the need for consistency and mutual supportiveness between the applicable rules’ (para. 133) and iii) Article 31(3)(c) VCLT, according to which account shall be taken of “any relevant rules of international law applicable in the relations between the parties” (para. 135).

Relevant Rules Across Non-Identical Parties

In the following sections, I will focus on the third point of entry, i.e. Article 31(3)(c) VCLT. ITLOS, although not going into detail, seems to take the correct approach of focusing more on whether the rule is ‘relevant’ than whether it is applicable between the ‘parties’, or at least takes quite a liberal view of what ‘parties’ means. Some tribunals and authors have argued that the correct interpretation of Article 31(3)(c) VCLT, and its customary counterpart, would be ‘parties to the treaty’ (for an analysis see here). This would require that all the parties of the treaty being interpreted, in this case UNCLOS, also be parties to the treaty being referred to under Article 31(3)(c). As far as customary international law is concerned (and general principles as well, since those also fall under Article 31(3)(c) VCLT) this is not an issue (unless, of course, we are dealing with a situation of persistent objection or regional customary law). However, issues may arise with such an approach to ‘parties’ where treaties are concerned.

This ‘parties to the treaty’ approach has been adopted by some tribunals, for example in the 2008 EC-Biotech case, and more recently in Vattenfall AB and others v Germany. However, several others have taken a more flexible approach and interpretation of ‘parties’, where the identity, near identity, or mere partial overlap of parties is just one of the many factors that can assist in determining relevance. For them the crucial aspect of Article 31(3)(c) is the term ‘relevant’ not the term ‘parties’ (see in more detail here). Consistent with this approach, in the Advisory Opinion on Climate Change, ITLOS rightly focuses more on the relevance of a particular treaty, rather than whether all UNCLOS parties have signed onto it, thus avoiding the very strict interpretation of Article 31(3)(c) VCLT seen in some other cases. This is evident from the treaties ITLOS refers to. In its interpretation of the content of UNCLOS provisions, the Tribunal referred to the United Nations Framework Convention (UNFCCC), the Kyoto Protocol, the Paris AgreementMARPOL, the Convention on Biological Diversity (CBD) and the Montreal Protocol, including the Kigali Amendment.

If one were to strictly interpret Article 31(3)(c) VCLT as referring only to ‘parties to the treaty’, some of the above treaties could not be taken into account under the principle of systemic integration. For instance, Yemen has signed and ratified UNCLOS, but not the Paris Agreement. Thus, by virtue of that one State, the Paris Agreement would not be a ‘relevant rule’ for the interpretation of UNCLOS according to strict understanding of the term ‘parties’ in Article 31(3)(c). The list grows longer with respect to the Kyoto Protocol, MARPOL, and the Kigali Amendment.

Consequently, while ITLOS did not elaborate on its interpretation of Article 31(3)(c), its decision to consider the Paris Agreement and the other instruments above reflects an expansive view of the article. The Tribunal’s approach conforms not only with pre-existing and current international case-law, but also with the spirit of Articles 31-33 VCLT. In this regard, the Tribunal noted:

“[t]he Tribunal is of the view that, subject to article 293 of the Convention, the provisions of the Convention and external rules should, to the extent possible, be interpreted consistently. In this context, the Tribunal notes that the Study Group of the International Law Commission (hereinafter “the ILC”), in its 2006 Report on the Fragmentation of International Law, concluded that “[i]t is a generally accepted principle that when several norms bear on a single issue they should, to the extent possible, be interpreted so as to give rise to a single set of compatible obligations” (Fragmentation of International Law, Report of the Study Group of the ILC, 2006, p. 8; see also Guideline 9 of the 2021 ILC Guidelines on the protection of the atmosphere.” (para. 136, emphasis added)

Harmony vs Cacophony

It is worth underscoring this principle of a harmonious interpretation, which ITLOS refers to without explicating on whether it is a separate principle, an outcome of Article 31(3)(c) VCLT, or whether Article 31(3)(c) VCLT is itself a manifestation of this wider interpretative principle. It seems that ITLOS is of the view that, by taking into account relevant external rules, we can minimize ‘apparent conflicts’ of norms, and that the treaty being interpreted can speak in one voice without at the same time creating a cacophony of voices when taken with other treaties or rules regulating similar subject-matters. Why is this important? Well, because it was precisely the opposite that was presented in Vattenfall AB and others v Germany. In that case, the European Commission argued that EU law could be considered a ‘relevant rule’ for the purposes of interpreting the Energy Charter Treaty (ECT). The investment tribunal held that referring to relevant rules to which not all parties to the ECT have ratified or acceded to

“would potentially allow for different interpretations of the same ECT treaty provision. The Tribunal considers that this would be an incoherent and anomalous result and inconsistent with the object and purpose of the ECT and with the rules of international law on treaty interpretation and application”(para. 155).

The investment tribunal further held that understanding ‘parties,’ as used in Article 31(3)(c) VCLT, as ‘parties to the treaty’ is mandated by “[t]he need for coherence, and for a single unified interpretation of each treaty provision” (para. 156). There are many problems with this reasoning. Perhaps most notably, the Vattenfall tribunal suggests that the principle of systemic integration and its term ‘parties’ should be understood as ‘parties to the treaty’, because otherwise this would lead to incoherence. In the investment tribunal’s view the same provision would have a different meaning “depending on the independent legal obligations entered into by one State or another, and depending on the parties to a particular dispute” (para. 156). However, it is questionable, to say the least, that a more flexible understanding of the principle of systemic integration would lead to incoherence. Moreover, this view presumes that the principle of systemic integration is entirely based on the term ‘parties’, and if that is satisfied immediately, the rule would be taken into account for interpretative purposes, but that is not the case. Article 31(3)(c) VCLT makes clear that external rules are only to be taken into account if they are “relevant.” This requires consideration of  a number of factors to determine whether other external rules can be of utility in the interpretative process.

Contrary to the suggestion in Vattenfall, a rigid interpretation of Article 31(3)(c), requiring all the parties of the treaty being interpreted to be also parties to the treaty being referred to, is the one more likely to lead to incoherence and different interpretations. This is because the ‘relevant rules’ that one would take into account would constantly change depending on the decision of even one State to either join or leave a treaty. Consider the above example of Yemen. If Yemen were to join the Paris Agreement, it could then be used to inform the interpretation of UNCLOS. But, if another party to UNCLOS were to subsequently withdraw from the Paris Agreement, it would again be deemed irrelevant. Similar considerations and results would apply if states withdrew from or joined UNCLOS (For a more detailed explanation of the problems with Vattenfall’s understanding of the principle of systemic integration see here, and for how relevance is determinative and determined see here.)

Clearly, such scenarios are untenable and will lead to significant uncertainty, lack of harmony, and interpretative cacophony. ITLOS clearly recognizes this. By referring to treaties that have not been ratified or accepted by all parties to UNCLOS, ITLOS has taken a principled and, in this author’s view, correct stance that for the purposes of Article 31(3)(c) VCLT relevance is the most important aspect (and not identity of parties). This will ensure harmony both within the provisions of UNCLOS but also of those provisions with their wider normative environment.

Evolutive Interpretation: ‘Relevant Rules’ of When?

One more relevant point relates to the character of UNCLOS as a living instrument. ITLOS noted

“that many participants in the present proceedings have emphasized the open character of the Convention and its constitutional and framework nature. In the Tribunal’s view, coordination and harmonization between the Convention and external rules are important to clarify, and to inform the meaning of, the provisions of the Convention and to ensure that the Convention serves as a living instrument” (para. 130).

There are some issues with relying on the ‘nature’ of a treaty to make a heuristic argument in favour of its evolutive interpretation. Such an interpretation is better substantiated by reference to the ‘generic’ nature of the treaty’s terms, its duration, its object and purpose and the intention of its parties (which ITLOS alludes to in various parts of its Opinion). The issue of evolutive interpretation has been addressed in other contributions to this Symposium (see Peel’s and Voigt’s contributions). I will thus restrict myself here to some thoughts on the importance of evolutive interpretation in connection with the external ‘relevant rules’. The Tribunal states:

“… article 31, paragraph 3(c), of the VCLT … requires that account be taken, together with the context, of any relevant rules of international law applicable in the relations between the parties. This method of interpretation ensures, as observed by the International Court of Justice …, that treaties do not operate in isolation but are “interpreted and applied within the framework of the entire legal system prevailing at the time of the interpretation” (Legal Consequences for States of the Continued Presence of South Africa in Namibia (South West Africa) notwithstanding Security Council Resolution 276 (1970), Advisory Opinion, I.C.J. Reports 1971, p. 16, at p. 31, para. 53)” (para. 135, emphasis added).

This is not entirely accurate. In the sentences preceding the quoted language, the International Court of Justice had concluded that the concepts embodied in the Covenant “were not static, but were by definition evolutionary, … [and that the] parties to the Covenant must consequently be deemed to have accepted them as such”. The International Court of Justice, thus, adopted an evolutive interpretation based on the evolutive nature of the Covenant’s terms and the intention of the parties. Only after that did it also say that the Covenant needs to be interpreted within the legal framework at the time of interpretation. This may seem like splitting hairs but it is an important distinction. The reason is that, when Article 31(3)(c) VCLT was being discussed, an early draft called for consideration of ‘relevant rules applicable in the relations between the parties at the time of the conclusion of the treaty’. This was later taken out on the understanding that the choice between relevant rules at the time of the conclusion of the treaty and relevant rules at the time of the interpretation of the treaty would be made on a case-by-case basis, depending on whether the treaty was to be interpreted contemporaneously or evolutively (see in more detail here, Chapter 4). ITLOS in the aforementioned passage seems to suggest that, as far as Article 31(3)(c) VCLT is concerned, it will be the rules at the time of interpretation that will always be relevant, but that is not accurate. This will depend on whether evolutive interpretation is allowed for under the particular provision or treaty. In the ITLOS Advisory Opinion reference to, for instance, the Kyoto Protocol and the Paris Agreement was allowed under Article 31(3)(c) VCLT, but only because the relevant UNCLOS provisions were subject to an evolutive interpretation.


Advisory Opinions like that issued by ITLOS are always closely scrutinized. There are, undoubtedly, elements of the ITLOS opinion that could have been more clearly or fully explained. Indeed, some of the judges themselves seem to recognize this. For example, in his Declaration, Judge Pawlak rightly lamented the fact that the Advisory Opinion did not reflect “the broader implications of recent developments in climate change justice, such as the decision of the UN Human Rights Committee issued on 22 September 2022 in the Torres Strait Islanders case and the very important judgment of the European Court of Human Rights of 9 April 2024 in the Verein KlimaSeniorinnen Schweiz and Others v. Switzerland case.”

Despite this, ITLOS should be applauded for taking a principled stance on interpretation. The Tribunal rightly viewed UNCLOS not as an isolated instrument but as part of a wider normative environment, which exercises an interpretative pull on it, just as UNCLOS also exercises an interpretative pull to that environment. It is only through such a systemic interpretative approach that we can ensure, especially for global issues such as climate change, that the multiplicity of instruments does not lead to incoherence, disorder and cacophony, but can also assist in achieving harmony and (as in a musical orchestra) dare I say, a symphony.

Panos Merkouris
Professor of International Law at University of Groningen

Panos is a Professor of International Law at University of Groningen and is the Principal Investigator of the TRICI-Law Project.