16 - Interpreting Customary International Law

Summary

Customary international law is one of the formal sources of international law and plays a pivotal role in the existence and functioning of the international legal system. Although for a rule of CIL to emerge a widespread, representative, constant and virtually uniform state practice is required, accompanied by the requisite opinio juris, that does not necessarily mean that CIL is a slow and archaic process, which has been overcome by extensive treaty-making. On the contrary, CIL remains a vital element in the corpus of international law that is open to refinement, clarification, development and evolution. This process does not happen only through the classical emergence and/or subsequent modification of the rule, but also and perhaps most importantly through the process of interpretation. This chapter demonstrates this by showing that CIL interpretation is neither problematic from a theoretical perspective, nor is it the only example of interpretation of unwritten rules. This is further reinforced by jurisprudence taken from both the domestic and the international legal system, although such interpretative exercises are not without their limits. What emerges from this analysis is that CIL interpretation, as Sur has beautifully noted, is of a negentropic nature that constantly nourishes and updates CIL.

1 Introduction

Oceans of ink have been spilt over both treaty interpretation and customary international law (CIL). Yet the point of convergence between these two areas, that is, CIL interpretation, remains somewhat woefully under-examined. The almost obsessive focus on the formation stage of CIL, with its two elements, state practice and opinio juris may have something to do with that. As perhaps does the fact, stemming from the above obsession, that CIL is often cursorily dismissed as not being interpretable. The present contribution aims to question these assumptions, and demonstrate that CIL interpretation is not only plausible, but has been occurring both in international and domestic legal systems. It is a process that is inextricably linked to the life cycle of every rule, irrespective of its source, and it is one that can also breathe life and ensure the relevance of rules across wide swathes of the temporal landscape.

Section 2 will start with an examination of some of the basic objections raised against the interpretability of CIL and will also investigate whether in international law there are other examples of non-written rules that are nonetheless accepted to be interpretable. Section 3 will dive into domestic and international legislation and case-law that evidence that CIL interpretation is actually occurring. Domestic law and case law will also be examined, as we often tend to forget that the interaction between the international and the domestic legal system is not one-way but rather an amphidromous one. In fact, domestic legal systems with a rich and much longer tradition than that of international law, may have significant insights to offer in how customary law (both domestic and international) functions. Section 3 will also highlight some key interpretative approaches that seem to emerge from the examined jurisprudence. This will lead us to Section 4, where the outer limits of such an interpretative exercise will be demarcated. As with any interpretation of any rule, so CIL interpretation should not be construed as a carte blanche to the judges, that allows then to substitute the states in the creation of norms. This section will focus on these limits, which if exceeded we transgress to judicial lawmaking. Section 5 will offer some concluding thoughts.

2 International Law’s Approach to Interpretation of Non-written Rules

The literature on CIL tends to be dominated by inquiries into the formative stage of CIL and/or whether the existing two-element model is a functional one or falls prey to inherent pitfalls. That is not to say that analysis on CIL interpretation is not present, with scholars arguing both against and in favour of CIL’s interpretability.¹ Let us, however, examine what the main arguments against the interpretability of CIL are.

Stemming from the doctrinal focus on the two-element approach, an argument often invoked against the interpretability of CIL is that ‘content merges with existence’, namely that the identification of CIL through a strict application of the two-element approach in and of itself satisfies the content-determinative aspect of interpretation, and thus there is no need for interpretation.² This approach, however, seems to accept as a given a degree of specificity and precision that even written texts and long-negotiated treaties are incapable of achieving. The requirements of widespread, representative, constant and uniform state practice accompanied by opinio juris would never be precise enough to account for newly emerging situations, that in any other case (and especially in the case of written instruments) would be easily addressed through the process of interpretation. Add to that the fact that CIL is often criticised for being vague,³ and it becomes evident that even more so in the case of CIL interpretation is a sine qua non, as it is the only process that allows for lifting this ‘penumbra of doubt’.⁴ This seems to be summed up by the International Court of Justice (ICJ) itself in the Gulf of Maine when it stated that

[a] body of detailed rules is not to be looked for in customary international law … It is therefore unrewarding … to look to general international law to provide a readymade set of rules that can be used for solving any delimitation problems that arise. A more useful course is to seek a better formulation of the fundamental norm, on which the Parties were fortunate enough to be agreed.⁵

In the same vein, Sur, more recently, in his General Course in the Hague Academy of International also reaffirmed the content-determinative importance of interpretation for CIL when he noted that ‘[i]nterpretation of customary rules allows the formulation of a statement that specifies their content and meaning’.⁶

The other main strand of objection to the interpretability of CIL is it being non-written. ‘[T]he irrelevance of linguistic expression excludes interpretation as a necessary operation in order to apply [customary rules].’⁷ But is this truly so? This would seem to be based on an understanding of interpretation as entirely based on text. Yet, a simple browsing of Articles 31–33 Vienna Convention on the Law of Treaties (VCLT) reveals a cornucopia of other non-textual elements that exist on par with the text, even more so if one considers the International Law Commission’s (ILC) ‘crucible approach’ to interpretation that these articles reflect. Second, let us consider the following scenario. There are two identical rules at a particular point in time. One is a CIL rule, and the other one is a rule that exists in a codification treaty. The latter rule would be open to interpretation. So the interpreter would be able to refer to the object and purpose, to intention, to other relevant rules and all the other elements enshrined in Articles 31–33 VCLT. The former rule’s content, on the other hand, if one accepts the argument that the non-written nature of CIL bars it from being interpretable, would have to be determined solely on the model of state practice and opinio juris.⁸ The end result being, the written rule having the ability to be further content-determined through the process of interpretation, whereas the CIL rule would not, and situations that could be addressed through the written rule, through a teleological or evolutive interpretation, would remain outside the scope of the CIL rule, despite the fact that our original starting point was that both these rules were identical. This seems to be an illogical result, that militates in favour of the interpretability CIL.

Logical exercises are not the only reason why the linguistic irrelevancy of CIL is not a bar to its interpretability. Interpretation of non-written elements that, nonetheless, create binding rules of international law are nihil novum sub sole. Oral treaties, also known as verbal treaties or verbal/oral agreements⁹ are one such example. The binding character of oral agreements has been recognized in international jurisprudence, as for instance in Mavrommatis Jerusalem Concessions,¹⁰ and did not cause any waves during the preparatory work of the VCLT as can also be seen by the final adopted text.¹¹

Article 2(1)(a) VCLT defines treaties as ‘an international agreement concluded between States in written form and governed by international law, whether embodied in a single instrument or in two or more related instruments and whatever its particular designation’.¹² However, that is not to say that the VCLT rejects the potentiality of existence of other types of treaties that do not meet the strict criteria of Article 2(1)(a). So much so in fact, that Article 3 is explicitly devoted to this as it stipulates that the fact that the VCLT ‘does not apply to international agreements concluded between States and other subjects of international law or between such other subjects of international law, or to international agreements not in written form’ does not affect either the legal force of such agreements or the application to them of customary rules relating to the law of treaties.¹³ The reason why the VCLT focused only on written treaties was merely in the interest of clarity and simplicity.¹⁴

Although the VCLT seems to have taken a rather expansive interpretation of how strict the ‘written form’ requirement should be, by including even oral agreements that are evidenced in writing, as in the case of an oral agreement that is documented by a third party, which has so been authorized by the parties to the agreement.¹⁵ However, if no such authorized transcription exists, for example as in the case of (video)-taped understandings or oral answers to written proposals, these still remain oral agreements.¹⁶

Oral agreements were more common in the pre-Westphalian era, but have unsurprisingly been on the decline in the last two centuries, not only, as Schmalenbach rightly points out, due to the existence of an obligation to register treaties¹⁷ but also to ensure greater clarity and certainty as to their international obligations.¹⁸ That is not to say that oral agreements do not emerge in international practice, as evidenced by the famous 1919 Ihlen Declaration between the Ministers of Foreign Affairs of Norway and Denmark,¹⁹ and the telephone agreement of 1992 between the prime ministers of Denmark and Finland regarding the Great Belt Bridge.²⁰

The customary rules on the law of treaties apply to such oral agreements as long as they are not tied to the written form requirement and, since text is but one of the many elements to be taken into account during interpretation, this would also include the rules of interpretation.²¹

In the same vein, another set of non-written acts that have raised no concerns as to their interpretability are unilateral acts of states capable of creating international obligations. From 1996 to 2006 the ILC worked on the topic of ‘Unilateral Acts of States’ and their capacity to create binding international obligations. In its Guiding Principle 5, the ILC specified that the form of the declaration, oral or in writing, was immaterial.²² Thirty years earlier the ICJ had stated the same thing in Nuclear Tests; ‘[w]hether a statement is made orally or in writing makes no essential difference … Thus the question of form is not decisive.’²³ What makes this relevant for the purposes of our analysis is that the ILC also adopted rules of interpretation applicable to such unilateral declarations, again without making any distinction as to whether the declaration is oral or in writing.²⁴

As the previous examples demonstrate, interpretation of non-written rules is neither prohibited nor a first for international law. But even the non-written (linguistic irrelevance) objection is not as clear cut as one would think. Alland referring also to Müller and Kolb underscores this point, when he writes that ‘it is difficult to think of a custom independently of any linguistic expression, of any “lexical garment”, to use [Müller’s] wonderful expression. In fact, even if we do not put the customary rule in a codification convention, it must be formulated and, from this formulation, it may appear that we are interpreting linguistic signs expressing a customary rule.’²⁵ This is also something that we shall see in the next sections being a common pattern in the interpretation of CIL by international and domestic courts.

3 CIL Interpretation in International and Domestic Legal Systems

As shown in the previous section, interpretation of non-written rules is not something that international law is unfamiliar with. But is CIL interpretation something that is actually taking place either in the international or domestic legal systems? In order to answer this, we will now turn our attention to the practice of international and domestic courts to examine whether when applying CIL or domestic customary law they engage in a process of interpretation. This issue is also touched upon in a number of other chapters in this volume.²⁶ To avoid overlap only a few additional cases will be mentioned here, highlighting some common interpretative patterns; the reader however is strongly encouraged to consult those chapters as well in order to get a complete picture of the pervasiveness of CIL interpretation in both the international and domestic legal arena.

3.1 The Interpretability of CIL as Evidenced in Written Instruments

Where one could first look for acknowledgement of the interpretability of CIL is within instruments regulating the judicial process or identifying the sources of applicable domestic or international law. Article 21 of the Rome Statute,²⁷ for instance, which sets out the law applicable by the International Criminal Court (ICC) makes no distinction between the various sources of law (treaties, custom and general principles). In fact, Article 21(2) clearly spells out that ‘[t]he Court may apply principles and rules of law as interpreted in its previous decisions’, while Article 21(3) builds on this uniform approach when it simply refers to ‘[t]he application and interpretation of law pursuant to this article’²⁸ without finding any reason to suggest that certain types of rules are not open to interpretation and should be approached differently. The ICC has also followed this line of reasoning when it refers to principles and rules as having been interpreted in the ICC’s previous judgments.²⁹

A more explicit acknowledgement of the interpretability of CIL can also be found in the Statutes of the ICJ and the Permanent Court of International Justice (PCIJ) and their preparatory work. Article 36 of the ICJ Statute, which was almost verbatim reproduced from that of the PCIJ Statute, refers to the jurisdiction of the court in all legal disputes concerning ‘a. the interpretation of a treaty’ and ‘b. any question of international law’. One could reasonably arrive at the conclusion that the explicit avoidance of reference to the word ‘interpretation’ in sub-paragraph (b), was an intentional one and that this would indicate that the drafters of the PCIJ Statute, the Advisory Committee of Jurists, took a firm position on the matter through this differentiated wording. However, if one looks closely at the travaux préparatoires the true reason for this linguistic choice is revealed. What became Article 36 of the PCIJ Statute was based on a draft by Lord Phillimore.³⁰ While discussing this, another member of the Advisory Committee, Ricci-Busatti, suggested that the proposed version was problematic and should be amended so as to read ‘a. the interpretation or application of a treaty; b. the interpretation or application of a general rule of international law’.³¹ No member raised any objections as to the validity of Ricci-Busatti’s proposal;³² on the contrary some members, such as de la Pradelle and Hagerup, were vocal as to the linguistic defects of Lord Phillimore’s version, and the superiority of Ricci-Busatti’s proposal.³³ Despite this, the original version remained in place, and the reason was that the language used was copied directly from Article 13 of the Covenant of the League of Nations and the drafters wanted to ensure linguistic continuity as to the expressions used.³⁴ This notwithstanding, the fact remains that not only interpretation of CIL was actually proposed to be included in the text of the PCIJ Statute, but also it raised no objections from a theoretical standpoint, that is, that CIL is non-interpretable, and its eventual non-inclusion was based solely on linguistic continuity concerns, but not on substantive objections.

The examples offered so far demonstrate that in the statutes of international courts and tribunals and their preparatory work indicia can be found that demonstrate that interpretation is a process recognised by the drafters as an inherent element of the application of both conventional and customary rules. Similar evidence can also be traced within constitutions, legislation and codes of domestic legal systems. One point that has to be made here is that in domestic legal systems there is usually one or two caveats often introduced with respect to customary law, be it domestic or international. As with treaty interpretation, interpretation of customary law has certain limits. Although the limits to CIL interpretation will be analysed infra in Section 4, here it is worth noting that an approach that appears with relative frequency in domestic legal systems is that an interpretation or existence of a customary rule cannot conflict with a written rule of domestic law, and in case of such conflict the written rule prevails.³⁵ Of import here is that before acknowledging the existence of a conflict between rules, domestic courts always attempt to harmonize the content of the rules through interpretation,³⁶ a process not unique to domestic courts but equally applied by international courts and recognised by the ILC as well.³⁷

Apart from this ‘harmonisation through interpretation’ that we will see more of in Section 3.2, a more explicit acknowledgement of CIL interpretation can also be seen, for instance, in the case of Article 559(1) of the Greek Code of Civil Procedure. According to that Article ‘[a]n appeal is allowed only 1) if a rule of substantive law has been violated, which includes the rules of interpretation of legal acts, regardless of whether this entails a law or custom, Greek or foreign, of domestic or international law’.³⁸ This provision and ground of appeal has in fact been interpreted by the Supreme Civil and Criminal Court of Greece in the following manner: ‘The legal rule is violated, if it is not applied, … as well as if it is applied incorrectly … and the violation is manifested either by false interpretation [misinterpretation] or by incorrect application.’³⁹ It is of note that misinterpretation is one of the manifestations of violation of the rule, and neither the Greek Code of Civil Procedure nor the relevant jurisprudence differentiate in their approach on whether the rule is one of written law or a customary rule.⁴⁰

3.2 Patterns of CIL Interpretability in International and Domestic Case Law

Evidence from statutes and domestic pieces of legislation are useful, but not entirely decisive of the ubiquity of CIL interpretation. For this we shall now turn our attention to case law. The former Chess World Champion Mikhail Botvinik is often credited with the chess aphorism, ‘every Russian school boy knows’, which is used within chess circles to denote some basic knowledge that everyone has. Mutatis mutandis ‘every international law student knows’ that when talking about CIL two sets of cases are the ones most often used, Nicaragua and North Sea Continental Shelf, with the latter being the landmark case for the two-element approach of state practice and opinio juris. Ironically enough, even in these bastions of the classical two-element approach, one can find references to CIL interpretation. The Nicaragua case seems to be open to the interpretability of CIL, when the court opines that ‘[r]ules which are identical in treaty law and in customary international law are also distinguishable by reference to the methods of interpretation and application’.⁴¹ However, in the North Sea Continental Shelf cases this is much more explicit, when Judge Tanaka has the following to say regarding CIL interpretation: ‘Customary law, being vague and containing gaps compared with written law, requires precision and completion about its content. This task, in its nature being interpretative, would be incumbent upon the Court. The method of logical and teleological interpretation can be applied in the case of customary law as in the case of written law’.⁴² Although it is unclear the exact line, if any, drawn by Tanaka between logical and teleological interpretation,⁴³ the use of ‘logical interpretation’ is not so foreign. This word may not have found its way in the text of Articles 31–33 VCLT, but it was used in early jurisprudence and in the early codification attempts of the law of treaties and the rules of treaty interpretation. For instance, Fiore’s Draft Code suggested that treaty interpretation could be either grammatical or logical, a slightly different structure than that of Tanaka. In the rules of logical interpretation, one could find recourse to, for instance, intention of the parties, context, contra proferentem, equity, ut res magis valeat quam pereat, systemic/harmonious interpretation and teleology.⁴⁴

So let us examine if any of these interpretative tools emerge in cases where courts have been called to apply CIL. In Section 2, we discussed Alland’s view that CIL is always shrouded in a ‘lexical garment’. The practice of courts and tribunals, both international and domestic, seems to utilize this to compensate for the non-existence of a written rule in the case of CIL. Since textual interpretation stricto sensu is not possible, what they do is refer to documents which are allegedly reflective of CIL.⁴⁵ If one were to try and find an analogy with the rules of treaty interpretation, this would be akin to an application of the principle of systemic integration or in pari materia interpretation if the documents referred to were treaties.⁴⁶ This attempt at a ‘by proxy’/hybrid textual interpretation of CIL is sometimes taken even further, when courts use not only the language of the relevant provision that reflects CIL, but also other provisions of the referred instrument, as a type of context (again by proxy) to determine the meaning of the CIL rule.⁴⁷

However, that is not to say that reference to other treaties, CIL rules or general principles only happens in this context, that is, in a ‘by proxy’ textual interpretation. There are also several instances where courts and tribunals have interpreted CIL by reference to its normative environment in the traditional ‘systemic integration’ fashion.⁴⁸ The Supreme Court of Italy in Ferrini v. Germany summarized this very concisely: ‘However, it is unquestionably true that similar criteria [i.e. reference to relevant rules] apply to the interpretation of customary norms, which like the others are part of a system and therefore may only be correctly understood in relation to other norms that form an integral part of the same legal system.’⁴⁹ This interpretative method is often used to ensure that the normative environment is taken into account in order to avoid conflict and ensure ‘harmonization through interpretation’, as can be easily seen in a string of domestic cases, where state immunity was counter-balanced, for example, with the protection of fundamental human rights/values, and the prohibition of torture.⁵⁰

Another dominant pattern emerging from domestic and international case law is reference to either the telos of the rule or its rationale.⁵¹ In Her Majesty the Queen in Right of Canada v. Edelson and others, for instance, the Supreme Court of Israel was called to identify the content of the CIL rule on state immunity and the criteria to be used in distinguishing between acta jure imperii and acta jure gestionis. In the ‘Comfort Women’ case, the South Koran District Court also had to tackle issues of state immunity but in the context of whether such immunity could be invoked for crimes against humanity committed during World War II. In both cases, the domestic courts relied on the reasons underlying the existence and functioning of the CIL on state immunity in order to come to conclusions as to the content of the rule.⁵²

An interesting tendency in CIL interpretation is also that the telos referred to is not necessarily that of the CIL rule alone. Sometimes, courts and tribunals based their teleological interpretation of the CIL rule on the telos of an entire area of international law.⁵³ In such instances, such a lato sensu teleological interpretation becomes very similar to systemic interpretation.

However, as in treaty interpretation, where various interpretative maxims and approaches not explicitly mentioned in the VCLT are often utilised, these also make their appearance in cases of CIL interpretation. Ut res magis valeat quam pereat and ad absurdum arguments make regular appearances in the reasoning of courts when they interpret CIL. In Her Majesty the Queen in Right of Canada v. Edelson and others, the court emphasized that the reason why the purpose criterion was not the appropriate one for distinguishing between acta jure imperii and acta jure gestionis was that it would end up negating the distinction between private and state acts.⁵⁴ In A v. Swiss Federal Public Prosecutor, the court held that ‘it would be both contradictory and futile if, on the one hand, we affirmed that we wanted to fight against these serious violations of the fundamental values of humanity, and, on the other hand, we allowed a broad interpretation of the rules of functional immunity’.⁵⁵ While in the Sea Shepherd case the district court did not mince its words on what it thought of a broad interpretation of ‘piracy’; ‘[a]mong other nonsensical results, Defendants’ interpretation would allow any seaman with a special affinity for a sea creature – say, a tuna – to state a piracy claim against a fisherman’.⁵⁶ Other cases have also referred to CIL as being open to evolutive interpretation,⁵⁷ or even more dubiously to in dubio mitius constructions⁵⁸ and presumptions that promote interpretations in favour of internal jurisdiction.⁵⁹

As can be seen from the previous analysis, the examples offered were not meant to be an exhaustive list but rather a demonstration of the occurrence of CIL interpretation across the board and the multifariousness of interpretative tools used, which are, however, familiar from treaty interpretation. It is also of note that several of the cited cases do not use just one interpretative method but a number of them, which again also coincides with the ILC’s view of interpretation as a holistic exercise.

4 Limits of CIL Interpretation

The fact that CIL is open to interpretation does not mean that judges have a carte blanche when engaging in such interpretative exercises. As with interpretation of treaties and of other instruments, so CIL interpretation cannot go beyond certain limits. Certain of these limits are common to all rules irrespective of the source from which they have emerged. It is to these limits that we shall turn our attention.

The first and foremost such limit is a system-oriented one, that is, one that is imposed by the system and its, admittedly limited, hierarchical structure. Any interpretation of a rule cannot be such that it would go against a rule of jus cogens.⁶⁰ This limit is a very logical one, and stems also from the very definition of jus cogens rules, being rules from which no derogation is possible. It is such a fundamental limit that it even found its way into the Institut de Droit International’s resolution on ‘Intertemporal Law’, where it was stated that: ‘States and other subjects of international law shall, however, have the power to determine by common consent the temporal sphere of application of norms, … subject to any imperative norm of international law which might restrict that power.’⁶¹ Of course, both the cases mentioned in the first footnote to this section and the Institut’s resolution were focused on treaties, however the rationale behind the acceptance of jus cogens as an interpretative limit is equally applicable to CIL rules and obligations emerging from unilateral acts of states.

This can be seen in the recent works of the ILC, both on ‘Identification of CIL’ and on ‘Jus Cogens’. With respect to the former, both the commentary to Draft Conclusion 1 and the text of Draft Conclusion 15 made a point of underscoring that these draft conclusions were ‘without prejudice to questions of hierarchy among rules of international law, including those concerning peremptory norms of general international law (jus cogens)’.⁶² This has more recently become even clearer through the conclusions proposed by the Drafting Committee on ‘Jus Cogens’. Draft Conclusion 14 clarifies that with respect to CIL no such rule may come into existence if it conflicts with a peremptory norm of general international law, and ‘ceases to exist if and to the extent that it conflicts with a new peremptory norm of general international law’.⁶³ So Draft Conclusion 14 covers both ends of the spectrum, emergence and termination of CIL rules, but what of its interpretation? Draft Conclusion 20, which deals with the interpretation and application of rules in a manner consistent with peremptory norms of general international law, provides the answer to that: ‘Where it appears that there may be a conflict between a peremptory norm of general international law (jus cogens) and another rule of international law, the latter is, as far as possible, to be interpreted and applied so as to be consistent with the former.’⁶⁴ Of particular note here is that Draft Conclusion 20 makes absolutely no distinction between rules on the basis of their source, but considers that an interpretation that ensures harmony with existing jus cogens rules is an interpretative limit for rules irrespective of the type of source from which they emerged.

The second limit is one that derives from the classical distinction between interpretation and revision/modification. In treaty interpretation, for instance, whereas interpretation aims to give flesh to the intention of the parties,⁶⁵ revision of a treaty falls outside its outer limits as it changes the content and identity of a rule in ways that could not be arrived at through a normal interpretative exercise. Because revision amounts to creating a new rule, as it exceeds the rule’s ‘natural limits’,⁶⁶ interpretation may never amount to a revision of the rule.⁶⁷ Treaty revision falls squarely within the exclusive competence of the parties to the treaty (or any body so authorized by the parties), not of the judges. Consequently, an interpretation that would lead to a revision of the rule, would be equivalent to the judges exercising a pouvoir de légiférer, a power that they have not been imbued with.⁶⁸ As Dupuy very eloquently put it, ‘[m]emory must remain loyal and not serve to rewrite history; a treaty belongs to its authors and not to the judge’.⁶⁹ The ILC also confirmed this recently through Draft Conclusion 7(3) on ‘Subsequent Agreements and Subsequent Practice in Relation to Interpretation of Treaties’.⁷⁰ According to the ILC, if the limits of interpretation are crossed, then we may be in the realm of treaty modification, although the ILC admitted that the lines may be difficult to draw and was agnostic as to whether modification of a treaty by subsequent practice of the parties was customary law.⁷¹

This differentiation between the existing rule and its modification/revision seems also to lie at the heart of the Hadžihasanović case. The tribunal, on the one hand, felt that there was no sufficient evidence of state practice and opinio juris to demonstrate that the existing content of the CIL rule on command responsibility, covered also situations where a change in the command structure had occurred, and therefore that any such reading/interpretation of the rule would amount to an unacceptable and impermissible revision/modification. A number of judges, on the other hand, were of the view that a teleological interpretation of the rule inexorably led to an inclusion of that situation within the regulatory framework of the rule.⁷²

The same line in the sand distinction between interpretation and revision/modification seems to be the driving force behind judge ad hoc Kreća’s analysis in the Croatia-Serbia Genocide case as well. His main objection to certain of the pronouncements of the ICTY and its ‘interpretation’ of CIL was that the methods used were incoherent and subjective, and that the establishment of the content of a CIL rule resembled ‘a quasi-customary law exercise based on deductive reasoning driven by meta-legal and extra-legal principles … [that] has resulted in judicial law-making through purposive, adventurous interpretation’.⁷³ Leaving aside that judge ad hoc Kreća also recognises the interpretability of CIL, his objection stems not from the interpretative exercise per se and the use of teleological interpretation, but rather from the fact that such an interpretation is not interpretation in the proper sense, but rather a revision of the rule, which amounts to an exercise by the judges of a pouvoir de légiférer (judicial lawmaking). In essence, this objection is an affirmation of the second limit of CIL interpretation, and interpretation in general.

Another limit that needs to be examined in this context is that any interpretation ‘can only apply in the observation of the general rule of interpretation laid down in Article 31 of the Vienna Convention on the Law of Treaties’.⁷⁴ This was identified by Judge Bedjaoui in the Gabčíkovo-Nagymaros Project case in the context of an evolutive interpretation of a written instrument, but it applies equally in the case of CIL. The following cases may help illustrate this point.

On 8 January 2021 the Central District Court of Seoul issued its judgment, now final, regarding compensation of South Korean women, who had been forced into sexual slavery and euphemistically known as ‘comfort women’, during World War II. A key issue was whether state immunity could be upheld even in cases where grave crimes against humanity had been perpetrated.⁷⁵ Although, as analysed above in Section 3, the Central District Court also engaged in a logical and teleological interpretation of CIL, it based part of its reasoning on a somewhat ‘systemic-type’ of interpretation but of an inward focus, that is, it focused on the potential of harmonization or conflict of an expansive interpretation of state immunity with its domestic constitution. According to it,

if customary law is applied to exempt the Defendant from jurisdiction even in cases where the Defendant has committed grave crimes against humanity, it would be impossible to sanction a State for violating international conventions that prevent it from committing grave crimes against humanity against citizens of another state, thereby depriving victims of their right of access to courts guaranteed by the Constitution and not providing a remedy for their rights. Such results are unreasonable and unjust as they are not in accordance with the overall legal order that positions the Constitution as the highest norm.⁷⁶

Although the first part of this argument shows similarities with an ut res magis valeat approach to interpretation, the final part links it to its domestic legal order. Essentially, what the District Court of Seoul focused on was that: (a) an expansive interpretation of state immunity would lead to a non-prosecution of crimes against humanity and (b) such a result would be unreasonable as it would conflict with the right of access to courts guaranteed by the constitution. Consequently, the District Court of Seoul was of the view that a more restrictive interpretation of state immunity was the one that ensured both effectiveness and the harmony among the rules of its domestic legal order. What this boils down to is that the District Court of Seoul, following a mélange of ut res magis valeat quam pereat and ‘harmonious/systemic interpretation’ approaches, interpreted the CIL rule on state immunity in a way that did not allow for its invocation in situations of crimes against humanity. However, the crucial point is that the counterpart to the rule on state immunity against which ut res magis valeat and harmonious/systemic interpretation were evaluated were not other rules of international law but rather its own domestic law and in particular its own constitution.

The Supreme Court of Israel in Her Majesty the Queen in Right of Canada v. Edelson and others when discussing the criteria to be applied in distinguishing between acta jure imperii and acta jure gestionis also referred to its domestic legal order but with a slight twist compared to the previous case. The Supreme Court, although in earlier paragraphs engaged in a teleological interpretation of the CIL rule, it then felt the need to buttress its findings by reference to its domestic legal order, not as a way to avoid normative conflict, but rather as a way to fill a potential lacuna.

[P]ending the development of a standard international practice regarding this issue, it is inevitable that each State will apply its own locally accepted criteria in accordance with its existing national jurisprudence … It is incumbent upon us to formulate a distinction that accounts for basic values such as individual rights, equality before the law and the rule of law. This having been said, we will allow the foreign State to realize its sovereign objectives, without subjecting them to judicial review in a foreign state’s courts. The balance struck between these conflicting considerations is far from simple and is certainly not immutable. It would seem that, for the time being, it is sufficient to determine that, when in doubt, we must rule in favor of recognizing internal jurisdiction. In any case, the tendency should be towards restricting immunity. This is our practice regarding any domestic matter.⁷⁷

A final case that needs to be mentioned in this context is Sentenza No 238/2014, where the Italian Constitutional Court had to grapple with the aftermath of the Jurisdictional Immunities case.⁷⁸ This case is very interesting as the Italian Constitutional Court did not object to the ‘interpretation’⁷⁹ on jurisdictional immunities adopted by the ICJ as ‘[i]nternational custom is external to the Italian legal order, and its application by the government and/or the judge, as a result of the referral of Article 10, para 1 of the Constitution, must respect the principle of conformity, ie must follow the interpretation given in its original legal order, that is the international legal order’.⁸⁰ What it tried to do was determine whether the interpretation of the CIL rule given by the ICJ could be harmonized with the Italian constitutional order and its fundamental principles.⁸¹ The Constitutional Court came to the conclusion that this was not possible and that therefore the CIL rule as interpreted by the ICJ had not entered the Italian legal order, through Article 10 para 1 of the Italian Constitution, and, thus, did not have any effect therein.⁸² The Constitutional Court, then turned its attention to Article 1 of the Law of Adaptation No 848/1957, and declared it unconstitutional, insofar as it concerned the execution of Article 94 of the UN Charter, and that as well exclusively to the extent that it obliged Italian courts to comply with the ICJ judgment in Jurisdictional Immunities.⁸³ The manner in which the Italian Constitutional Court approached the issue of CIL rule on jurisdictional immunities bears similarities both with the Solange⁸⁴ and Kadi⁸⁵ cases. With Solange in the sense that it determines the applicability of the CIL rule in Italian domestic legal order by applying the ‘limit’ of the concordance with fundamental principles of the state’s own constitutional order,⁸⁶ and with Kadi in the sense that the Italian Constitutional Court avoided engaging directly with an interpretation of the CIL rule on jurisdictional immunities different from that given by the ICJ, but rather decided to focus on the unconstitutionality of two domestic laws, through which the ICJ judgment and its interpretation would have become effective in the Italian domestic legal order.

The aforementioned three cases are not entirely identical, as they cover a wide spectrum of situations where CIL rules and their interpretation were considered, ranging from an attempt to harmonize the rule with the constitutional order (Case No 2016 Ga-Hap 505092), to filling lacunae of the CIL rule by reference to the domestic legal order (Her Majesty the Queen in Right of Canada v. Edelson and others) and including the CIL rule not entering the legal order as it cannot be harmonized with the limit of fundamental constitutional principles (Sentenza No 238/2014). These differences aside, a common thread remains an attempt at content-determination⁸⁷ of the CIL rule by reference to the state’s own domestic legal system. This from an internal, domestic-oriented point of view may not be as problematic,⁸⁸ although this is not to say that such an approach is entirely problem-free. This can be seen from the fact that a CIL rule should be interpreted using the rules/methods endemic to that international legal order. While this point was rightly so in Sentenza No 238/2014 it was not resorted to in the other two cases we discussed. This point also highlights why, from an international perspective, an interpretative approach to CIL focusing only on the domestic legal system of one state raises serious concerns. In all the cases mentioned above, the point being made was an effort to achieve a harmonious interpretation, that by taking into account other relevant rules would ensure that a normative conflict would be avoided. What this amounts to is an attempt at applying the principle of systemic integration in the context of CIL interpretation. However, the system of a CIL rule would refer to international rules (treaties, custom, general principles), but not to domestic rules of a single state. The only potential scenario where domestic rules may come into play is if an argument could be made that these reflected a ‘general principle’ shared by domestic legal systems. Leaving aside the issues of which domestic legal systems need to be considered,⁸⁹ by any stretch of imagination considering just one legal system would not be enough. Ryngaert calls this approach a ‘reverse’ consistent interpretation, and rightly points out the fact that it is a misapplication or disregard of the interpretative methods of international law.⁹⁰ Such an approach, thus, at least from the international perspective, seems to go against the limit of following the rules of interpretation.

As a final thought, it has to be noted that several of the cases cited in this section were also mentioned in Section 3. This is not surprising. In fact, it is demonstrative of why this discussion on CIL interpretation is not only inevitable but quintessential. The same way that the discussion on the rules of treaty interpretation helped and continues to help streamline and clarify the interpretative exercise and led to a common language being used, so can this occur with respect to CIL interpretation.

5 Conclusion

Customary international law is one of the formal sources of international law and plays a pivotal role in the existence and functioning of the international legal system. Although for a rule of CIL to emerge a widespread, representative, constant and virtually uniform state practice is required, accompanied by thе requisite opinio juris, that does not necessarily mean that CIL is a slow and archaic process, that has been overcome by extensive treaty-making. On the contrary, CIL remains a vital element in the corpus of international law, that is open to refinement, clarification, development and evolution. This process does not happen only through the classical emergence and/or subsequent modification of the rule, but also and perhaps most importantly through the process of interpretation.

In the previous sections what was shown was that not only is CIL interpretable (as are other non-written rules), but also that such an interpretation has and continues to occur with frequency in courts across different international legal regimes and different legal systems. Of course, the variety of interpretative approaches and the differences in the language/terminology used is not something unexpected. After all, if one examines the jurisprudence pre-VCLT, they would reach the same conclusion. But that is why further explorations and increased awareness of CIL interpretation is the key to further clarifying and refining the CIL interpretative process and prompt judicial bodies to be aware of and provide more clearly reasoned explications of the manner in which they interpret CIL.

As Sur very beautifully put it, CIL interpretation and its exploration is vital because whereas treaty interpretation is entropic, ‘[t]he interpretation of custom is creative or negentropic [i.e., reduces entropy], because it constantly nourishes and updates it [i.e., CIL], softening the distinction between formation and application’.⁹¹ Interpretation has, continues and will always be an integral part of the life cycle of CIL,⁹² or in simple terms, CIL will never walk alone.