The Theory and Practice of Customary International Law: What’s Interpretation Got to Do with It?

By Nina Mileva
Published on 25 November 2022

In the introduction to this symposium, Kammerhofer and Merkouris end their post by reminding us that reports of the death of the conceptual and theoretical problems of customary international law (CIL) have been greatly exaggerated. Indeed, having read the insightful posts by Xuexia Liao, Alvin Yap, Michael Wood and Omri Sender, and Kristi How, I cannot help but emphatically agree.

Let me begin by thanking CIL Dialogues for hosting this symposium and providing us with a platform to discuss the many issues surrounding the theory and practice of CIL. I am also grateful to the authors of the four posts of the symposium, for taking some of their valuable time to engage anew with this subject. Customary international law has been at the centre of my research agenda for the better part of the last four years, and I am glad to read that the research which has occupied my daily life for so long finds resonance among colleagues.

That the four posts of this symposium each identify an aspect of the theory and practice of CIL that the authors feel deserves more scrutiny, could be seen as an indicator that our edited volume is a timely and relevant addition to the ongoing discourse. At the same time, I would have liked to have seen deeper engagement with the idea, presented in various chapters of The Theory, Practice, and Interpretation of Customary International Law, that CIL can be subject to interpretation as a separate operation distinct from its identification. In this response, I would like to lead with this idea, and engage with each post of the symposium from that perspective. In particular, I would like to engage with them by departing from the premise that CIL rules can be interpreted, and that their interpretation is an operation which can be distinguished from their identification.

The broader theoretical underpinnings of this premise can be found in the chapters of the edited volume authored by Merkouris, Fortuna, Di Marco, Ryngaert, and myself. For the purposes of the present post, the premise can be briefly sketched in the following way. Interpretation is the operation of determining the content and scope of customary rules and takes place after their identification through state practice and opinio juris. Interpretation can be differentiated from identification because its aim is not to ascertain the existence (or not) of a customary rule, but rather to determine how a general customary rule operates in particular circumstances. While there is theoretical opposition to the idea that CIL rules can be interpreted, its interpretation is ubiquitous in the practice of both international and national courts. An examination of this practice indicates that courts regularly engage in the interpretation of custom as separate from its identification, and often rely on interpretive methodologies familiar from the interpretation of other sources (such as treaties) only adapting them to the peculiarities of custom. Beyond the identification of examples where judges engage in the interpretation of custom, accounting for the process of interpretation bears a lot of theoretical relevance as well. In the absence of an interpretive process, there is no viable explanation for the continued existence and operation of customary rules. Moreover, when we do not account for interpretation as a possible operation in the application of CIL, or as a plausible analytical framework in the theory of custom, we run the risk of providing an incomplete picture of the way customary rules operate in the legal system. How, then, does this idea of custom interpretation interact with the issues identified in the four posts of the present symposium?

The first post by Xuexia Liao deals with the role of treaties in the identification of custom. Liao argues that courts rely on the text of treaties ‘not just as a starting point of assessing the customary process, but as a methodological shortcut in the identification procedure’. This, as Liao rightfully argues, is problematic. Particularly so when courts rely on the text of treaties to declare customary rules which would bind non-parties, or when courts rely on the relationship between treaty provision to declare the entire regime stipulated by a treaty-customary. Liao’s post raises valid concerns about the methodological soundness of courts’ reliance on treaties for the identification of CIL. The only insight I would wish to add to its analysis is that in certain scenarios the reliance on treaties by courts may not be methodologically problematic, if the court is doing this in order to interpret the customary rule. In particular, with certain treaties which are reflective of custom, courts may rely on their provisions in order to provide a lexical expression of a customary rule. In this sense, the courts may be said to compensate for the lack of text of CIL rules by relying on relevant treaty provisions as a substitute. This is not inherently problematic when these provisions are widely reflected in the practice of states. In such instances, courts may in fact be engaging in systemic integration or an in pari materia interpretation of the customary rule.

The second post by Alvin Yap focuses on the conceptual and practical problems in the change or termination of CIL rules. With respect to the relationship between the identification of a CIL rule and its change or termination, Yap notes that ‘in practice, once a rule has been identified to exist under customary international law, there appears to be a presumption that the rule continues to exist, at least for some time’. In this way, Yap argues, the identification of CIL rules by authoritative bodies may inhibit the process by which the rule develops or terminates. While the identification of CIL by bodies such as the International Court of Justice or the International Law Commission (ILC) carries substantive weight, this does not necessarily bar the development of the rule. To the contrary, a general customary rule identified by an authoritative body may continue to evolve through interpretation. As I have argued elsewhere in a joint post with Panos Merkouris, interpretation plays a central role in the evolution of a customary rule. It is through interpretation that the scope of an older CIL rule can be expanded to cover new situations. In this regard, it is not necessary to always look for new state practice or opinio juris in order to establish the development of a CIL rule. This is not to say that interpretation will account for instances of rule modification or termination, and Yap’s post raises valid concerns in this regard. Nevertheless, by accounting for interpretation, we can see that the identification of a general customary rule by an authoritative body does not necessarily inhibit its development, nor is it necessary to always look for new state practice or opinio juris to establish the evolution of the rule. Rather, interpretation allows for a natural degree of flexibility which is necessary for customary rules to continue operating in the constantly evolving international landscape.

The third post by Omri Sender and Michael Wood explicitly touches on the question of CIL interpretation. Sender and Wood argue that the subject of CIL’s interpretation is sufficiently addressed by conclusion 2 of the ILC Conclusions on Identification of Customary Law. Moreover, they observe that ‘the notion of interpretation, which is commonly employed in the context of giving meaning to a text, appears intrinsically inapplicable to the unwritten rules of [CIL]’. While conclusion 2 is an authoritative guideline when it comes to the identification of a customary rule, it offers little guidance on what happens to the CIL rule after identification. If we only follow the adage of the two-element theory, we are left with no account of the myriad cases where a customary rule which had previously been identified is now being interpreted. These include both cases where a court has first identified a CIL rule by looking at practice and opinio juris and has then proceeded to interpret it (see Canada v Edelson), and cases where a court simply interpreted an existing customary rule (see Public Committee against Torture or Former consular employee v Croatia). In this context, it also seems somewhat contradictory to accept the ‘interpretation’ of state practice but not that of the rule, especially since several instances of the former may also be unwritten. Note as well, the use of teleological interpretation in several of these cases, which is always connected to the object and purpose of a rule, and not to the object and purpose of state practice. In this sense, while I agree with Sender and Wood that the ILC’s conclusions have given us ample guidance, we must be wary of relying on them as reason to dismiss further research on the subject. This is especially true, when states through their acts, be they judicial decisions or other (see, for instance, Germany’s recent explicit acceptance of and reference to interpretation of customary rules), have explicitly accepted and applied interpretation of customary rules.

A related consideration in this regard is that the identification of CIL may also include a degree of interpretive reasoning. The fourth post by Kristi How discusses the identification of CIL from the perspective of a governmental advisor. How points out that since the identification exercise is inductive in nature, there will necessarily be an element of subjective value judgment on the part of the advisor in determining the weight of the information before her. Depending on the factual background underpinning the search for a CIL rule, certain evidence could be ascribed more weight than other. Furthermore, How argues, the assessment of practice and opinio juris may be influenced by the aim of the advisor undertaking the identification exercise. This certainly describes a degree of interpretation that takes place at the stage of identification. To this, I would only add that this must be differentiated from the interpretation of a CIL rule stricto sensu, as the two differ with respect to both their content and their outcome. In the stage of CIL identification there may naturally be some rudimentary elements of an interpretative exercise taking place. The reasoning employed here is concerned with the relevance and weight to be given to evidence of state practice and opinio juris, and the outcome is a binary one–a CIL rule is either determined to exist or not. This is similar to what happens in the case of determining whether a treaty exists–some rudimentary interpretation occurs at that stage, but nobody suggests that due to it there is no need for treaty interpretation proper. Similarly, then, once a CIL rule is identified there may be, and often is, need for interpretation proper. The reasoning employed at the stage of interpretation is concerned with the scope and content of the CIL rule, and it may have a variety of outcomes depending on the rule being interpreted and the legal and factual circumstances it is being interpreted in.

It was a pleasure reading these four posts and engaging with their ideas. I hope that this symposium inspires colleagues to consider the idea of CIL interpretation further and brings about a more in-depth engagement with the subject.

Acknowledgments: This contribution is based on research conducted in the context of the project ‘The Rules of Interpretation of Customary International Law’ (‘TRICI-Law’). This project received funding from the European Research Council (‘ERC’) under the European Union’s Horizon 2020 Research and Innovation Programme (Grant Agreement No. 759728).